����JFIFXX�����    $.' ",#(7),01444'9=82<.342  2!!22222222222222222222222222222222222222222222222222����"��4�� ���,�PG"Z_�4�˷����kjز�Z�,F+��_z�,�© �����zh6�٨�ic�fu���#ډb���_�N�?��wQ���5-�~�I���8����TK<5o�Iv-�����k�_U_�����~b�M��d����Ӝ�U�Hh��?]��E�w��Q���k�{��_}qFW7HTՑ��Y��F�?_�'ϔ��_�Ջt��=||I ��6�έ"�����D���/[�k�9���Y�8ds|\���Ҿp6�Ҵ���]��.����6�z<�v��@]�i%��$j��~�g��J>��no����pM[me�i$[����s�o�ᘨ�˸ nɜG-�ĨU�ycP�3.DB�li�;��hj���x7Z^�N�h������N3u{�:j�x�힞��#M&��jL P@_���� P��&��o8������9�����@Sz6�t7#O�ߋ �s}Yf�T���lmr����Z)'N��k�۞p����w\�Tȯ?�8`�O��i{wﭹW�[�r�� ��Q4F�׊���3m&L�=��h3����z~��#�\�l :�F,j@�� ʱ�wQT����8�"kJO���6�֚l����}���R�>ډK���]��y����&����p�}b��;N�1�m�r$�|��7�>e�@B�TM*-iH��g�D�)� E�m�|�ؘbҗ�a��Ҿ����t4���o���G��*oCN�rP���Q��@z,|?W[0�����:�n,jWiE��W��$~/�hp\��?��{(�0���+�Y8rΟ�+����>S-S����VN;�}�s?.����� w�9��˟<���Mq4�Wv'��{)0�1mB��V����W[�����8�/<� �%���wT^�5���b��)iM� pg�N�&ݝ��VO~�q���u���9� ����!��J27����$O-���! �:�%H��� ـ����y�ΠM=t{!S�� oK8������t<����è:a������[�����ա�H���~��w��Qz`�po�^ ����Q��n� �,uu�C�$ ^���,������8�#��:�6��e�|~���!�3�3.�\0��q��o�4`.|� ����y�Q�`~;�d�ׯ,��O�Zw�������`73�v�܋�<���Ȏ�� ـ4k��5�K�a�u�=9Yd��$>x�A�&�� j0� ���vF��� Y�|�y��� ~�6�@c��1vOp�Ig����4��l�OD���L����� R���c���j�_�uX6��3?nk��Wy�f;^*B� ��@�~a�`��Eu������+���6�L��.ü>��}y���}_�O�6�͐�:�YrG�X��kG�����l^w���~㒶sy��Iu�!� W ��X��N�7BV��O��!X�2����wvG�R�f�T#�����t�/?���%8�^�W�aT��G�cL�M���I��(J����1~�8�?aT ���]����AS�E��(��*E}� 2��#I/�׍qz��^t�̔���b�Yz4x���t�){ OH��+(E��A&�N�������XT��o��"�XC��'���)}�J�z�p� ��~5�}�^����+�6����w��c��Q�|Lp�d�H��}�(�.|����k��c4^�"�����Z?ȕ ��a<�L�!039C� �Eu�C�F�Ew�ç ;�n?�*o���B�8�bʝ���'#Rqf���M}7����]����s2tcS{�\icTx;�\��7K���P���ʇ Z O-��~��c>"��?�������P��E��O�8��@�8��G��Q�g�a�Վ���󁶠�䧘��_%#r�>�1�z�a��eb��qcPѵ��n���#L��� =��׀t� L�7�`��V���A{�C:�g���e@�w1 Xp3�c3�ġ����p��M"'-�@n4���fG��B3�DJ�8[Jo�ߐ���gK)ƛ��$���� ���8�3�����+���� �����6�ʻ���� ���S�kI�*KZlT _`���?��K����QK�d����B`�s}�>���`��*�>��,*@J�d�oF*����弝��O}�k��s��]��y�ߘ��c1G�V���<=�7��7����6�q�PT��tXԀ�!9*4�4Tހ3XΛex�46���Y��D ����� �BdemDa����\�_l,��G�/���֌7���Y�](�xTt^%�GE�����4�}bT���ڹ�����;Y)���B�Q��u��>J/J �⮶.�XԄ��j�ݳ�+E��d ��r�5�_D�1 ��o�� �B�x�΢�#���<��W�����8���R6�@g�M�.��� dr�D��>(otU��@x=��~v���2� ӣ�d�oBd��3�eO�6�㣷�����ݜ6��6Y��Qz`��S��{���\P�~z m5{J/L��1������<�e�ͅPu�b�]�ϔ���'������f�b� Zpw��c`"��i���BD@:)ִ�:�]��hv�E�w���T�l��P���"Ju�}��وV J��G6��. J/�Qgl߭�e�����@�z�Zev2u�)]կ�����7x���s�M�-<ɯ�c��r�v�����@��$�ޮ}lk���a���'����>x��O\�ZFu>�����ck#��&:��`�$�ai�>2Δ����l���oF[h��lE�ܺ�Πk:)���`�� $[6�����9�����kOw�\|���8}������ބ:��񶐕��I�A1/�=�2[�,�!��.}gN#�u����b��� ~��݊��}34q����d�E��Lc��$��"�[q�U�硬g^��%B �z���r�pJ�ru%v\h1Y�ne`ǥ:g���pQM~�^�Xi� ��`S�:V29.�P���V�?B�k�� AEvw%�_�9C�Q����wKekPؠ�\�;Io d�{ ߞo�c1eP����\� `����E=���@K<�Y���eڼ�J���w����{av�F�'�M�@/J��+9p���|]�����Iw &`��8���&M�hg��[�{��Xj��%��Ӓ�$��(����ʹN���<>�I���RY���K2�NPlL�ɀ)��&e����B+ь����( � �JTx���_?EZ� }@ 6�U���뙢ط�z��dWI�n` D����噥�[��uV��"�G&Ú����2g�}&m��?ċ�"����Om#��������� ��{�ON��"S�X��Ne��ysQ���@Fn��Vg���dX�~nj�]J�<�K]:��FW��b�������62�=��5f����JKw��bf�X�55��~J �%^����:�-�QIE��P��v�nZum� z � ~ə ���� ���ة����;�f��\v���g�8�1��f24;�V���ǔ�)����9���1\��c��v�/'Ƞ�w�������$�4�R-��t���� e�6�/�ġ �̕Ecy�J���u�B���<�W�ַ~�w[B1L۲�-JS΂�{���΃������A��20�c#��@ 0!1@AP"#2Q`$3V�%45a6�FRUq��� ����^7ׅ,$n�������+��F�`��2X'��0vM��p�L=������5��8������u�p~���.�`r�����\���O��,ư�0oS ��_�M�����l���4�kv\JSd���x���SW�<��Ae�IX����������$I���w�:S���y���›R��9�Q[���,�5�;�@]�%���u�@ *ro�lbI �� ��+���%m:�͇ZV�����u�̉����θau<�fc�.����{�4Ա� �Q����*�Sm��8\ujqs]{kN���)qO�y�_*dJ�b�7���yQqI&9�ԌK!�M}�R�;������S�T���1���i[U�ɵz�]��U)V�S6���3$K{�ߊ<�(� E]Զ[ǼENg�����'�\?#)Dkf��J���o��v���'�%ƞ�&K�u�!��b�35LX�Ϸ��63$K�a�;�9>,R��W��3�3� d�JeTYE.Mϧ��-�o�j3+y��y^�c�������VO�9NV\nd�1 ��!͕_)a�v;����թ�M�lWR1��)El��P;��yوÏ�u 3�k�5Pr6<�⒲l�!˞*��u־�n�!�l:����UNW ��%��Chx8vL'��X�@��*��)���̮��ˍ��� ���D-M�+J�U�kvK����+�x8��cY������?�Ԡ��~3mo��|�u@[XeY�C�\Kp�x8�oC�C�&����N�~3-H���� ��MX�s�u<`���~"WL��$8ξ��3���a�)|:@�m�\���^�`�@ҷ)�5p+��6���p�%i)P M���ngc�����#0Aruz���RL+xSS?���ʮ}()#�t��mˇ!��0}}y����<�e� �-ή�Ԩ��X������ MF���ԙ~l L.3���}�V뽺�v�����멬��Nl�)�2����^�Iq��a��M��qG��T�����c3#������3U�Ǎ���}��לS�|qa��ڃ�+���-��2�f����/��bz��ڐ�� �ݼ[2�ç����k�X�2�* �Z�d���J�G����M*9W���s{��w���T��x��y,�in�O�v��]���n����P�$�JB@=4�OTI�n��e�22a\����q�d���%�$��(���:���: /*�K[PR�fr\nڙdN���F�n�$�4�[�� U�zƶ����� �mʋ���,�ao�u 3�z� �x��Kn����\[��VFmbE;�_U��&V�Gg�]L�۪&#n%�$ɯ�dG���D�TI=�%+AB�Ru#��b4�1�»x�cs�YzڙJG��f��Il��d�eF'T� iA��T���uC�$����Y��H?����[!G`}���ͪ� �纤Hv\������j�Ex�K���!���OiƸ�Yj�+u-<���'q����uN�*�r\��+�]���<�wOZ.fp�ێ��,-*)V?j-kÊ#�`�r��dV����(�ݽBk�����G�ƛk�QmUڗe��Z���f}|����8�8��a���i��3'J�����~G_�^���d�8w������ R�`(�~�.��u���l�s+g�bv���W���lGc}��u���afE~1�Ue������Z�0�8�=e�� f@/�jqEKQQ�J��oN��J���W5~M>$6�Lt�;$ʳ{���^��6�{����v6���ķܰg�V�cnn �~z�x�«�,2�u�?cE+Ș�H؎�%�Za�)���X>uW�Tz�Nyo����s���FQƤ��$��*�&�LLXL)�1�" L��eO��ɟ�9=���:t��Z���c��Ž���Y?�ӭV�wv�~,Y��r�ۗ�|�y��GaF�����C�����.�+� ���v1���fήJ�����]�S��T��B��n5sW}y�$��~z�'�c ��8 ��� ,! �p��VN�S��N�N�q��y8z˱�A��4��*��'������2n<�s���^ǧ˭P�Jޮɏ�U�G�L�J�*#��<�V��t7�8����TĜ>��i}K%,���)[��z�21z ?�N�i�n1?T�I�R#��m-�����������������1����lA�`��fT5+��ܐ�c�q՝��ʐ��,���3�f2U�եmab��#ŠdQ�y>\��)�SLY����w#��.���ʑ�f��� ,"+�w�~�N�'�c�O�3F�������N<���)j��&��,-� �љ���֊�_�zS���TǦ����w�>��?�������n��U仆�V���e�����0���$�C�d���rP �m�׈e�Xm�Vu� �L��.�bֹ��� �[Դaզ���*��\y�8�Է:�Ez\�0�Kq�C b��̘��cө���Q��=0Y��s�N��S.���3.���O�o:���#���v7�[#߫ ��5�܎�L���Er4���9n��COWlG�^��0k�%<���ZB���aB_���������'=��{i�v�l�$�uC���mƎҝ{�c㱼�y]���W�i ��ߧc��m�H� m�"�"�����;Y�ߝ�Z�Ǔ�����:S#��|}�y�,/k�Ld� TA�(�AI$+I3��;Y*���Z��}|��ӧO��d�v��..#:n��f>�>���ȶI�TX��� 8��y����"d�R�|�)0���=���n4��6ⲑ�+��r<�O�܂~zh�z����7ܓ�HH�Ga롏���nCo�>������a ���~]���R���̲c?�6(�q�;5%� |�uj�~z8R=X��I�V=�|{v�Gj\gc��q����z�؋%M�ߍ����1y��#��@f^���^�>N�����#x#۹��6�Y~�?�dfPO��{��P�4��V��u1E1J �*|���%���JN��`eWu�zk M6���q t[�� ��g�G���v��WIG��u_ft����5�j�"�Y�:T��ɐ���*�;� e5���4����q$C��2d�}���� _S�L#m�Yp��O�.�C�;��c����Hi#֩%+) �Ӎ��ƲV���SYź��g |���tj��3�8���r|���V��1#;.SQ�A[���S������#���`n�+���$��$I �P\[�@�s��(�ED�z���P��])8�G#��0B��[ى��X�II�q<��9�~[Z멜�Z�⊔IWU&A>�P~�#��dp<�?����7���c��'~���5 ��+$���lx@�M�dm��n<=e�dyX��?{�|Aef ,|n3�<~z�ƃ�uۧ�����P��Y,�ӥQ�*g�#먙R�\���;T��i,��[9Qi歉����c>]9�� ��"�c��P�� �Md?٥��If�ت�u��k��/����F��9�c*9��Ǎ:�ØF���z�n*�@|I�ށ9����N3{'��[�'ͬ�Ҳ4��#}��!�V� Fu��,�,mTIk���v C�7v���B�6k�T9��1�*l� '~��ƞF��lU��'�M ����][ΩũJ_�{�i�I�n��$���L�� j��O�dx�����kza۪��#�E��Cl����x˘�o�����V���ɞ�ljr��)�/,�߬h�L��#��^��L�ф�,íMƁe�̩�NB�L�����iL����q�}��(��q��6IçJ$�W�E$��:������=#����(�K�B����zђ <��K(�N�۫K�w��^O{!����)�H���>x�������lx�?>Պ�+�>�W���,Ly!_�D���Ō�l���Q�!�[ �S����J��1��Ɛ�Y}��b,+�Lo�x�ɓ)����=�y�oh�@�꥟/��I��ѭ=��P�y9��� �ۍYӘ�e+�p�Jnϱ?V\SO%�(�t� ���=?MR�[Ș�����d�/ ��n�l��B�7j� ��!�;ӥ�/�[-���A�>�dN�sLj ��,ɪv��=1c�.SQ�O3�U���ƀ�ܽ�E����������̻��9G�ϷD�7(�}��Ävӌ\�y�_0[w ���<΍>����a_��[0+�L��F.�޺��f�>oN�T����q;���y\��bՃ��y�jH�<|q-eɏ�_?_9+P���Hp$�����[ux�K w�Mw��N�ی'$Y2�=��q���KB��P��~������Yul:�[<����F1�2�O���5=d����]Y�sw:���Ϯ���E��j,_Q��X��z`H1,#II ��d�wr��P˂@�ZJV����y$�\y�{}��^~���[:N����ߌ�U�������O��d�����ؾe��${p>G��3c���Ė�lʌ�� ת��[��`ϱ�-W����dg�I��ig2��� ��}s ��ؤ(%#sS@���~���3�X�nRG�~\jc3�v��ӍL��M[JB�T��s3}��j�Nʖ��W����;7��ç?=X�F=-�=����q�ߚ���#���='�c��7���ڑW�I(O+=:uxq�������������e2�zi+�kuG�R��������0�&e�n���iT^J����~\jy���p'dtG��s����O��3����9* �b#Ɋ�� p������[Bws�T�>d4�ۧs���nv�n���U���_�~,�v����ƜJ1��s�� �QIz��)�(lv8M���U=�;����56��G���s#�K���MP�=��LvyGd��}�VwWBF�'�à �?MH�U�g2�� ����!�p�7Q��j��ڴ����=��j�u��� Jn�A s���uM������e��Ɔ�Ҕ�!)'��8Ϣ�ٔ��ޝ(��Vp���צ֖d=�IC�J�Ǡ{q������kԭ�߸���i��@K����u�|�p=..�*+����x�����z[Aqġ#s2a�Ɗ���RR�)*HRsi�~�a &f��M��P����-K�L@��Z��Xy�'x�{}��Zm+���:�)�) IJ�-i�u���� ���ܒH��'�L(7�y�GӜq���� j��� 6ߌg1�g�o���,kر���tY�?W,���p���e���f�OQS��!K�۟cҒA�|ս�j�>��=⬒��˧L[�� �߿2JaB~R��u�:��Q�] �0H~���]�7��Ƽ�I���(}��cq '�ήET���q�?f�ab���ӥvr� �)o��-Q��_'����ᴎo��K������;��V���o��%���~OK ����*��b�f:���-ťIR��`B�5!RB@���ï�� �u �̯e\�_U�_������� g�ES��3�������QT��a����x����U<~�c?�*�#]�MW,[8O�a�x��]�1bC|踤�P��lw5V%�)�{t�<��d��5���0i�XSU��m:��Z�┵�i�"��1�^B�-��P�hJ��&)O��*�D��c�W��vM��)����}���P��ܗ-q����\mmζZ-l@�}��a��E�6��F�@��&Sg@���ݚ�M����� ȹ 4����#p�\H����dYDo�H���"��\��..R�B�H�z_�/5˘����6��KhJR��P�mƶi�m���3�,#c�co��q�a)*Pt����R�m�k�7x�D�E�\Y�閣_X�<���~�)���c[[�BP����6�Yq���S��0����%_����;��Àv�~�| VS؇ ��'O0��F0��\���U�-�d@�����7�SJ*z��3n��y��P����O���������m�~�P�3|Y��ʉr#�C�<�G~�.,! ���bqx���h~0=��!ǫ�jy����l�O,�[B��~��|9��ٱ����Xly�#�i�B��g%�S��������tˋ���e���ې��\[d�t)��.+u�|1 ������#�~Oj����hS�%��i.�~X���I�H�m��0n���c�1uE�q��cF�RF�o���7� �O�ꮧ� ���ۛ{��ʛi5�rw?׌#Qn�TW��~?y$��m\�\o����%W� ?=>S�N@�� �Ʈ���R����N�)�r"C�:��:����� �����#��qb��Y�. �6[��2K����2u�Ǧ�HYR��Q�MV��� �G�$��Q+.>�����nNH��q�^��� ����q��mM��V��D�+�-�#*�U�̒ ���p욳��u:�������IB���m���PV@O���r[b= �� ��1U�E��_Nm�yKbN�O���U�}�the�`�|6֮P>�\2�P�V���I�D�i�P�O;�9�r�mAHG�W�S]��J*�_�G��+kP�2����Ka�Z���H�'K�x�W�MZ%�O�YD�Rc+o��?�q��Ghm��d�S�oh�\�D�|:W������UA�Qc yT�q������~^�H��/��#p�CZ���T�I�1�ӏT����4��"�ČZ�����}��`w�#�*,ʹ�� ��0�i��課�Om�*�da��^gJ݅{���l�e9uF#T�ֲ��̲�ٞC"�q���ߍ ոޑ�o#�XZTp����@ o�8��(jd��xw�]�,f���`~�|,s��^����f�1���t��|��m�򸄭/ctr��5s��7�9Q�4�H1꠲BB@l9@���C�����+�wp�xu�£Yc�9��?`@#�o�mH�s2��)�=��2�.�l����jg�9$�Y�S�%*L������R�Y������7Z���,*=�䷘$�������arm�o�ϰ���UW.|�r�uf����IGw�t����Zwo��~5 ��YյhO+=8fF�)�W�7�L9lM�̘·Y���֘YLf�큹�pRF���99.A �"wz��=E\Z���'a� 2��Ǚ�#;�'}�G���*��l��^"q��+2FQ� hj��kŦ��${���ޮ-�T�٭cf�|�3#~�RJ����t��$b�(R��(����r���dx� >U b�&9,>���%E\� Ά�e�$��'�q't��*�א���ެ�b��-|d���SB�O�O��$�R+�H�)�܎�K��1m`;�J�2�Y~9��O�g8=vqD`K[�F)k�[���1m޼c��n���]s�k�z$@��)!I �x՝"v��9=�ZA=`Ɠi �:�E��)`7��vI��}d�YI�_ �o�:ob���o ���3Q��&D&�2=�� �Ά��;>�h����y.*ⅥS������Ӭ�+q&����j|UƧ����}���J0��WW< ۋS�)jQR�j���Ư��rN)�Gű�4Ѷ(�S)Ǣ�8��i��W52���No˓� ۍ%�5brOn�L�;�n��\G����=�^U�dI���8$�&���h��'���+�(������cȁ߫k�l��S^���cƗjԌE�ꭔ��gF���Ȓ��@���}O���*;e�v�WV���YJ\�]X'5��ղ�k�F��b 6R�o՜m��i N�i����>J����?��lPm�U��}>_Z&�KK��q�r��I�D�Չ~�q�3fL�:S�e>���E���-G���{L�6p�e,8��������QI��h��a�Xa��U�A'���ʂ���s�+טIjP�-��y�8ۈZ?J$��W�P� ��R�s�]��|�l(�ԓ��sƊi��o(��S0��Y� 8�T97.�����WiL��c�~�dxc�E|�2!�X�K�Ƙਫ਼�$((�6�~|d9u+�qd�^3�89��Y�6L�.I�����?���iI�q���9�)O/뚅����O���X��X�V��ZF[�یgQ�L��K1���RҖr@v�#��X�l��F���Нy�S�8�7�kF!A��sM���^rkp�jP�DyS$N���q��nxҍ!U�f�!eh�i�2�m���`�Y�I�9r�6� �TF���C}/�y�^���Η���5d�'��9A-��J��>{�_l+�`��A���[�'��յ�ϛ#w:݅�%��X�}�&�PSt�Q�"�-��\縵�/����$Ɨh�Xb�*�y��BS����;W�ջ_mc�����vt?2}1�;qS�d�d~u:2k5�2�R�~�z+|HE!)�Ǟl��7`��0�<�,�2*���Hl-��x�^����'_TV�gZA�'j� ^�2Ϊ��N7t�����?w�� �x1��f��Iz�C-Ȗ��K�^q�;���-W�DvT�7��8�Z�������� hK�(P:��Q- �8�n�Z���܃e貾�<�1�YT<�,�����"�6{/ �?�͟��|1�:�#g��W�>$����d��J��d�B��=��jf[��%rE^��il:��B���x���Sּ�1հ��,�=��*�7 fcG��#q� �eh?��2�7�����,�!7x��6�n�LC�4x��},Geǝ�tC.��vS �F�43��zz\��;QYC,6����~;RYS/6���|2���5���v��T��i����������mlv��������&� �nRh^ejR�LG�f���? �ۉҬܦƩ��|��Ȱ����>3����!v��i�ʯ�>�v��オ�X3e���_1z�Kȗ\<������!�8���V��]��?b�k41�Re��T�q��mz��TiOʦ�Z��Xq���L������q"+���2ۨ��8}�&N7XU7Ap�d�X��~�׿��&4e�o�F��� �H����O���č�c�� 懴�6���͉��+)��v;j��ݷ�� �UV�� i��� j���Y9GdÒJ1��詞�����V?h��l����l�cGs�ځ�������y�Ac�����\V3�? �� ܙg�>qH�S,�E�W�[�㺨�uch�⍸�O�}���a��>�q�6�n6����N6�q������N ! 1AQaq�0@����"2BRb�#Pr���3C`��Scst���$4D���%Td�� ?���N����a��3��m���C���w��������xA�m�q�m���m������$����4n淿t'��C"w��zU=D�\R+w�p+Y�T�&�պ@��ƃ��3ޯ?�Aﶂ��aŘ���@-�����Q�=���9D��ռ�ѻ@��M�V��P��܅�G5�f�Y<�u=,EC)�<�Fy'�"�&�չ�X~f��l�KԆV��?�� �W�N����=(� �;���{�r����ٌ�Y���h{�١������jW����P���Tc�����X�K�r��}���w�R��%��?���E��m�� �Y�q|����\lEE4���r���}�lsI�Y������f�$�=�d�yO����p�����yBj8jU�o�/�S��?�U��*������ˍ�0������u�q�m [�?f����a�� )Q�>����6#������� ?����0UQ����,IX���(6ڵ[�DI�MNލ�c&���υ�j\��X�R|,4��� j������T�hA�e��^���d���b<����n�� �즇�=!���3�^�`j�h�ȓr��jẕ�c�,ٞX����-����a�ﶔ���#�$��]w�O��Ӫ�1y%��L�Y<�wg#�ǝ�̗`�x�xa�t�w��»1���o7o5��>�m뭛C���Uƃߜ}�C���y1Xνm�F8�jI���]����H���ۺиE@I�i;r�8ӭ����V�F�Շ| ��&?�3|x�B�MuS�Ge�=Ӕ�#BE5G�����Y!z��_e��q�р/W>|-�Ci߇�t�1ޯќd�R3�u��g�=0 5��[?�#͏��q�cf���H��{ ?u�=?�?ǯ���}Z��z���hmΔ�BFTW�����<�q�(v� ��!��z���iW]*�J�V�z��gX֧A�q�&��/w���u�gYӘa���; �i=����g:��?2�dž6�ى�k�4�>�Pxs����}������G�9��3 ���)gG�R<>r h�$��'nc�h�P��Bj��J�ҧH� -��N1���N��?��~��}-q!=��_2hc�M��l�vY%UE�@|�v����M2�.Y[|y�"Eï��K�ZF,�ɯ?,q�?v�M 80jx�"�;�9vk�����+ ֧�� �ȺU��?�%�vcV��mA�6��Qg^M����A}�3�nl� QRN�l8�kkn�'�����(��M�7m9و�q���%ޟ���*h$Zk"��$�9��: �?U8�Sl��,,|ɒ��xH(ѷ����Gn�/Q�4�P��G�%��Ա8�N��!� �&�7�;���eKM7�4��9R/%����l�c>�x;������>��C�:�����t��h?aKX�bhe�ᜋ^�$�Iհ �hr7%F$�E��Fd���t��5���+�(M6�t����Ü�UU|zW�=a�Ts�Tg������dqP�Q����b'�m���1{|Y����X�N��b �P~��F^F:����k6�"�j!�� �I�r�`��1&�-$�Bevk:y���#yw��I0��x��=D�4��tU���P�ZH��ڠ底taP��6����b>�xa����Q�#� WeF��ŮNj�p�J* mQ�N����*I�-*�ȩ�F�g�3 �5��V�ʊ�ɮ�a��5F���O@{���NX��?����H�]3��1�Ri_u��������ѕ�� ����0��� F��~��:60�p�͈�S��qX#a�5>���`�o&+�<2�D����: �������ڝ�$�nP���*)�N�|y�Ej�F�5ټ�e���ihy�Z �>���k�bH�a�v��h�-#���!�Po=@k̆IEN��@��}Ll?j�O������߭�ʞ���Q|A07x���wt!xf���I2?Z��<ץ�T���cU�j��]��陎Ltl �}5�ϓ��$�,��O�mˊ�;�@O��jE��j(�ا,��LX���LO���Ц�90�O �.����a��nA���7������j4 ��W��_ٓ���zW�jcB������y՗+EM�)d���N�g6�y1_x��p�$Lv:��9�"z��p���ʙ$��^��JԼ*�ϭ����o���=x�Lj�6�J��u82�A�H�3$�ٕ@�=Vv�]�'�qEz�;I˼��)��=��ɯ���x �/�W(V���p�����$ �m�������u�����񶤑Oqˎ�T����r��㠚x�sr�GC��byp�G��1ߠ�w e�8�$⿄����/�M{*}��W�]˷.�CK\�ުx���/$�WPw���r� |i���&�}�{�X� �>��$-��l���?-z���g����lΆ���(F���h�vS*���b���߲ڡn,|)mrH[���a�3�ר�[1��3o_�U�3�TC�$��(�=�)0�kgP���� ��u�^=��4 �WYCҸ:��vQ�ר�X�à��tk�m,�t*��^�,�}D*� �"(�I��9R����>`�`��[~Q]�#af��i6l��8���6�:,s�s�N6�j"�A4���IuQ��6E,�GnH��zS�HO�uk�5$�I�4��ؤ�Q9�@��C����wp�BGv[]�u�Ov���0I4���\��y�����Q�Ѹ��~>Z��8�T��a��q�ޣ;z��a���/��S��I:�ܫ_�|������>=Z����8:�S��U�I�J��"IY���8%b8���H��:�QO�6�;7�I�S��J��ҌAά3��>c���E+&jf$eC+�z�;��V����� �r���ʺ������my�e���aQ�f&��6�ND��.:��NT�vm�<- u���ǝ\MvZY�N�NT��-A�>jr!S��n�O 1�3�Ns�%�3D@���`������ܟ 1�^c<���� �a�ɽ�̲�Xë#�w�|y�cW�=�9I*H8�p�^(4���՗�k��arOcW�tO�\�ƍR��8����'�K���I�Q�����?5�>[�}��yU�ײ -h��=��% q�ThG�2�)���"ו3]�!kB��*p�FDl�A���,�eEi�H�f�Ps�����5�H:�Փ~�H�0Dت�D�I����h�F3�������c��2���E��9�H��5�zԑ�ʚ�i�X�=:m�xg�hd(�v����׊�9iS��O��d@0ڽ���:�p�5�h-��t�&���X�q�ӕ,��ie�|���7A�2���O%P��E��htj��Y1��w�Ѓ!����  ���� ࢽ��My�7�\�a�@�ţ�J �4�Ȼ�F�@o�̒?4�wx��)��]�P��~�����u�����5�����7X ��9��^ܩ�U;Iꭆ 5 �������eK2�7(�{|��Y׎ �V��\"���Z�1� Z�����}��(�Ǝ"�1S���_�vE30>���p;� ΝD��%x�W�?W?v����o�^V�i�d��r[��/&>�~`�9Wh��y�;���R��� ;;ɮT��?����r$�g1�K����A��C��c��K��l:�'��3 c�ﳯ*"t8�~l��)���m��+U,z��`(�>yJ�?����h>��]��v��ЍG*�{`��;y]��I�T� ;c��NU�fo¾h���/$���|NS���1�S�"�H��V���T���4��uhǜ�]�v;���5�͠x��'C\�SBpl���h}�N����� A�Bx���%��ޭ�l��/����T��w�ʽ]D�=����K���ž�r㻠l4�S�O?=�k �M:� ��c�C�a�#ha���)�ѐxc�s���gP�iG��{+���x���Q���I= �� z��ԫ+ �8"�k�ñ�j=|����c ��y��CF��/��*9ж�h{ �?4�o� ��k�m�Q�N�x��;�Y��4膚�a�w?�6�>e]�����Q�r�:����g�,i"�����ԩA�*M�<�G��b�if��l^M��5� �Ҩ�{����6J��ZJ�����P�*�����Y���ݛu�_4�9�I8�7���������,^ToR���m4�H��?�N�S�ѕw��/S��甍�@�9H�S�T��t�ƻ���ʒU��*{Xs�@����f�����֒Li�K{H�w^���������Ϥm�tq���s� ���ք��f:��o~s��g�r��ט� �S�ѱC�e]�x���a��) ���(b-$(�j>�7q�B?ӕ�F��hV25r[7 Y� }L�R��}����*sg+��x�r�2�U=�*'WS��ZDW]�WǞ�<��叓���{�$�9Ou4��y�90-�1�'*D`�c�^o?(�9��u���ݐ��'PI&� f�Jݮ�������:wS����jfP1F:X �H�9dԯ���˝[�_54 �}*;@�ܨ�� ð�yn�T���?�ןd�#���4rG�ͨ��H�1�|-#���Mr�S3��G�3�����)�.᧏3v�z֑��r����$G"�`j �1t��x0<Ɔ�Wh6�y�6��,œ�Ga��gA����y��b��)��h�D��ß�_�m��ü �gG;��e�v��ݝ�nQ� ��C����-�*��o���y�a��M��I�>�<���]obD��"�:���G�A��-\%LT�8���c�)��+y76���o�Q�#*{�(F�⽕�y����=���rW�\p���۩�c���A���^e6��K������ʐ�cVf5$�'->���ՉN"���F�"�UQ@�f��Gb~��#�&�M=��8�ט�JNu9��D��[̤�s�o�~������ G��9T�tW^g5y$b��Y'��س�Ǵ�=��U-2 #�MC�t(�i� �lj�@Q 5�̣i�*�O����s�x�K�f��}\��M{E�V�{�υ��Ƈ�����);�H����I��fe�Lȣr�2��>��W�I�Ȃ6������i��k�� �5�YOxȺ����>��Y�f5'��|��H+��98pj�n�.O�y�������jY��~��i�w'������l�;�s�2��Y��:'lg�ꥴ)o#'Sa�a�K��Z� �m��}�`169�n���"���x��I ��*+� }F<��cГ���F�P�������ֹ*�PqX�x۩��,� ��N�� �4<-����%����:��7����W���u�`����� $�?�I��&����o��o��`v�>��P��"��l���4��5'�Z�gE���8���?��[�X�7(��.Q�-��*���ތL@̲����v��.5���[��=�t\+�CNܛ��,g�SQnH����}*F�G16���&:�t��4ُ"A��̣��$�b �|����#rs��a�����T�� ]�<�j��BS�('$�ɻ� �wP;�/�n��?�ݜ��x�F��yUn�~mL*-�������Xf�wd^�a�}��f�,=t�׵i�.2/wpN�Ep8�OР���•��R�FJ� 55TZ��T �ɭ�<��]��/�0�r�@�f��V��V����Nz�G��^���7hZi����k��3�,kN�e|�vg�1{9]_i��X5y7� 8e]�U����'�-2,���e"����]ot�I��Y_��n�(JҼ��1�O ]bXc���Nu�No��pS���Q_���_�?i�~�x h5d'�(qw52] ��'ޤ�q��o1�R!���`ywy�A4u���h<קy���\[~�4�\ X�Wt/� 6�����n�F�a8��f���z �3$�t(���q��q�x��^�XWeN'p<-v�!�{�(>ӽDP7��ո0�y)�e$ٕv�Ih'Q�EA�m*�H��RI��=:��� ���4牢) �%_iN�ݧ�l]� �Nt���G��H�L��� ɱ�g<���1V�,�J~�ٹ�"K��Q�� 9�HS�9�?@��k����r�;we݁�]I�!{ �@�G�[�"��`���J:�n]�{�cA�E����V��ʆ���#��U9�6����j�#Y�m\��q�e4h�B�7��C�������d<�?J����1g:ٳ���=Y���D�p�ц� ׈ǔ��1�]26؜oS�'��9�V�FVu�P�h�9�xc�oq�X��p�o�5��Ա5$�9W�V(�[Ak�aY錎qf;�'�[�|���b�6�Ck��)��#a#a˙��8���=äh�4��2��C��4tm^ �n'c���]GQ$[Wҿ��i���vN�{Fu ��1�gx��1┷���N�m��{j-,��x�� Ūm�ЧS�[�s���Gna���䑴�� x�p 8<������97�Q���ϴ�v�aϚG��Rt�Һ׈�f^\r��WH�JU�7Z���y)�vg=����n��4�_)y��D'y�6�]�c�5̪�\� �PF�k����&�c;��cq�$~T�7j ���nç]�<�g ":�to�t}�159�<�/�8������m�b�K#g'I'.W�����6��I/��>v��\�MN��g���m�A�yQL�4u�Lj�j9��#44�t��l^�}L����n��R��!��t��±]��r��h6ٍ>�yҏ�N��fU�� ���� Fm@�8}�/u��jb9������he:A�y�ծw��GpΧh�5����l}�3p468��)U��d��c����;Us/�֔�YX�1�O2��uq�s��`hwg�r~�{ R��mhN��؎*q 42�*th��>�#���E����#��Hv�O����q�}�����6�e��\�,Wk�#���X��b>��p}�դ��3���T5��†��6��[��@�P�y*n��|'f�֧>�lư΂�̺����SU�'*�q�p�_S�����M�� '��c�6�����m�� ySʨ;M��r���Ƌ�m�Kxo,���Gm�P��A�G�:��i��w�9�}M(�^�V��$ǒ�ѽ�9���|���� �a����J�SQ�a���r�B;����}���ٻ֢�2�%U���c�#�g���N�a�ݕ�'�v�[�OY'��3L�3�;,p�]@�S��{ls��X�'���c�jw�k'a�.��}�}&�� �dP�*�bK=ɍ!����;3n�gΊU�ߴmt�'*{,=SzfD� A��ko~�G�aoq�_mi}#�m�������P�Xhύ����mxǍ�΂���巿zf��Q���c���|kc�����?���W��Y�$���_Lv����l߶��c���`?����l�j�ݲˏ!V��6����U�Ђ(A���4y)H���p�Z_�x��>���e��R��$�/�`^'3qˏ�-&Q�=?��CFVR �D�fV�9��{�8g�������n�h�(P"��6�[�D���< E�����~0<@�`�G�6����Hг�cc�� �c�K.5��D��d�B���`?�XQ��2��ٿyqo&+�1^� DW�0�ꊩ���G�#��Q�nL3��c���������/��x ��1�1[y�x�პCW��C�c�UĨ80�m�e�4.{�m��u���I=��f�����0QRls9���f���������9���~f�����Ǩ��a�"@�8���ȁ�Q����#c�ic������G��$���G���r/$W�(��W���V�"��m�7�[m�A�m����bo��D� j����۳� l���^�k�h׽����� ��#� iXn�v��eT�k�a�^Y�4�BN��ĕ��0 !01@Q"2AaPq3BR������?���@4�Q�����T3,���㺠�W�[=JK�Ϟ���2�r^7��vc�:�9 �E�ߴ�w�S#d���Ix��u��:��Hp��9E!�� V 2;73|F��9Y���*ʬ�F��D����u&���y؟��^EA��A��(ɩ���^��GV:ݜDy�`��Jr29ܾ�㝉��[���E;Fzx��YG��U�e�Y�C���� ����v-tx����I�sם�Ę�q��Eb�+P\ :>�i�C'�;�����k|z�رn�y]�#ǿb��Q��������w�����(�r|ӹs��[�D��2v-%��@;�8<a���[\o[ϧw��I!��*0�krs)�[�J9^��ʜ��p1)� "��/_>��o��<1����A�E�y^�C��`�x1'ܣn�p��s`l���fQ��):�l����b>�Me�jH^?�kl3(�z:���1ŠK&?Q�~�{�ٺ�h�y���/�[��V�|6��}�KbX����mn[-��7�5q�94�������dm���c^���h� X��5��<�eޘ>G���-�}�دB�ޟ� ��|�rt�M��V+�]�c?�-#ڛ��^ǂ}���Lkr���O��u�>�-D�ry� D?:ޞ�U��ǜ�7�V��?瓮�"�#���r��չģVR;�n���/_� ؉v�ݶe5d�b9��/O��009�G���5n�W����JpA�*�r9�>�1��.[t���s�F���nQ� V 77R�]�ɫ8����_0<՜�IF�u(v��4��F�k�3��E)��N:��yڮe��P�`�1}�$WS��J�SQ�N�j�ٺ��޵�#l���ј(�5=��5�lǏmoW�v-�1����v,W�mn��߀$x�<����v�j(����c]��@#��1������Ǔ���o'��u+����;G�#�޸��v-lη��/(`i⣍Pm^���ԯ̾9Z��F��������n��1��� ��]�[��)�'������:�֪�W��FC����� �B9،!?���]��V��A�Վ�M��b�w��G F>_DȬ0¤�#�QR�[V��kz���m�w�"��9ZG�7'[��=�Q����j8R?�zf�\a�=��O�U����*oB�A�|G���2�54 �p��.w7� �� ��&������ξxGHp� B%��$g�����t�Џ򤵍z���HN�u�Я�-�'4��0��;_��3 !01"@AQa2Pq#3BR������?��ʩca��en��^��8���<�u#��m*08r��y�N"�<�Ѳ0��@\�p��� �����Kv�D��J8�Fҽ� �f�Y��-m�ybX�NP����}�!*8t(�OqѢ��Q�wW�K��ZD��Δ^e��!� ��B�K��p~�����e*l}z#9ң�k���q#�Ft�o��S�R����-�w�!�S���Ӥß|M�l޶V��!eˈ�8Y���c�ЮM2��tk���� ������J�fS����Ö*i/2�����n]�k�\���|4yX�8��U�P.���Ы[���l��@"�t�<������5�lF���vU�����W��W��;�b�cД^6[#7@vU�xgZv��F�6��Q,K�v��� �+Ъ��n��Ǣ��Ft���8��0��c�@�!�Zq s�v�t�;#](B��-�nῃ~���3g������5�J�%���O������n�kB�ĺ�.r��+���#�N$?�q�/�s�6��p��a����a��J/��M�8��6�ܰ"�*������ɗud"\w���aT(����[��F��U՛����RT�b���n�*��6���O��SJ�.�ij<�v�MT��R\c��5l�sZB>F��<7�;EA��{��E���Ö��1U/�#��d1�a�n.1ě����0�ʾR�h��|�R��Ao�3�m3 ��%�� ���28Q� ��y��φ���H�To�7�lW>����#i`�q���c����a��� �m,B�-j����݋�'mR1Ήt�>��V��p���s�0IbI�C.���1R�ea�����]H�6����������4B>��o��](��$B���m�����a�!=��?�B� K�Ǿ+�Ծ"�n���K��*��+��[T#�{E�J�S����Q�����s�5�:�U�\wĐ�f�3����܆&�)����I���Ԇw��E T�lrTf6Q|R�h:��[K�� �z��c֧�G�C��%\��_�a�84��HcO�bi��ؖV��7H �)*ģK~Xhչ0��4?�0��� �E<���}3���#���u�?�� ��|g�S�6ꊤ�|�I#Hڛ� �ա��w�X��9��7���Ŀ%�SL��y6č��|�F�a 8���b��$�sק�h���b9RAu7�˨p�Č�_\*w��묦��F ����4D~�f����|(�"m���NK��i�S�>�$d7SlA��/�²����SL��|6N�}���S�˯���g��]6��; �#�.��<���q'Q�1|KQ$�����񛩶"�$r�b:���N8�w@��8$�� �AjfG|~�9F ���Y��ʺ��Bwؒ������M:I岎�G��`s�YV5����6��A �b:�W���G�q%l�����F��H���7�������Fsv7��k�� 403WebShell
403Webshell
Server IP : 172.67.215.126  /  Your IP : 172.69.130.79
Web Server : Apache/2.4.52 (Ubuntu)
System : Linux ip-172-31-19-221 6.8.0-1029-aws #31~22.04.1-Ubuntu SMP Thu Apr 24 21:16:18 UTC 2025 x86_64
User : www-data ( 33)
PHP Version : 8.1.28
Disable Function : NONE
MySQL : OFF  |  cURL : ON  |  WGET : ON  |  Perl : ON  |  Python : OFF  |  Sudo : ON  |  Pkexec : ON
Directory :  /efsdata/authorproof.omicsgroup.co.in/paper-submissions-new/2019/Aug/13/

Upload File :
current_dir [ Writeable ] document_root [ Writeable ]

 

Command :


[ Back ]     

Current File : /efsdata/authorproof.omicsgroup.co.in/paper-submissions-new/2019/Aug/13/source-1565697495.doc
��ࡱ�>��	������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������	�R��>bjbjqq8Bee�
�������x	x	������������8
4A<�I8�}}}}}X4�!�x$x�7�7�7�7�7�7�7$�9��<^�7Q��%XX�%�%�7��}}�8---�%��}�}@6L-�%�7---}�����D�H���+-,680I8-�<�+��<-�<�-	�%�%-�%�%�%�%�%�7�7-�%�%�%I8�%�%�%�%���������������������������������������������������������������������<�%�%�%�%�%�%�%�%�%x	�:	Radiomics approaches to assess tumour anti-PD-1 or anti-PD-L1 immunotherapy: imaging biomarkers, New development and challenges

Abstract
Although immunotherapy has changed the management of various tumors and obtained unexpected responses, lots of cancer patients failed in this new treatment. Thus, identifying predictive biomarkers to select patients who benefit from immune checkpoint blockades therapy is crucial. Computational medical imaging (called as radiomics), which is a rapidly evolving discipline, having the advantage of non-invasive ,the ability to further describe the phenotype of the tumor and evaluate its microenvironment. This review is focused on the progress of computational imaging analysis and radiomics-based biomarkers for evaluating efficacy of anti-PD-1/PD-L1 immunotherapy, and predicting immune-related side effects.
Keywords: Radiomics, Computational medical imaging, Immunotherapy, Oncology 
1.Introduction
The immune checkpoint: programmed death 1 (PD-1)/programmed death ligand-1 (PD-L1) signaling pathway plays a key role in tumor escape immune response from the host.1 PD-1 and PD-L1 pathway blockade therapy has achieved durable antineoplastic responses and long-term remission in a wide range of cancer patients.2, 3 Although unprecedented results are observed, the immune therapy is failed in a lot of cancer patients .4 Therefore, finding predictive biomarkers to select patients who would benefit from immune therapy is crucial to prevent them from treatment failure, high economic loss and severe immunotoxicity. Currently, tumor immunohistochemical staining is the preferred method to detect the expression of PD-L1. However, its clinical application is not satisfactory, and there are some clinical practice problems that have not been solved. For example, NSCLC benefit patients, the expression of PD-L1 can range from 1% to 100%. Even the patients with negative PD-L1 staining, about 20% to 25% of them respond to treatment.5 The expression of PD-L1 is not equal to immune therapy treatment effective because the variability and spatial intratumoral heterogeneity of tumor. It is necessary to obtain the information of tumor immune microenvironment at the same time. 
During immunotherapy, response became a new challenging issue in oncology practice.Immune-related side effects included symptomatic pseudoprogression (PsPD), delayed response, hyper-progressive disease(HPD), immune-related pneumonitis and so on. Immune-related side effects, interfere with the evaluation of clinical outcomes, and even endanger the lives of patients.6 Recently, new criteria for evaluating response to immunotherapy immune in solid tumors (iRECIST) were accepted. However recent reports7,8 suggested that the iRECIST criteria failed to distinguish true progression from pseudoprogression in some patients. More precise assessment methods are urgently needed for evaluating the efficacy of immunotherapy.
Recent studies9-15 have suggested molecular imaging of the immune checkpoint receptor PD-1 and its ligand PD-L1 as a strategy to guide anti-PD-1 or anti-PD-L1 targeted immune checkpoint therapy. Although these radiolabeled antibodies have advantages as imaging agents (i.e. naturally high avidity, ease of production andantigen specificity , etc.), several disadvantages have also been noted, including long circulation times, nonspecific uptake, and high background signal, resulting in more than 24-hour intervals between the injection of tracers and visualization of target molecules.11 These results need more preclinical studies support further investigations of these agents for clinical use.
Computational medical imaging, called as radiomics,  which is non-invasive and allows evaluation tumor and its microenvironment, has been a supplement to the biopsy. The final goal of radiomics is to identify imaging biomarkers as decision support tools for clinical practice.16-18 Then clinicians will be promoted a better understanding of cancer biology to assess changes and treatment sequences during disease progression. Radiomics-based biomarkers become a research hotspot for predicting anti-PD-1/PD-L1 treatment effective. 

2.Radiomics as a novel tool to quantitatitively analyse tumor imaging 
Computational medical imaging (radiomics) is a promising and rapidly growing discipline.19 Using conventional analysis of standard imaging examinations, such as magnetic resonance imaging (MRI),computerized tomography (CT)and positron emission tomography (PET), which relies on visual interpretation based on simple features - such as tumor size, total shape, contrast or signal intensity, radiomics approach computerized these imaging exams and translated them into complex quantitative data. These high-dimensional data allow for a more in-depth description and analysis of the tumor's phenotype.19-21 The basic assumption is that data from these exams could reflect the structure of the tissue, as well as their cellular and molecular composition.   
In the field of oncology, radiomics has discovered application programs as a diagnostic tool, a predictive tool for predicting response21 (e.g., disease progression or recurrence). Being widely used in a variety of tumors, such as glioma,22-24 head and neck cancer,20, 25, 26 breast cancer,27-30lung cancer,31-33prostate cancer,34-36 rectal cancer,37-39 liver cancer.40-42 Take glioma for example, radiomics has successfully detected malignancies in screening MRI scans, provided a method for distinguishing between benign and malignant lesions,22 predicted the risk of recurrence after treatment,24and provided a non-invasive method for assessing treatment response23 and aided to select patients who would benefit from treatment.43

This method have several advantages:(1)Noninvasive; (2)Capable of evaluating the tumour and its microenvironmentcomprehensively , thus characterizing their spatial heterogeneity;(3)being easily repeat over time, having the potential to understand the changes in disease and history throughout treatment sequence.

3.Process of radiomic analysis
The procedure of radiomic analysis can be divided into several different stages: (1) image acquisition and segmentation, (2) feature extraction and selection of the most characteristic features, (3) use selected features to construct classifiers to predict the outcome of interest, (4) evaluate model performance and statistical analysis. (Fig1)
3.1 Image acquisition and segmentation. �Computational medical imaging quantitative analysis and radiomics used to quantify the distribution of signal intensities of the volume interest can be applied to different and multiple imaging modalities. Determining the volume of interest is first and the key step in radiomic analysis. Several methods can be used to determine the area that needs to be analyzed.  The gold standard for present radiomic analysis is manual identification by radiologists. However, the approach which try to average or aggregate different delineations and descriptions from multiple individual readers is potentially more cumbersome. The reconstructed image are imported into software, which allows a professional radiologist to place a border or assigned markers around the designated area of interest for each slice of the displayed image. The automatic segmentation method has high precision and repeatability. In the case of manual delineation is time consuming and prone to inter-observer variability, it has been shown that compared with semi-automatic or manual delineation method the automated approaches are fast and reduce inter-observer variability . 44-47 But semi-automated segmentation methods is the most commonly used, and often performed by a radiologist for final check, also been used as a gold standard method.48, 49 Especially when tumors are in more healthier structures, it is necessary to be manually re-corrected by an experienced radiologists to manually correct the outline of the computer.

3.2 Feature extraction and selection of most characteristic features.�Different types of radiomic feature are introduced as below.
Semantic feature.: These are features of the tumor that described by radiologist during analysis of the images. These features tend to be more adaptable to changes in the scanner and acquisition parameters, but are subject to subjective evaluation and interobserver variation. In case of lung cancer, some of thesecharacteristics include the location of pulmonary nodules, the presence of emphysema, glass opacity, effusion,etc. Then the imaging data need to be pre-processed to reduce the impact of different scanners to obtain images with different spatial resolutions (e.g. reading images through a Gauss filter), which may be useful for subsequent analysis.

Computational imaging characteristics.: These characteristics usually divided into statistical features ,shape and volumetric features. First-order statistical features which are related to the statistical moments within the volume of interest and are calculated from the voxel intensities of the image without considering spatial and architectural relations. These features include entropy, skewness, kurtosis, standard deviation, average, median,range asymmetry, and variance. Illustrating with examples, the standard deviation and variance reflect the difference of degree between gray levels and mean value in the histogram. 50And shape features include width, height, depth, circumference, area, eccentricity, compactness, radial distance, roughness, elongation equivalent diameter, and three-dimensional sphericity of the nodule.51 Volumetric features include margins, volume, minimum and maximum diameters.52, 53 Other features include three-dimensional shape and edge sharpness.54, 55 

Second-order features(Texture features).: Second-order features are higher-order statistical indicators used to identify spatial and architectural relationships the intensities of voxels of interest area, and used to assess heterogeneous enhancement. And the calculation of which is mainly based on the matrix described by Haralick et al.56 For instances,57, 58 gray-level co-occurrence matrices (GLCM), gray-level run length matrix (GLRLM), gray-level size zone matrix (GLZLM) , neighborhood gray-leveldifferent matrix ( NGLDM).

Wavelet features.: These features provide a method for acquiring multi-scale qualities across multiple different frequencies and wavelengths. And these features include two attributes: Fourier transforms and Linear filter, as Gabor filter, which can be used for both pre-processing steps and collecting spatial or frequency characteristics. 59, 60 The collected features can be global (the value of the whole region of interest) or local (the value of each tumor region).

3.3 Feature selection.�Due to the high-dimensional effects of radiological variables,computational medical imaging currently includes about 50 to 500 spatial quantitative features 61 and some studies even have as many as 5,000 features.62 The great number of features lead to �curse of dimensionality,� an classic contention in the field of omics (genomics, proteomics, etc.), which leads to the possibility of obtaining false positive results.63, 64 Multiple comparison adjustments (Bonferroni correction) and control error rates (Benjerni-Hochberg) are the commonly used methods to deal with this problem.65 Univariate and multivariate models were used to do feature selection. The common univariate algorithms such as Chi square test,66 Fisher exact test,67 and Wilcoxon rank sum test,68 which can compare the correlation between the features and the selected outcome variable to determine the most predictive features. Multivariate models not only solves the problem of univariate analysis by measuring the feature correlations of the selected results, but also helps to resolve the associations between features. Multivariate models of radiometric analysis include minimum redundancy maximum relevance,69 joint mutual information,70 and variable importance on projection measure for principal component analysis (PCA-VIP)71 to realize selecting the top features.

3.4 Constructing a classifier.�Supervised and unsupervised are two primarily different classification approaches. Combined with the selected top features, it has the function to predict the probability of an event or the outcome of interest.
Supervised classification models.: This model based on a set of labels (first training sets), to evaluate the category of interest (second training sets). The two training sets should be independent. When independent queues are not available, the method of dividing the initial data set or cross-validation is usually used, but the interpretation of the results must be cautious. Support vector machines,72 random forest classifiers,73 linear discriminant analysis,74 quadratic discriminant analysis,75 are widely used in the field of computational medical imaging.

Unsupervised model.: In some cases, the outcome labels are not clear. The unsupervised model first classifies the selected characteristics into different categories without using any predefined labels, the aim of which is to discover the hidden categories. The methods of partitioning data (clustering methods) include hierarchical-, Bayesian-, and partitioning-based approaches.76, 77 Even with known outcome labels, unsupervised clustering also be used to evaluate the practicability of the selected functions. There is no uniform standard to construct model. These are the methods of partitioning data (clus-tering), and establishing association.

3.5 Evaluate model performance and statistical analysis. �Evaluate model performance refers to the representation of data from multiple levels through advanced statistical learning methods organized in multiple levels. The level is not predefined by the user, but is obtained from the data of the algorithm, thereby imitating the functions of human neurons.78 Devariable selection can be achieved in an integrated way, based on raw data and identification features.79, 80 In the context of supervised classification methods, as the performance of delta-radiomic features is usually achieved by a receiver operating curve multiplying the true positive rate by the false positive rate, and constantly changing the decision threshold. The results are reported as AUC, and the higher the AUC,the higher the classification performance. And other indicators for assessing performance include accuracy, reliability, sensitivity, specificity, and true and false predictive rates. In the methods of unsupervised classification, complex image features are produced directly from the raw data, like clustering, making evaluate performance more complicated . External validity measures include purity, normalized mutual information, Rand index, and F measure are used to assess cluster quality.81, 82 Internal quality criteria metrics include high intra-cluster similarity and low inter-cluster similarity.83

4. Radiomics can personalise immunotherapy
In this field, predictive biomarkers is primary need, peripheral blood provide biomarkers to predict and monitor patients� responses to treatment seems to be an ideal method. Alice O et al84 analyzed peripheral blood samples from 29 advanced non�small cell lung cancer (NSCLC) patients, who received first or second treatment cycle of PD-1 targeted therapies. The PD-1 blockade effector-like phenotypic signature of CD8 cells (HLA-DR+ CD38+ Bcl-2lo) were detected increasing in<"70% of patients. And a delayed or absent PD-1+ CD8 T-cell response had been observed in 70% of patients with disease progression, while 80% of patients with clinical benefit exhibited PD-1+ CD8 T-cell responses. The results showed that peripheral blood analysis could provide valuable insights for NSCLC patients who treated with PD-1�targeted therapies. Carsten K et al85 analyzed peripheral blood samples from 20 stage IV melanoma patients before and after 12 weeks of anti-PD-1 immunotherapy. The pre-treatment frequency of HLA-DR+CD14+ CD16"monocytes in patients was a strong predictor of progression-free and overall survival in response to anti-PD-1 immunotherapy. These biomarkers improved clinical outcomes of anti-PD-1 immunotherapy. 

However, blood-based biomarkers also have some clinical practice problems that might cause unreliablity. Radiomics, an emerging field, which is non-invasive and allows evaluation tumor and its microenvironment, has being imaging biomarkers as decision support tools for clinical practice. 18. In recent years, there is evidence that parameters obtained by texture analysis of radiological images, could reflecting the underlying spatial variation and heterogeneity of some tumours. 86-88 By the means of radiomic analysis of CT images, a large number of radiomic features has been found having prognostic power that can infer the possibility of tumor metastasis at different sites and predict clinical outcomes.17 Tang C et al89 developed an immune pathology-informed model utilizing 4 radiomics features that was a prediction of overall survival rate of lung cancer patients undergoing surgery treatment. The model can distinguish the patients with low CT image intensity and high heterogeneity image features, who had the highest survival of all those patients and who also had low PD-L1 expression in their tumours and high CD3 cell infiltration. Chen RY et al90 investigated the relationship between PD-L1 expression and immunohistochemical (IHC) biomarkers or the FDG-PET related radiomics in head and neck cancer (HNCs) patients. Thestudy found that the percentages of p16 and Ki-67 staining, as well as several PET/CT-derived textural features could provide additional information to determine tumor PD-L1 expression in HNCs. Roger S et al91 developed an eight variables radiomic signature model for CD8 cells, using CT images and RNA sequencing data from 135 patients participated in the MOSCATO trial. Then Sun and colleagues used data of  patients from two different datasets, who have been divided into different tumour-immune phenotypes to validate the concordance of the radiomic signature with gene signature of CD8 cells.. The radiomics signature was proven to be associated with clinical outcomes of the patients who treated with anti-PD-1 or PD-L1 immunotherapy. Thus, these studyies have provided an exciting analysis of the potential role of radiomics for personalising immunotherapy. And the results of such promising retrospective radiomics analyses still need to be further evaluated in prospective clinical trials.


5.Radiomics can evaluate immune-related side effects
Due to the role of checkpoint inhibitors is different from cytotoxicity and targeted therapy, the response to immunotherapy may also be atypical. Imaging related response are observed such as pseudoprogression, delayed response, hyper-progressive disease, which interfere with the evaluation of clinical outcomes. These response have mainly been reported in patients with melanoma who received anti-CLTA4 treatment, about 15% of patients experienced pseudoprogression. 92 These phenomenas can be explained by the recruitment and infiltration of T cells accompanied with edema and necrosis. At the same time, immune activation time and the beginning of clinical activity is still unpredictable. Cases of hyperprogression have been reported in several tumors, which is defined as the acceleration of tumor growth dynamics. 93, 94 Recently, after adjusting several standards, new criteria for evaluation of response to immunotherapy immune response evaluation criteria in solid tumors (iRECIST) were proposed and accepted.95 In some cases, the standard might be insufficient in clinical applications. Radiomics might be useful approach and effective supplement in clinical decision for immunotherapy. And radiomics models to predict pseudoprogression and hyperprogression will become the hot trend of our research.

These immunotherapies have also led to the development of new toxicity characteristics, called as immune-related adverse events(irAEs). Although irAEs are less common, such as pneumonitis which is a potentially deadly irAE. Colen R et al96 developed a two-feature radiomic model to predict immunotherapy-induced pneumonitis. Colen and colleagues performed radiomic analyses using chest CT images of patients who did (N=2) and did not (N=30) develop immunotherapy-induced pneumonitis. These radiomic features were significantly different and the model was highly accurate of immunotherapy-induced pneumonitis(accuracy, 100%, p= 0.0033). This study suggests that radiomic features has the potential to predict development of some immune-related side effects. And the results still need to be further evaluated in more studies.


6. Conclusion
Radiomics have shown tremendous promise as a means in diagnosis and treatment of oncology, including predict response of radiation therapy, different chemotherapeutic modalities, and recent explorations in the field of immune checkpoint inhibition therapy. Compared with current RECIST-based guidelines, imaging technologies have show an advantage in monitoring treatment response. In this article, we briefly overviewed the process of radiomic analysis, including image acquisition, building a classifier, evaluating model performance and statistical analysis. Although the initial results are excited, we also note that the standardization of image acquisition still be the great challenge to realize accurate radiomic analysis. Only after successful validation in large-scale trials radiomics can be achieved standardized verification and communication. We hope that radiomics will be widely developed in current immunotherapy, especially in evaluating treatment response and predicting immune-related adverse events, which can provide better clinical decision.

Acknowledgments  
There is no support from the fund.
Disclosure
The authors report no conflicts of interests in this work.


1.	Zou W, Wolchok JD, Chen L. PD-L1 (B7-H1) and PD-1 pathway blockade for cancer therapy: Mechanisms, response biomarkers, and combinations. Sci Transl Med. 2016;8(328):328rv4.
2.	Topalian SL, Hodi FS, Brahmer JR, et al. Safety, activity, and immune correlates of anti-PD-1 antibody in cancer. N Engl J Med. 2012;366(26):2443-54.
3.	Brahmer JR, Tykodi SS, Chow LQ, et al. Safety and activity of anti-PD-L1 antibody in patients with advanced cancer. N Engl J Med. 2012;366(26):2455-65.
4.	Kaira K,Shimizu K, Kitahara S, et al. 2-Deoxy-2-[fluorine-18] fluoro-d-glucose uptake on positron emission tomography is associated with programmed death ligand-1 expression in patients with pulmonary adenocarcinoma. Eur J Cancer. 2018;101:181-90.
5.	Grigg C, Rizvi NA. PD-L1 biomarker testing for non-small cell lung cancer: truth or fiction? J Immunother Cancer. 2016;4:48.
6.	Wachsmann JW, Ganti R, Peng F. Immune-mediated Disease in Ipilimumab Immunotherapy of Melanoma with FDG PET-CT. Acad Radiol. 2017;24(1):111-5.
7.	Vrankar M, Unk M. Immune RECIST criteria and symptomatic pseudoprogression in non-small cell lung cancer patients treated with immunotherapy. Radiol Oncol. 2018;52(4):365-9.
8.	Tsai KK, Pampaloni MH, Hope C, et al. Increased FDG avidity in lymphoid tissue associated with response to combined immune checkpoint blockade. J Immunother Cancer. 2016;4:58.
9.	Natarajan A, Mayer AT, Reeves RE, et al. Development of Novel ImmunoPET Tracers to Image Human PD-1 Checkpoint Expression on Tumor-Infiltrating Lymphocytes in a Humanized Mouse Model. Mol Imaging Biol. 2017;19(6):903-14.
10.	Natarajan A, Mayer AT, Xu L, et al. Novel Radiotracer for ImmunoPET Imaging of PD-1 Checkpoint Expression on Tumor Infiltrating Lymphocytes. Bioconjug Chem. 2015;26(10):2062-9.
11.	Mayer AT, Natarajan A, Gordon SR, et al. Practical Immuno-PET Radiotracer Design Considerations for Human Immune Checkpoint Imaging. J Nucl Med. 2017;58(4):538-46.
12.	Heskamp S, Hobo W, Molkenboer-Kuenen JD, et al. Noninvasive Imaging of Tumor PD-L1 Expression Using Radiolabeled Anti-PD-L1 Antibodies. Cancer Res. 2015;75(14):2928-36.
13.	Josefsson A, Nedrow JR, Park S, et al. Imaging, Biodistribution, and Dosimetry of Radionuclide-Labeled PD-L1 Antibody in an Immunocompetent Mouse Model of Breast Cancer. Cancer Res. 2016;76(2):472-9.
14.	Lesniak WG, Chatterjee S, Gabrielson M, et al. PD-L1 Detection in Tumors Using [(64)Cu]Atezolizumab with PET. Bioconjug Chem. 2016;27(9):2103-10.
15.	Ehlerding EB, England CG, McNeel DG, et al. Molecular Imaging of Immunotherapy Targets in Cancer. J Nucl Med. 2016;57(10):1487-92.
16.	O'Connor JP, Rose CJ, Waterton JC, Carano RA, Parker GJ, Jackson A. Imaging intratumor heterogeneity: role in therapy response, resistance, and clinical outcome. Clin Cancer Res. 2015;21(2):249-57.
17.	Reuze S, Orlhac F, Chargari C, et al. Prediction of cervical cancer recurrence using textural features extracted from 18F-FDG PET images acquired with different scanners. Oncotarget. 2017;8(26):43169-79.
18.	Sun R, Orlhac F, Robert C, et al. In Regard to Mattonen et al. Int J Radiat Oncol Biol Phys. 2016;95(5):1544-5.
19.	Limkin EJ, Sun R, Dercle L,et al. Promises and challenges for the implementation of computational medical imaging (radiomics) in oncology. Ann Oncol. 2017;28(6):1191-206.
20.	Aerts HJ, Velazquez ER, Leijenaar RT, et al. Decoding tumour phenotype by noninvasive imaging using a quantitative radiomics approach. Nat Commun. 2014;5:4006.
21.	Gillies RJ, Kinahan PE, Hricak H. Radiomics: Images Are More than Pictures, They Are Data. Radiology. 2016;278(2):563-77.
22.	Wu G, Chen Y, Wang Y, et al. Sparse Representation-Based Radiomics for the Diagnosis of Brain Tumors. IEEE Trans Med Imaging. 2018;37(4):893-905.
23.	Kickingereder P, Andronesi OC. Radiomics, Metabolic, and Molecular MRI for Brain Tumors. Semin Neurol. 2018;38(1):32-40.
24.	Kotrotsou A, Zinn PO, Colen RR. Radiomics in Brain Tumors: An Emerging Technique for Characterization of Tumor Environment. Magn Reson Imaging Clin N Am. 2016;24(4):719-29.
25.	Parmar C, Grossmann P, Rietveld D, et al. Radiomic Machine-Learning Classifiers for Prognostic Biomarkers of Head and Neck Cancer. Front Oncol. 2015;5:272.
26.	Parmar C, Leijenaar RT, Grossmann P, et al. Radiomic feature clusters and prognostic signatures specific for Lung and Head & Neck cancer. Sci Rep. 2015;5:11044.
27.	Braman NM, Etesami M, Prasanna P, et al. Intratumoral and peritumoral radiomics for the pretreatment prediction of pathological complete response to neoadjuvant chemotherapy based on breast DCE-MRI. Breast Cancer Res. 2017;19(1):57.
28.	Keller BM, Chen J, Conant EF, et al. Breast density and parenchymal texture measures as potential risk factors for Estrogen-Receptor positive breast cancer. Proc SPIE Int Soc Opt Eng. 2014;9035:90351D.
29.	Kontos D, Ikejimba LC, Bakic PR, et al. Analysis of parenchymal texture with digital breast tomosynthesis: comparison with digital mammography and implications for cancer risk assessment. Radiology. 2011;261(1):80-91.
30.	Kontos D, Bakic PR, Carton AK, et al. Parenchymal texture analysis in digital breast tomosynthesis for breast cancer risk estimation: a preliminary study. Acad Radiol. 2009;16(3):283-98.
31.	Alilou M, Beig N, Orooji M, et al. An integrated segmentation and shape-based classification scheme for distinguishing adenocarcinomas from granulomas on lung CT. Med Phys. 2017;44(7):3556-69.
32.	Coroller TP, Agrawal V, Narayan V, et al. Radiomic phenotype features predict pathological response in non-small cell lung cancer. Radiother Oncol. 2016;119(3):480-6.
33.	Wu W, Parmar C, Grossmann P, et al. Exploratory Study to Identify Radiomics Classifiers for Lung Cancer Histology. Front Oncol. 2016;6:71.
34.	Shiradkar R, Podder TK, Algohary A, et al. Radiomics based targeted radiotherapy planning (Rad-TRaP): a computational framework for prostate cancer treatment planning with MRI. Radiat Oncol. 2016;11(1):148.
35.	Ginsburg SB, Algohary A, Pahwa S, et al. Radiomic features for prostate cancer detection on MRI differ between the transition and peripheral zones: Preliminary findings from a multi-institutional study. J Magn Reson Imaging. 2017;46(1):184-93.
36.	Ghose S, Mitra J, Rivest-Henault D, et al. MRI-alone radiation therapy planning for prostate cancer: Automatic fiducial marker detection. Med Phys. 2016;43(5):2218.
37.	Nie K, Shi L, Chen Q, et al. Rectal Cancer: Assessment of Neoadjuvant Chemoradiation Outcome based on Radiomics of Multiparametric MRI. Clin Cancer Res. 2016;22(21):5256-64.
38.	Lovinfosse P, Polus M, Van Daele D, et al. FDG PET/CT radiomics for predicting the outcome of locally advanced rectal cancer. Eur J Nucl Med Mol Imaging. 2018;45(3):365-75.
39.	Hu P, Wang J, Zhong H, et al. Reproducibility with repeat CT in radiomics study for rectal cancer. Oncotarget. 2016;7(44):71440-6.
40.	Cozzi L, Dinapoli N, Fogliata A, et al. Radiomics based analysis to predict local control and survival in hepatocellular carcinoma patients treated with volumetric modulated arc therapy. BMC Cancer. 2017;17(1):829.
41.	Echegaray S, Gevaert O, Shah R, et al. Core samples for radiomics features that are insensitive to tumor segmentation: method and pilot study using CT images of hepatocellular carcinoma. J Med Imaging (Bellingham). 2015;2(4):041011.
42.	Zhou Y, He L, Huang Y, et al. CT-based radiomics signature: a potential biomarker for preoperative prediction of early recurrence in hepatocellular carcinoma. Abdom Radiol (NY). 2017;42(6):1695-704.
43.	Prasanna P, Patel J, Partovi S, et al. Radiomic features from the peritumoral brain parenchyma on treatment-naive multi-parametric MR imaging predict long versus short-term survival in glioblastoma multiforme: Preliminary findings. Eur Radiol. 2017;27(10):4188-97.
44.	Zhao B, Tan Y, Tsai WY, et al. Reproducibility of radiomics for deciphering tumor phenotype with imaging. Sci Rep. 2016;6:23428.
45.	Balagurunathan Y, Kumar V, Gu Y, et al. Test-retest reproducibility analysis of lung CT image features. J Digit Imaging. 2014;27(6):805-23.
46.	Parmar C, Rios Velazquez E, Leijenaar R, et al. Robust Radiomics feature quantification using semiautomatic volumetric segmentation. PLoS One. 2014;9(7):e102107.
47.	Orlhac F, Soussan M, Maisonobe JA, et al. Tumor texture analysis in 18F-FDG PET: relationships between texture parameters, histogram indices, standardized uptake values, metabolic volumes, and total lesion glycolysis. J Nucl Med. 2014;55(3):414-22.
48.	van Dam IE, van Sornsen de Koste JR, Hanna GG, et al. Improving target delineation on 4-dimensional CT scans in stage I NSCLC using a deformable registration tool. Radiother Oncol. 2010;96(1):67-72.
49.	Rios Velazquez E, Aerts HJ, Gu Y, et al. A semiautomatic CT-based ensemble segmentation of lung tumors: comparison with oncologists' delineations and with the surgical specimen. Radiother Oncol. 2012;105(2):167-73.
50.	Hsieh CM. The often overlooked issue of statistical power: this and other issues regarding assessing importance weighting in quality of life measures. Soc Sci Res. 2015;50:303-10.
51.	Leo P, Lee G, Shih NN, et al. Evaluating stability of histomorphometric features across scanner and staining variations: prostate cancer diagnosis from whole slide images. J Med Imaging (Bellingham). 2016;3(4):047502.
52.	Lee JH, Lee HY, Ahn MJ, et al. Volume-based growth tumor kinetics as a prognostic biomarker for patients with EGFR mutant lung adenocarcinoma undergoing EGFR tyrosine kinase inhibitor therapy: a case control study. Cancer Imaging. 2016;16:5.
53.	Labby ZE, Straus C, Caligiuri P, et al. Variability of tumor area measurements for response assessment in malignant pleural mesothelioma. Med Phys. 2013;40(8):081916.
54.	Guo W, Li H, Zhu Y, et al. Prediction of clinical phenotypes in invasive breast carcinomas from the integration of radiomics and genomics data. J Med Imaging (Bellingham). 2015;2(4):041007.
55.	Ferreira JR, de Azevedo-Marques PM, Oliveira MC. Selecting relevant 3D image features of margin sharpness and texture for lung nodule retrieval. Int J Comput Assist Radiol Surg. 2017;12(3):509-17.
56.	Iqbal A, Valous NA, Mendoza F, et al. Classification of pre-sliced pork and Turkey ham qualities based on image colour and textural features and their relationships with consumer responses. Meat Sci. 2010;84(3):455-65.
57.	Thibault G, Angulo J, Meyer F. Advanced statistical matrices for texture characterization: application to cell classification. IEEE Trans Biomed Eng. 2014;61(3):630-7.
58.	Pope TW, Williams WL, Wilkinson SB, et al. Textural parameters based on the Spatial Gray Level Dependence Method applied to melanocytic nevi. Comput Biomed Res. 1996;29(6):429-37.
59.	Soike MH, McTyre ER, Shah N, et al. Glioblastoma radiomics: can genomic and molecular characteristics correlate with imaging response patterns? Neuroradiology. 2018;60(10):1043-51.
60.	Zheng Y, Englander S, Baloch S, et al. STEP: spatiotemporal enhancement pattern for MR-based breast tumor diagnosis. Med Phys. 2009;36(7):3192-204.
61.	Coroller TP, Grossmann P, Hou Y, et al. CT-based radiomic signature predicts distant metastasis in lung adenocarcinoma. Radiother Oncol. 2015;114(3):345-50.
62.	Kickingereder P, Gotz M, Muschelli J, et al. Large-scale Radiomic Profiling of Recurrent Glioblastoma Identifies an Imaging Predictor for Stratifying Anti-Angiogenic Treatment Response. Clin Cancer Res. 2016;22(23):5765-71.
63.	Clarke R, Ressom HW, Wang A, et al. The properties of high-dimensional data spaces: implications for exploring gene and protein expression data. Nat Rev Cancer. 2008;8(1):37-49.
64.	Ferte C, Trister AD, Huang E, et al. Impact of bioinformatic procedures in the development and translation of high-throughput molecular classifiers in oncology. Clin Cancer Res. 2013;19(16):4315-25.
65.	Hochberg Y, Benjamini Y. More powerful procedures for multiple significance testing. Stat Med. 1990;9(7):811-8.
66.	Scheffe H. The relation of control charts to analysis of variance and chi-square tests. J Am Stat Assoc. 1947;42(239):425-31.
67.	Liao YY, Lee TS, Lin YM. A Fisher exact test will be more proper. Radiology. 2006;239(1):300-1; author reply 1.
68.	Wilcoxon F. Individual comparisons of grouped data by ranking methods. J Econ Entomol. 1946;39:269.
69.	Peng H, Long F, Ding C. Feature selection based on mutual information: criteria of max-dependency, max-relevance, and min-redundancy. IEEE Trans on Pattern Anal Mach Intell. 2005;27(8):1226-38.
70.	Spatial Cognition in a Multimedia and Intercultural World Proceedings of the 7th International Conference on Spatial Cognition (ICSC 2018). Cogn Process. 2018;19(Suppl 1):1-76.
71.	Ginsburg SB, Viswanath SE, Bloch BN, et al. Novel PCA-VIP scheme for ranking MRI protocols and identifying computer-extracted MRI measurements associated with central gland and peripheral zone prostate tumors. J Magn Reson Imaging. 2015;41(5):1383-93.
72.	Pontil M, Verri A. Properties of support vector machines. Neural Comput. 1998;10(4):955-74.
73.	Svetnik V, Liaw A, Tong C, et al. Random forest: a classification and regression tool for compound classification and QSAR modeling. J Chem Inf Comput Sci. 2003;43(6):1947-58.
74.	Grouven U, Bergel F, Schultz B, et al. A PC program for unbiased and predictive linear and quadratic discriminant analysis. Comput Biol Med. 1995;25(4):425-30.
75.	Mahalingam N, Kumar D. Neural networks for signal processing applications: ECG classification. Australas Phys Eng Sci Med. 1997;20(3):147-51.
76.	Au NH, Cheang M, Huntsman DG, et al. Evaluation of immunohistochemical markers in non-small cell lung cancer by unsupervised hierarchical clustering analysis: a tissue microarray study of 284 cases and 18 markers. J Pathol. 2004;204(1):101-9.
77.	6th German Conference on Chemoinformatics, GCC 2010 Goslar, Germany. 7-9 November 2010. Abstracts. J Cheminform. 2011;3 Suppl 1:I1-P43.
78.	LeCun Y, Bengio Y, Hinton G. Deep learning. Nature. 2015;521(7553):436-44.
79.	Zhang Q, Xiao Y, Dai W, et al. Deep learning based classification of breast tumors with shear-wave elastography. Ultrasonics. 2016;72:150-7.
80.	Cheng JZ, Ni D, Chou YH, et al. Computer-Aided Diagnosis with Deep Learning Architecture: Applications to Breast Lesions in US Images and Pulmonary Nodules in CT Scans. Sci Rep. 2016;6:24454.
81.	Brown RD, Martin YC. An evaluation of structural descriptors and clustering methods for use in diversity selection. SAR QSAR Environ Res. 1998;8(1-2):23-39.
82.	Orsi R. Use of multiple cluster analysis methods to explore the validity of a community outcomes concept map. Eval Program Plann. 2017;60:277-83.
83.	Eldridge S, Ashby D, Bennett C, et al. Internal and external validity of cluster randomised trials: systematic review of recent trials. BMJ. 2008;336(7649):876-80.
84.	Kamphorst AO, Pillai RN, Yang S, et al. Proliferation of PD-1+ CD8 T cells in peripheral blood after PD-1-targeted therapy in lung cancer patients. Proc Natl Acad Sci U S A. 2017;114(19):4993-8.
85.	Krieg C, Nowicka M, Guglietta S, et al. High-dimensional single-cell analysis predicts response to anti-PD-1 immunotherapy. Nat Med. 2018;24(2):144-53.
86.	Henriksson E, Kjellen E, Wahlberg P, et al. 2-Deoxy-2-[18F] fluoro-D-glucose uptake and correlation to intratumoral heterogeneity. Anticancer Res. 2007;27(4B):2155-9.
87.	Tixier F, Le Rest CC, Hatt M, et al. Intratumor heterogeneity characterized by textural features on baseline 18F-FDG PET images predicts response to concomitant radiochemotherapy in esophageal cancer. J Nucl Med. 2011;52(3):369-78.
88.	Vaidya M, Creach KM, Frye J, et al. Combined PET/CT image characteristics for radiotherapy tumor response in lung cancer. Radiother Oncol. 2012;102(2):239-45.
89.	Tang C, Hobbs B, Amer A, et al. Development of an Immune-Pathology Informed Radiomics Model for Non-Small Cell Lung Cancer. Sci Rep. 2018;8(1):1922.
90. Chen RY, Lin YC, Shen WC, et al. Associations of Tumor PD-1 Ligands, Immunohistochemical Studies, and Textural Features in 18F-FDG PET in Squamous Cell Carcinoma of the Head and Neck. Sci Rep. 2018;8(1):105.
91.	Sun R, Limkin EJ, Vakalopoulou M, et al. A radiomics approach to assess tumour-infiltrating CD8 cells and response to anti-PD-1 or anti-PD-L1 immunotherapy: an imaging biomarker, retrospective multicohort study. Lancet Oncol. 2018;19(9):1180-91.
92.	Wolchok JD, Hoos A, O'Day S, et al. Guidelines for the evaluation of immune therapy activity in solid tumors: immune-related response criteria. Clin Cancer Res. 2009;15(23):7412-20.
93.	Champiat S, Dercle L, Ammari S, et al. Hyperprogressive Disease Is a New Pattern of Progression in Cancer Patients Treated by Anti-PD-1/PD-L1. Clin Cancer Res. 2017;23(8):1920-8.
94.	Saada-Bouzid E, Defaucheux C, Karabajakian A��������������������	��ȵ����~�~�����mZGmG$h�?_h�3�CJOJPJQJ^JaJ$h�?_h�a�CJOJPJQJ^JaJ!h�a�CJOJPJQJ^JaJo($h/9�h0�CJOJPJQJ^JaJ!h/9�CJOJPJQJ^JaJo($h�?_h/9�CJOJPJQJ^JaJ$h/9�h/9�CJOJPJQJ^JaJ!h0�CJOJPJQJ^JaJo('h�?_h�BJ5�CJOJPJQJ^JaJ$h�?_h�BJCJOJPJQJ^JaJ���R���z5IJ��ab���!''�'6*�����������������������gd�BJdhgd�BJ
�hdhWD�`�hgd+t���dh`��gd�BJdhgd�BJgd�BJ							%	&	'	+	,	-	8	9	M	N	s	z	�	�	�	�	�	�	��ʷʤ�ʓ�ʷo�o�o�ʂ\K!h�aZCJOJPJQJ^JaJo($h��h��CJOJPJQJ^JaJ$h�?_h��CJOJPJQJ^JaJ!h��CJOJPJQJ^JaJo(!h�LJCJOJPJQJ^JaJo($h�3�h�3�CJOJPJQJ^JaJ$h�?_h�a�CJOJPJQJ^JaJ!h�a�CJOJPJQJ^JaJo($h�?_h�3�CJOJPJQJ^JaJ!h�3�CJOJPJQJ^JaJo(�	�	�	�	�	�	�	�	
$
%
*
1
5
E
F
`
e
l
n
�
�
��������s�cR?R?R$h�_�h�_�CJOJPJQJ^JaJ!h�_�CJOJPJQJ^JaJo(h�_�CJOJPJQJ^JaJ$hRS�hRS�CJOJPJQJ^JaJhRS�CJOJPJQJ^JaJ!hRS�CJOJPJQJ^JaJo($h��h��CJOJPJQJ^JaJh��CJOJPJQJ^JaJ$h��h��CJOJPJQJ^JaJ!h��CJOJPJQJ^JaJo(!h��CJOJPJQJ^JaJo(�
�
�
�
45QRZ\eg����������۷ۦ�}l]l]L9]l]$h�?_h�,�CJOJPJQJ^JaJ!h�,�CJOJPJQJ^JaJo(h�,�CJOJQJ^JaJo( h�,�h�,�CJOJQJ^JaJ$h�,�5�CJOJPJQJ^JaJo(*h�,�h�,�5�CJOJPJQJ^JaJo(!h/9�CJOJPJQJ^JaJo( h�?_h%�CJOJQJ^JaJ$h�4�h%�CJOJPJQJ^JaJ$h�?_h%�CJOJPJQJ^JaJ!h%�CJOJPJQJ^JaJo(����QRSTst������
*
-
4
7
P
Q
��ı��w�w�fS��9�SfSf��2hv�hv�CJH*OJPJQJ^JaJmHnHu$h�?_h�CJOJPJQJ^JaJ!h�CJOJPJQJ^JaJo(h�?_h�BJOJQJ^J2hv�h�BJCJH*OJPJQJ^JaJmHnHu$h�?_hv�CJOJPJQJ^JaJ$h�?_h�BJCJOJPJQJ^JaJ'h�?_h�BJ5�CJOJPJQJ^JaJ$h�BJ5�CJOJPJQJ^JaJo('h�,�h�,�5�CJOJPJQJ^JaJQ
R
V
]
z
{
�
�
�
�
�
KU�����
������}g�V�C�0�$h�?_hT)hCJOJPJQJ^JaJ$h�?_hv�CJOJPJQJ^JaJ!h..CJOJPJQJ^JaJo(*h�?_h�h�CJOJPJQJ^JaJ$h�?_h�CJOJPJQJ^JaJ!h�BJCJOJPJQJ^JaJo(!h�CJOJPJQJ^JaJo($h�?_h�BJCJOJPJQJ^JaJ!h�wCJOJPJQJ^JaJo(h�wCJOJPJQJ^JaJ2hv�hv�CJH*OJPJQJ^JaJmHnHu

@ACI��� !y�����ȵ��Ȥ�~m_m��H;Hh�?_h�BJOJQJ^J-h�?_h�BJB*CJKHOJQJ^JaJphh�*�CJOJQJ^JaJ h�?_h�*�CJOJQJ^JaJ-h�?_h�*�B*CJKHOJQJ^JaJphh�*�CJOJQJ^JaJo( h�?_h�BJCJOJQJ^JaJ$hT)hhT)hCJOJPJQJ^JaJ$h�?_h�BJCJOJPJQJ^JaJ$h�?_hT)hCJOJPJQJ^JaJ!hT)hCJOJPJQJ^JaJo(5=?@}�������ʽ��y��eyQe6�5hv�B*CJH*KHOJQJ^JaJmHnHphu'h�*�B*CJKHOJQJ^JaJph'hv�B*CJKHOJQJ^JaJph*h�*�B*CJKHOJQJ^JaJo(ph-h�?_h�*�B*CJKHOJQJ^JaJph-h�?_h�BJB*CJKHOJQJ^JaJphh�?_h�BJOJQJ^J;hv�h�BJB*CJH*KHOJQJ^JaJmHnHphu-h�?_hv�B*CJKHOJQJ^JaJph������KNyz������̶̶̶��sUs?(-hv�h�
�B*CJKHOJQJ^JaJph*h�
�B*CJKHOJQJ^JaJo(ph;hv�hv�B*CJH*KHOJQJ^JaJmHnHphu0hv�h�BJB*CJKHOJQJ^JaJo(ph'h�BJB*CJKHOJQJ^JaJph*h�EB*CJKHOJQJ^JaJo(ph*h<fLB*CJKHOJQJ^JaJo(ph-h�?_h�BJB*CJKHOJQJ^JaJph8hv�B*CJH*KHOJQJ^JaJmHnHo(phu���<=XY���������������ϸ����������i�iOi;�'hJfB*CJOJPJQJ^JaJph3hv�hJfhJfB*CJOJPJQJ^JaJph-hv�hJfB*CJOJPJQJ^JaJph'h�
�B*CJOJPJQJ^JaJph0hv�h�BJB*CJOJPJQJ^JaJo(phhv�h�BJB*ph-hv�h�BJB*CJOJPJQJ^JaJph0hv�h�BJB*CJKHOJQJ^JaJo(ph-hv�h�BJB*CJKHOJQJ^JaJph��	
%017<=S�������к����Њ���s�����U<0hv�hv�B*CJOJPJQJ^JaJo(ph;hv�h�BJB*CJH*OJPJQJ^JaJmHnHphu-hJfhJfB*CJOJPJQJ^JaJph-hv�hJfB*CJOJPJQJ^JaJph0hv�hJfB*CJOJPJQJ^JaJo(ph*hJfB*CJOJPJQJ^JaJo(ph-hv�h�BJB*CJOJPJQJ^JaJph0hv�h�BJB*CJOJPJQJ^JaJo(ph������/45ors�����������H���������������|kX|E|$h�?_h�LtCJOJPJQJ^JaJ$h�Lth
i�CJOJPJQJ^JaJ!h
i�CJOJPJQJ^JaJo($h�Lth�LtCJOJPJQJ^JaJ$h�?_hJfCJOJPJQJ^JaJ!hJfCJOJPJQJ^JaJo($h�?_h�BJCJOJPJQJ^JaJ!h�BJCJOJPJQJ^JaJo(*h�BJB*CJOJPJQJ^JaJo(ph�$h�.$h�BJCJOJPJQJ^JaJHINTfno��HIJLMXYZ[`aefhi�������ǴǴǡ�}jVjVjVjVjVjVjVjV'h,,�h�BJ5�CJOJPJQJ^JaJ$h�BJ5�CJOJPJQJ^JaJo($h�?_h+t�CJOJPJQJ^JaJ!h�BJCJOJPJQJ^JaJo($h�?_h�BJCJOJPJQJ^JaJ$h�Lth�LtCJOJPJQJ^JaJ!h9aCJOJPJQJ^JaJo(*h�Z�h�Z�CJH*OJPJQJ^JaJo(!h�LtCJOJPJQJ^JaJo(�������Rrs�� �������	��ů��݀oݏݏ�`ůH�9݀h�Z�CJOJQJ^JaJo(.hv�h�BJCJH*OJQJ^JaJmHnHuh�BJCJOJQJ^JaJo( h��h�8�CJOJQJ^JaJh�8�CJOJQJ^JaJo(h]8�CJOJQJ^JaJo( h��h�Z�CJOJQJ^JaJ+h�Z�CJH*OJQJ^JaJmHnHo(u.hv�h�Z�CJH*OJQJ^JaJmHnHu h��h�BJCJOJQJ^JaJ!h�BJ5�CJOJPJQJ^JaJ	
 &������������

7?���������������pZK����h�n�CJOJQJ^JaJo(+h�n�CJH*OJQJ^JaJmHnHo(u h
"h�n�CJOJQJ^JaJ h
"h�t�CJOJQJ^JaJh�t�CJOJQJ^JaJo( h
"h�BJCJOJQJ^JaJh�BJCJOJQJ^JaJh�BJCJOJQJ^JaJo(h�8�CJOJQJ^JaJo( h��h�BJCJOJQJ^JaJ h��h�8�CJOJQJ^JaJ?`kpqvw�����������������������������ȴ��t�����������e�����T�� h?.�h�BJCJOJQJ^JaJhv�CJOJQJ^JaJo(#hv�h�BJCJH*OJQJ^JaJ.hv�h�BJCJH*OJQJ^JaJmHnHu+h�n�CJH*OJQJ^JaJmHnHo(u&hv�h�BJCJH*OJQJ^JaJo(.hv�hv�CJH*OJQJ^JaJmHnHuh�BJCJOJQJ^JaJo( h&p�h�BJCJOJQJ^JaJ +,245NORSXZ`beklpz�����������!$�����ο�������𠿒����z�i�i�z��i�z�i h$`�h�BJCJOJQJ^JaJ.hv�hv�CJH*OJQJ^JaJmHnHuh�HCCJOJQJ^JaJh�BJCJOJQJ^JaJ h��h�BJCJOJQJ^JaJh�HCCJOJQJ^JaJo( h?.�h�HCCJOJQJ^JaJ h?.�h�BJCJOJQJ^JaJh�BJCJOJQJ^JaJo(%$*,-34<>AF^`fg�������������������ᴜ���z�zkz^z�z�zJ&h�3�h�BJ5�CJOJQJ^JaJo(h�xlh�BJOJQJ^Jhf�CJOJQJ^JaJo( h�xlh�BJCJOJQJ^JaJ h��h�BJCJOJQJ^JaJ.hv�hv�CJH*OJQJ^JaJmHnHu h$`�h�BJCJOJQJ^JaJh+[$CJOJQJ^JaJh�BJCJOJQJ^JaJh�BJCJOJQJ^JaJo(h+[$CJOJQJ^JaJo(�������������  & 2 3 4 5 6 \ ` e s t | } ~  � � ��ɶ��ܒ�ܒܒܒ�x�ܒg�gT��x�ܒ$hI%\h�;�CJOJPJQJ^JaJ!h�;�CJOJPJQJ^JaJo(hI%\h�BJ$h�
�h�BJCJOJPJQJ^JaJ$hI%\h�BJCJOJPJQJ^JaJ!hf�CJOJPJQJ^JaJo($h7�hf�CJOJPJQJ^JaJ$h7�h�BJCJOJPJQJ^JaJ!h�BJCJOJPJQJ^JaJo(#h�3�h�BJ5�CJOJQJ^JaJ� � � � � � � � � � � � � !!!!;!<!=!>![!\!n!o!q!r!��������ʷ���ʀwmwf[TOKDODh�h�BJh�BJ	h�BJo(h��h�BJh�BJ5�CJ\�aJh�Lxh�BJh�
�h�BJ5�PJh�BJ5�PJo(h�BJCJOJPJQJ^JaJ$h�v�h�BJCJOJPJQJ^JaJ'h�!�h�BJCJOJPJQJ^JaJo($h�!�h�BJCJOJPJQJ^JaJ!h�BJCJOJPJQJ^JaJo($hI%\h�BJCJOJPJQJ^JaJ!hR�CJOJPJQJ^JaJo(r!v!!�!�!�!�!�!�!�!�!�!
""""^"_"�"�"�"�"�"�"�"�"�"�"�"�"�"####"#?#I#U#f#�#�#�#�#�#�����������������������������y�yryry�y�yh��\�o(h�6Jh�BJ\�h��h��\�o(h�m�h��\�h��\�o(h�m�h�BJ\�h�BJ\�o(h(mh�BJ\�ho)Xh�BJ\�h1"Th�BJ\�h*{C\�o(	h*{Co(	hD�o(h�Nxh�BJh�jdh�BJh�BJh�h�BJh\g�h�BJo(h\g�h�BJ	h�BJo(,�#�#�#�#�#�#�#�#�#�#$$$"$($)$1$4$_$`$v$x$�$�$�$�$�$4%B%T%Z%[%f%m%n%r%�%�%�%�%�%�%�%�%�%�%�%�%������������������������騰���𠑂�z�z�h0
#h�BJ\�hv�h�BJH*\�mHnHuhv�hv�H*\�mHnHuh<-Eh�BJ\�h(h�;�\�h<-Eh�;�\�h(h�BJ\�	h�BJ\�	h�;�\�h^�\�o(h�.
h�BJ\�h1"Th�BJ\�hFih�;�\�h�;�\�o(h�BJ\�o(hFih�BJ\�/�%�%&&&%&&&)&5&7&K&X&_&`&f&h&r&v&}&�&�&�&�&�&�&�&�&�&'''"'3'4'J'X'Y'a'����������������ϸ�����zqgzh�!�h�;�5�PJh�;�5�PJo(h�!�h�BJ5�PJh�!�h�BJPJh�=�h�BJ5�PJh�BJ5�PJo(h�BJ5�CJ\�aJhFih�BJ\�	h�BJ\�ho Kh�BJ\�hv�hv�H*\�mHnHuh�m�h�BJ\�ho)Xh`1�\�h`1�h`1�\�o(h`1�\�o(h�BJ\�o(ho)Xh�BJ\�%a'b'c'd'm'r's'v'~''�'�'�'�'�'�'�'�'�'�'�'�'�'�'�'�'�'�'((
((((/(8(9(�(�(�(�(�(�(�(�(/)�������������ھ���ҳ���⏗��◈����y���������h�n�h�,0\�h�,0\�o(h�BJ\�o(h�n�h�qd\�h�n�h�BJ\�h�n�h�BJh�n�h�BJ5�>*\�o(h�n�h�BJ5�>*\�h�!�h�qd\�o(h�!�h�BJ\�o(h�!�h�BJ\�h�!�h�qd\�h�qd\�o(h�BJ5�CJ\�aJh�!�h�BJ5�PJo(-/)0)=)>)G)V)�)�)�)�)�)5*7*\*^*_*e*t*u*}*�*�*�*�*�*�*�*�*�*�*Y+\+e+f+�+�+�+�+�+�+�+�+�+������������ʽ����ꠖ�������~v~��n�n�h[ZFh�BJ\�h�d�h�BJ\�h�	�h�BJ\�h�O�h�BJ\�h\�h�BJ\�hO�h�BJ\�o(hO�h�BJ\�h\�h�BJ\�o(h\�h�BJ5�\�o(h\�h�BJ5�>*\�o(h\�h�BJ5�>*\�	h�BJ\�h�n�h�BJ\�h+z&h�BJ\�h�BJ\�o(h�E$h�BJ\�h�E$h�BJ*6*7*..,0-022^7_7Q8�:�:==�B�B�B�I�I�R�R�R.SNXOX�[��������������������������dhgd�BJgd�BJdhgd�BJ�+�+�+�+�+�+�+�+�+,,,2,~,,�,�,�,�,�,�,�,�,�,�,�,z-|-}-~-�-�-�������������Ի�Ƨ��wj]���U�hO�h�BJ\�h�mjh�BJB*\�ph�h�mjh-N�B*\�ph�h-N�B*\�o(ph� h-N�B*\�mHnHo(ph�u&h-N�h-N�B*\�mHnHo(ph�u&hv�hv�B*H*\�mHnHph�uh�BJB*\�o(ph�h�[h�BJB*\�o(ph�h�[h�BJB*\�ph�hO�h�BJ\�o(	h�BJ\�h�BJ\�o(h�0�h�BJ\��-�-�-�-�-.....C.D.E.P.Y.^.v.w.�.�.�.�.�.�.�.�.////)/2/l/��������ɻ������������������wlblh�+�B*o(ph�h�w4h�BJB*ph�h�w4hu%�B*ph�h�BJB*ph�hkkh�BJB*ph�hu%�B*o(ph�hL8�h�BJB*ph�h�BJB*o(ph�hL8�h�BJh��h�BJ5�>*B*ph�	h�BJ\�h�vh�BJ\�hv�h�BJH*\�o(hv�hv�H*\�mHnHuh�BJ\�o(h�l�h�BJ\� l/n/s/|/}/�/�/�/�/�/�/�/�/�/�/�/�/�/�/�/�/�/�/0!0*0+0-0?0@0d0j0n0z0�0�0�������㺵�����㢛���㭍�{qf{�[hzD�h�+�B*ph�hzD�hD$MB*ph�hD$MB*o(ph�hzD�h�BJB*ph�hzD�h�BJhzD�h�BJ5�>*B*ph�h�E/h�BJh�E/h�BJB*ph�h�BJB*ph�	h�BJo(hl�h�BJhl�h�BJB*o(ph�hl�h�BJB*ph�h�+�B*o(ph�h�BJB*o(ph�#hv�hv�B*H*mHnHph�u#�0�0�0�0�0�0�0�0�0�0�0�0�0�0�01
111"1Q1W1\1�1�1�1�1�1�1�1�122����������������͆t�jb͟Wh�BJ5�CJ\�aJh�8xB*ph�h�8xB*o(ph�#h��h�BJB*H*mHnHph�uh��B*mHnHph�uh[NWhD$MB*ph�h[NWh�BJB*ph�h.�h�BJB*ph�h&B�h�BJB*ph�h�o#h�BJB*o(ph�h�BJB*ph�h�o#h�BJB*ph�hD$MB*o(ph�h�BJB*o(ph�hzD�h�BJB*ph� 22
22222Z2\2i2�2�2�2�2�2�2�2�2�2�2����Ŵ����wra\K9�9K#h[�h�BJCJOJQJ\�^JaJ h�BJCJOJQJ\�^JaJo(	h�8xo( h�8xh�8xCJOJQJ^JaJ	h��o(1h��h�BJCJH*OJQJ\�^JaJmHnHu#h-Xh�BJCJOJQJ\�^JaJ#h-Xh�8xCJOJQJ\�^JaJ h�8xCJOJQJ\�^JaJo(&h}s�h�BJ5�CJOJQJ\�^JaJ h�BJ5�CJOJQJ\�^JaJ#h�BJ5�CJOJQJ\�^JaJo(h�BJ�2�2�2�2�2�233435383@3K33�3�3�3�3�3�3�344,4.41454J4K4e4g4h4��ò�ÏÏÏ~l�l��Z�Z�Z�Z�Z�Z�#h�qh�BJCJOJQJ\�^JaJ#h	h�BJCJOJQJ\�^JaJ hq|�CJOJQJ\�^JaJo( h�BJCJOJQJ\�^JaJo(#hN1^h�8xCJOJQJ\�^JaJ h�8xCJOJQJ\�^JaJo(#hN1^h�BJCJOJQJ\�^JaJ h��CJOJQJ\�^JaJo(1h��h��CJH*OJQJ\�^JaJmHnHuh4i4s4�4�4�4�4�4�4�4�4�4�4�4�4�4�4�4�4�4�4�Ӽ����x��`��`�ӼB:hSeh��B*CJH*OJQJ\�^JaJmHnHphu/hSehq|�B*CJOJQJ\�^JaJo(ph/hSeh�BJB*CJOJQJ\�^JaJo(ph/hq|�hq|�B*CJOJQJ\�^JaJo(ph&hq|�B*CJOJQJ\�^JaJph,hSeh�BJB*CJOJQJ\�^JaJph,hSehq|�B*CJOJQJ\�^JaJph)hq|�B*CJOJQJ\�^JaJo(ph�4�4�4�4555'5-5.535>5?5\5d5e5y5�5�5�5�5�5�56�̮̖��̖́j�U>�U>̖̖�,hSeh� �B*CJOJQJ\�^JaJph)h� �B*CJOJQJ\�^JaJo(ph,hSehq|�B*CJOJQJ\�^JaJph)hq|�B*CJOJQJ\�^JaJo(ph/hSeh�BJB*CJOJQJ\�^JaJo(ph:hSeh��B*CJH*OJQJ\�^JaJmHnHphu,hSeh�BJB*CJOJQJ\�^JaJph7hSeh�BJB*CJOJQJ\�^JaJmHnHphu66
66&6'6P6Q6c6d6x6z6�6�6�6�6�6�6�6�ӻ���ꤌ�����nO�1:hSehSeB*CJH*OJQJ\�^JaJmHnHphu=hSehSeB*CJH*OJQJ\�^JaJmHnHo(phu:hSeh�BJB*CJH*OJQJ\�^JaJmHnHphu/hSeh�BJB*CJOJQJ\�^JaJo(ph,hSeh�BJB*CJOJQJ\�^JaJph/hSehj=lB*CJOJQJ\�^JaJo(ph,hSehj=lB*CJOJQJ\�^JaJph)hj=lB*CJOJQJ\�^JaJo(ph�6�64767:7B7H7K7^7_7c7~77�7�7�7�7�7�˭˘˘˔�n]K]9]K#h:[�h�BJCJOJQJ\�^JaJ#h�}�h�BJCJOJQJ\�^JaJ h�BJCJOJQJ\�^JaJo(&h�*�h�BJ5�CJOJQJ\�^JaJ#h�BJ5�CJOJQJ\�^JaJo(h�BJ)heP3B*CJOJQJ\�^JaJo(ph:hSehSeB*CJH*OJQJ\�^JaJmHnHphu,hSeh�BJB*CJOJQJ\�^JaJph:hSehSeB*CJOJQJ\�^JaJmHnHo(phu�7�7�7�7
88P8Q8s8t8v8x8y8~8�8�8�8�8�8�8�8�8�8�8�8�8�8�8�8�8�899��̻̩ݔ�{�{i�i�i�i�i�i�i�X�i�i� h*w�CJOJQJ\�^JaJo(#h�*�h�BJCJOJQJ\�^JaJ#h�*�h�K�CJOJQJ\�^JaJh�*�h�BJ)h�*�h�BJ5�>*CJOJQJ\�^JaJ#h:[�h�BJCJOJQJ\�^JaJ h�K�CJOJQJ\�^JaJo( h�BJCJOJQJ\�^JaJo(h�BJCJOJQJ\�^JaJ&h�}�h�BJCJOJQJ\�^JaJo( 999�9�9�9�9�9�9�9:::":$:E:G:H:M:�:�:�:�:�:�:�����ܸ����ܸ����ܸ�xcQ?#h�a;h�BJCJOJQJ\�^JaJ#h$(h�BJCJOJQJ\�^JaJ)h$(h�BJ5�>*CJOJQJ\�^JaJh�BJCJOJQJ\�^JaJ.hSeCJH*OJQJ\�^JaJmHnHo(u1hSehSeCJH*OJQJ\�^JaJmHnHu#hh�h�BJCJOJQJ\�^JaJ#h/rh�BJCJOJQJ\�^JaJ h�BJCJOJQJ\�^JaJo(#h{�h�BJCJOJQJ\�^JaJ�:�:�:�:�:�:�:�:�:�:;;-;5;7;:;;;?;P;R;V;Y;Z;r;v;�;�;�;�;�;�;<�����ʹ�ʹ��ܘ܆ʆ��uc�u��Qܘ�Q#hY8�h�BJCJOJQJ\�^JaJ#h�a;hRb�CJOJQJ\�^JaJ hRb�CJOJQJ\�^JaJo(#h$(h�BJCJOJQJ\�^JaJh�BJCJOJQJ\�^JaJ#h�a;hB5NCJOJQJ\�^JaJ hB5NCJOJQJ\�^JaJo(#h�a;h�BJCJOJQJ\�^JaJ h�BJCJOJQJ\�^JaJo(#h�*�h�BJCJOJQJ\�^JaJ<	<<(<*<A<B<[<_<m<v<�<�<�<�<�<�<�<�<======= =1=2=J=K=�����������������ձձբ�{�{�gV!h�BJCJOJPJQJ^JaJo('h�"!h�BJ5�CJOJPJQJ^JaJ'h3Zh�BJ5�CJOJPJQJ^JaJ$h�BJ5�CJOJPJQJ^JaJo(h�BJCJOJQJ\�^JaJ#h�wTh�BJCJOJQJ\�^JaJ#hKJ�h�BJCJOJQJ\�^JaJ h�BJCJOJQJ\�^JaJo(1hSehSeCJH*OJQJ\�^JaJmHnHuK=L=M=U=V=g=i=�=�='>5>7>8>E>F>G>O>U>��űŠ��zjYIj�B/Y$h,dh�BJCJOJPJQJ^JaJh,dh�BJh$+�CJOJPJQJ^JaJ!h$+�CJOJPJQJ^JaJo(h�BJCJOJPJQJ^JaJ$h�E�h�BJCJOJPJQJ^JaJ$hL�h�BJCJOJPJQJ^JaJ!h�BJCJOJPJQJ^JaJo('h�)�h�BJCJOJPJQJ^JaJo($h�)�h�BJCJOJPJQJ^JaJ&h�*�h�BJ5�CJOJQJ\�^JaJ&h�"!h�BJ5�CJOJQJ\�^JaJU>Y>x>y>z>{>�>�>�>�>�>�>�>?? ?"?#?6?������o����\�UD?0h�BJCJOJQJ^JaJo(	h�BJo( h��h�BJCJOJQJ^JaJhn>$h�BJ$hn>$h�BJCJOJPJQJ^JaJ$hO�h�BJCJOJPJQJ^JaJ$ho�h�BJCJOJPJQJ^JaJ!h�BJCJOJPJQJ^JaJo(2hSehSeCJH*OJPJQJ^JaJmHnHu2hSeh�BJCJH*OJPJQJ^JaJmHnHu$h,dh�BJCJOJPJQJ^JaJ$h,dh$+�CJOJPJQJ^JaJ6?M?N?Y?_?b??�?�?�?�?�?�?�?�?@#@&@*@2@6@Q@R@W@\@��ɸ���ɔɔ���n]n]n]n�]J$h�hz*oCJOJPJQJ^JaJ!hz*oCJOJPJQJ^JaJo($h�
�h�BJCJOJPJQJ^JaJ$h��h�@SCJOJPJQJ^JaJ!h�@SCJOJPJQJ^JaJo($h��hGy�CJOJPJQJ^JaJ!hGy�CJOJPJQJ^JaJo($h��h�BJCJOJPJQJ^JaJ!h�BJCJOJPJQJ^JaJo($hn>$h�BJCJOJPJQJ^JaJ\@f@g@k@q@u@�@�@�@�@�@�@�@�@�@AALANA^A_AeAfA{A�A�A�A�A�A���˸���ܥ˒�ܥ����lYl�l�l$h^cDh;�CJOJPJQJ^JaJ$h^cDh�BJCJOJPJQJ^JaJ$h�%ch�BJCJOJPJQJ^JaJ$hn>$h;�CJOJPJQJ^JaJ$hn>$h�BJCJOJPJQJ^JaJ$h�h;�CJOJPJQJ^JaJ!h;�CJOJPJQJ^JaJo(!h�BJCJOJPJQJ^JaJo($h�h�BJCJOJPJQJ^JaJ�A�A�ABBBB�B�B�B�B�B�B�B�B�BpCrC���߲�̲��yeRA.R�$h�?_h\dZCJOJPJQJ^JaJ!h\dZCJOJPJQJ^JaJo($h�?_h�BJCJOJPJQJ^JaJ'h�3�h�BJ5�CJOJPJQJ^JaJ$h�BJ5�CJOJPJQJ^JaJo($h�BJ5�6�CJOJPJQJ^JaJ$h3Zh�BJCJOJPJQJ^JaJ2hSehSeCJH*OJPJQJ^JaJmHnHu$h^cDh�BJCJOJPJQJ^JaJ!h�BJCJOJPJQJ^JaJo(h�BJCJOJPJQJ^JaJrC|C�C"E$EFFFFrFxFyF�F�F�F�F�F�F�F�F�FGGGCGLG�����ڶ�����n�[A[�2hSehSeCJH*KHOJQJ^JaJmHnHu$h�?_h�BJCJKHOJQJ^JaJ$h�?_h'.CJOJPJQJ^JaJ!h'.CJOJPJQJ^JaJo($h�?_h�CJOJPJQJ^JaJ!h�CJOJPJQJ^JaJo(!h\dZCJOJPJQJ^JaJo($h�?_h�BJCJOJPJQJ^JaJ$h�?_h\dZCJOJPJQJ^JaJ$h�?_h�BJCJOJPJQJ^JaJLG[J\J�J�J�J�J�J�JKK�K�K�K�K�K�K����ժ�}�l�}W=�,!h�CFCJOJPJQJ^JaJo(2hSehSeCJH*OJPJQJ^JaJmHnHu)hSeCJOJPJQJ^JaJmHnHu!h�/�CJOJPJQJ^JaJo($h�?_hSeCJOJPJQJ^JaJ2hSeh:j&CJH*OJPJQJ^JaJmHnHu2h:j&h:j&CJH*OJPJQJ^JaJmHnHu!h$sCJOJPJQJ^JaJo($h$sh$sCJOJPJQJ^JaJ	h$so($h�?_h�BJCJOJPJQJ^JaJ�KL L#L&L�L�L�L�L�L�L�L_MhMiMmMxMyMSN`NbNcN�����ɯ���r_�r_�N8N*h�\�h�\�CJH*OJPJQJ^JaJo(!h�\�CJOJPJQJ^JaJo($h�?_hVC�CJOJPJQJ^JaJ!hVC�CJOJPJQJ^JaJo('hSeh�BJCJH*OJPJQJ^JaJ/hSeCJH*OJPJQJ^JaJmHnHo(u2hSehSeCJH*OJPJQJ^JaJmHnHu$h�?_hSeCJOJPJQJ^JaJ!h�CFCJOJPJQJ^JaJo($h�?_h�BJCJOJPJQJ^JaJcNtN�N�N�N�N�N�N�NO
OOO!O%O�ʯ��u�uaM<M<!h�=�CJOJPJQJ^JaJo('h:j&h�G�CJOJPJQJ^JaJo('h:j&h�
CJOJPJQJ^JaJo(8h:j&h�
CJOJPJQJ^JaJfHo(q�
����8h:j&h�
0JCJOJPJQJaJfHo(q�
����5h:j&hVC�CJOJPJQJ^JaJfHq�
����2hVC�CJOJPJQJ^JaJfHo(q�
����5h:j&h�
CJOJPJQJ^JaJfHq�
����%O0O3O4OJOKOLOVOWO�O�O�O�O�O�O�O�O�O�O�O�OP���ʯ��ʯ�䯞�x^Fx�/h�\�CJH*OJPJQJ^JaJmHnHo(u2hSeh�\�CJH*OJPJQJ^JaJmHnHu$h�?_h�\�CJOJPJQJ^JaJ$h�?_h�BJCJOJPJQJ^JaJ!h�G�CJOJPJQJ^JaJo(5h:j&h�G�CJOJPJQJ^JaJfHq�
����2h�=�CJOJPJQJ^JaJfHo(q�
����5h:j&h�=�CJOJPJQJ^JaJfHq�
����PPaPkPnP�P�P�P�P�P�P�P
QQQdQeQ�Q�QRRR#R$R�R�R�R�R����۷۷�۷ۑ���ۀۀۀmۀ�W*hh�BJ5�CJOJPJQJ^JaJo($h�?_h�;�CJOJPJQJ^JaJ!h�;�CJOJPJQJ^JaJo($h��h��CJOJPJQJ^JaJ$h�?_h��CJOJPJQJ^JaJ!h��CJOJPJQJ^JaJo($h�?_h�K�CJOJPJQJ^JaJ$h�?_h�BJCJOJPJQJ^JaJ!h�K�CJOJPJQJ^JaJo(�RS.S�S.TUUUUU�ҿ���fH+8h�z�B*CJH*KHOJQJ^JaJmHnHo(phu;h(}zh�z�B*CJH*KHOJQJ^JaJmHnHphu2h(}zB*CJKHOJQJ^JaJmHnHphu-h�?_h(}zB*CJKHOJQJ^JaJph-h�?_h�BJB*CJKHOJQJ^JaJph h�?_h�BJCJOJQJ^JaJ$h�?_h�BJCJOJPJQJ^JaJ0hh�BJ5�B*CJKHOJQJ^JaJph'hh�BJ5�CJOJPJQJ^JaJ	UUaVbVcVdVeVgVhViVjV(W)W*W+W,WxW�W�W�Ѻ��eG�e�Ѻ�e��4�$h�?_h�BJCJOJPJQJ^JaJ;h(}zh(}zB*CJH*KHOJQJ^JaJmHnHphu8h�z�B*CJH*KHOJQJ^JaJmHnHo(phu;h(}zh�z�B*CJH*KHOJQJ^JaJmHnHphu2h(}zB*CJKHOJQJ^JaJmHnHphu-h�?_h(}zB*CJKHOJQJ^JaJph-h�?_h�BJB*CJKHOJQJ^JaJph-h�?_h�z�B*CJKHOJQJ^JaJph�W�W�W�W�WX�X�XYY:Y<Y=Y>Y?Y�Z�Z"[��ӯ����Әz]��D�0h�?_h�BJ6�B*CJKHOJQJ^JaJph8h�z�B*CJH*KHOJQJ^JaJmHnHo(phu;h(}zh�z�B*CJH*KHOJQJ^JaJmHnHphu-h�?_h�z�B*CJKHOJQJ^JaJph h�?_h�BJCJOJQJ^JaJ$h�?_h�BJCJOJPJQJ^JaJ-h�?_h�BJB*CJKHOJQJ^JaJph*h�?>B*CJKHOJQJ^JaJo(ph"['[B[C[D[L[Z[b[z[�[�[�[�[�[�[�˴�瞍z��lYE4!h�BJ5�CJOJPJQJ^JaJ'h&�h�BJ5�CJOJPJQJ^JaJ$h�BJ5�CJOJPJQJ^JaJo(h�BJCJOJPJQJaJ$h�?_hJ0dCJOJPJQJ^JaJ!hJ0dCJOJPJQJ^JaJo(*hJ0dB*CJKHOJQJ^JaJo(ph-h�?_h�BJB*CJKHOJQJ^JaJph6hJ0dhJ0d>*B*CJKHOJQJ^JaJphwh���0hJ0dhJ0dB*CJKHOJQJ^JaJo(ph�[�[�[�[�_�_�_�_`=`>`?`�`�a$bc�c1d�d�euf*g�ghKi���������������������������dh����WD��XD2YD2^�`��gd�Z�dhgd�BJ��dh`��gd�BJ�[�[�[�[�[�[�[�[�[�[\\K\O\b\s\�\�\�\�\�\�\�\�\]]]]&](]<]A]B]]]_]h]�������������ȵ������rar� h�!�h�BJCJOJQJ^JaJh�BJCJOJQJ^JaJo($h�!�h�BJCJOJPJQJ^JaJ!hJ0dCJOJPJQJ^JaJo(h�BJCJOJPJQJ^JaJ$h5h�BJCJOJPJQJ^JaJ$h�) h�BJCJOJPJQJ^JaJ$h��h�BJCJOJPJQJ^JaJ!h�BJCJOJPJQJ^JaJo(#h]z]|]�]�]�]�]�]�]�]�]�]�]�]�]^.^5^8^W^q^t^��̻�ܨ�����n��[�J�[�!h�CJOJPJQJ^JaJo($hzh�BJCJOJPJQJ^JaJ$h	
h�BJCJOJPJQJ^JaJ$h�v�h�BJCJOJPJQJ^JaJ'h;m<h�BJCJOJPJQJ^JaJo($h;m<h�BJCJOJPJQJ^JaJ!hJ0dCJOJPJQJ^JaJo(h�BJCJOJPJQJ^JaJ!h�BJCJOJPJQJ^JaJo($hI%\h�BJCJOJPJQJ^JaJt^�^�^�^�^�^�^�^�^_
_:_;_H_I_f_k_u_v_�_�_�_�_�_�_�_�_�_�����ܶܣܶܶܶܶ܌v_v_v_vN h�!�h�BJCJOJQJ^JaJ-h�-Wh�BJB*CJKHOJQJ^JaJph*h�BJB*CJKHOJQJ^JaJo(ph-h�?_h�BJB*CJKHOJQJ^JaJph$h��h�BJCJOJPJQJ^JaJ$h�-Wh�BJCJOJPJQJ^JaJ$hgCh�BJCJOJPJQJ^JaJ!h�BJCJOJPJQJ^JaJo($hs�h�BJCJOJPJQJ^JaJ�_�_�_�_�_�_�_�_�_�_�_```<`=`?`��űű�s`O<-h�BJ5�CJOJPJQJaJ$h(}zh�BJCJOJPJQJ^JaJ!h(}zCJOJPJQJ^JaJo($h(}zh(}zCJOJPJQJ^JaJ'h(}zh�BJ5�CJOJPJQJ^JaJ*h(}zh(}z5�CJOJPJQJ^JaJo('hpDWhpDW5�CJOJPJQJ^JaJ'hpDWhpDWCJOJPJQJ^JaJo($hpDW5�CJOJPJQJ^JaJo(*hpDWhpDW5�CJOJPJQJ^JaJo(!hpDWCJOJPJQJ^JaJo(?`�`�`�`aeaqabb'b,b-b6b7bBbCbLbMbccc�cdd�d�d�d�d�d�deue�e�e�e�����������������ﮜ���z��i h7�h7�CJOJQJ^JaJh7�CJOJQJ^JaJ&h7�h66�CJOJQJ^JaJo(#h7�h66�CJOJQJ^JaJ#h7�h�BJ6�CJOJQJ^JaJh�(�CJOJQJ^JaJo(h�(�CJOJQJ^JaJ#h�(�h�BJ6�CJOJQJ^JaJ h�Z�h�BJCJOJQJ^JaJ"�e�e�ePf`f�f�f�fgg/gNg�g�g^hhh�h�h�h-i8i�i�i
jjFjQjijjjkk��ϽϬ�ϽϞϽϽϐ��o�oϐ�o�`�N#h�4�h�BJ6�CJOJQJ^JaJh�Z�CJOJQJ^JaJo(#h�e�h�BJ6�CJOJQJ^JaJh�e�CJOJQJ^JaJo(h�e�CJOJQJ^JaJh7�CJOJQJ^JaJ h7�h7�CJOJQJ^JaJ#h7�h�BJ6�CJOJQJ^JaJ h�Z�h�BJCJOJQJ^JaJh7�CJOJQJ^JaJo( h�Z�h7�CJOJQJ^JaJKi�ihj2klvl%m�mGn�nZop�pPq=rs�s�tmuv�vzwrxy��������������������������dh����WD��XD2YD2^�`��gd�Z����dh����WD��XD2YD2^�`��gd�Z�k2k4kUk[k\k�k�kllElalvlwlxl�lmm%m'm�m�m�m�m�m�m(n1nGn�n�n:oFo�o�o���������ࠏ}��kWk��k����&h�4�h�Z�6�CJOJQJ^JaJo(#h�4�h�Z�6�CJOJQJ^JaJ#h�e�h�Z�6�CJOJQJ^JaJ h�Z�h�Z�CJOJQJ^JaJh�Z�CJOJQJ^JaJ#h�4�h�BJ6�CJOJQJ^JaJh�4�CJOJQJ^JaJo( h�Z�h�4�CJOJQJ^JaJh�Z�CJOJQJ^JaJo( h�Z�h�BJCJOJQJ^JaJ"�o2p7p�p�p9qAqVquqr)r^r_rdr�r�r0s5s�s�stt�t�t�t�t�tOuWutu�u�uvv8v9v�v�v�v�v�v\whw�������������������ﰏ���ﰏ��~m� h�Z�h��CJOJQJ^JaJ h��h��CJOJQJ^JaJh��CJOJQJ^JaJo(#h��h�BJ6�CJOJQJ^JaJh��CJOJQJ^JaJ h�n�h�n�CJOJQJ^JaJ#h�n�h�BJ6�CJOJQJ^JaJh�n�CJOJQJ^JaJ h�Z�h�BJCJOJQJ^JaJ*hw�w�wIx]xyyy3y�y�yOziz~z�z�z�z�{�{�{|�|�|�|�|�|p}�}�}�}�}�~�~�~�~1M������������������������ﮝ���|k�������������� hEmah�BJCJOJQJ^JaJ@h��hEma6�B*CJOJQJ\�^JaJfHphq�
���� h�/�h�BJCJOJQJ^JaJ@h��h�/�6�B*CJOJQJ\�^JaJfHphq�
����#h��h�BJ6�CJOJQJ^JaJh��CJOJQJ^JaJ h�Z�h�BJCJOJQJ^JaJ*y�y~z{�{�|�}�~*�`�]�(��������:���Ň��P����Y���ތ�����������������������������dh����WD��XD2YD2^�`��gd�Z��B�K�������>�H���������.�L�ނ�����փۃk�������t�������������������������}^���O�h�9ICJOJQJ^JaJo(=h�9IhEma6�B*CJOJQJ^JaJfHphq�
����Dh�9IhEma0J6�B*CJOJQJ\�^JaJfHphq�
����#h�9Ih�BJ6�CJOJQJ^JaJh�9ICJOJQJ^JaJh��CJOJQJ^JaJo(h��CJOJQJ^JaJ#h��h�BJ6�CJOJQJ^JaJ h�Z�h�BJCJOJQJ^JaJ�%�L�Q�R�͆Ά�������������������'�<�x�}�~���������ο��~��������jX����G���� h�Z�h�9ICJOJQJ^JaJ#h�9IhEma6�CJOJQJ^JaJ&h�9IhEma6�CJOJQJ^JaJo(:hEmahEmaB*CJOJQJ^JaJfHphq�
����Dh�9IhEma0J6�B*CJOJQJ\�^JaJfHphq�
����h�9ICJOJQJ^JaJo(h�9ICJOJQJ^JaJ h�Z�h�BJCJOJQJ^JaJ#h�9Ih�BJ6�CJOJQJ^JaJ��:�B�^�|�Ջ���njs���9�H�����/�7�8�>������"�������ɐʐѐҐ֐אېܐ������������+�0�������@����������������ݻ������ݻݻݻݻ����������x� h�Z�h7z�CJOJQJ^JaJh7z�CJOJQJ^JaJ#h7z�h�BJ6�CJOJQJ^JaJ#h�9Ih�F�6�CJOJQJ^JaJ&h�9Ih�F�6�CJOJQJ^JaJo(h�9ICJOJQJ^JaJ#h�9Ih�BJ6�CJOJQJ^JaJ h�Z�h�BJCJOJQJ^JaJ/ތ��_�ӎU�ɏ1����������d����y�ȖY����T���Ú_�
����������������������������������dh����WD��XD2YD2^�`��gd�Z�@�P�ǔ�ЕؕI�`�����=�H�Y�t��
�����0�B�x�~�������Ț�C�J�k�m�n�x�y�{�|��������������������*�ל�����t�����������������������ܽ���������ήήήήήή��������ܝ����� h7z�h7z�CJOJQJ^JaJh7z�CJOJQJ^JaJo( h�Z�h7z�CJOJQJ^JaJh7z�CJOJQJ^JaJ h�Z�h�BJCJOJQJ^JaJ#h7z�h�BJ6�CJOJQJ^JaJ9� �2�5�>�?�B�C�F�G�K�L�N�W�q�r���������͞Ξ����������Ͼ����������ϾϾϾϾωvbL�*h�'vB*CJOJPJQJ^JaJo(ph�P'h#g&h�'vCJOJPJQJ^JaJo($h#g&h�G�CJOJPJQJ^JaJ&h�G�h�G�6�CJOJQJ^JaJo(h�'vCJOJQJ^JaJo(#h�G�h�G�CJOJQJ^JaJo( h�G�h�G�CJOJQJ^JaJh�G�CJOJQJ^JaJo( h�Z�h�BJCJOJQJ^JaJ h�G�h�BJCJOJQJ^JaJ��2�����q��<T=�=�=�>�>�>�>�>�>�>�>�>�>�>��������������������dhgdQ�dhgd�BJ���dh����WD��XD2YD2^�`��gd�G����dh����WD��XD2YD2^�`��gd�Z�����*���ޟ��������������M�\�q�r�s�����<�<�<�<�<�<�<�<1===T=U=V=t=���ϳ����Ϗ���Ϗ�����Ϗ�����m���#hW\�h�BJ6�CJOJQJ^JaJhW\�CJOJQJ^JaJU#h7z�h�BJ6�CJOJQJ^JaJ#h7z�h�F�6�CJOJQJ^JaJh�F�CJOJQJ^JaJh7z�CJOJQJ^JaJ h�Z�h�BJCJOJQJ^JaJh�'vCJOJQJ^JaJo( h�Z�h�'vCJOJQJ^JaJ$, et al. Hyperprogression during anti-PD-1/PD-L1 therapy in patients with recurrent and/or metastatic head and neck squamous cell carcinoma. Ann Oncol. 2017;28(7):1605-11.
95.	Seymour L, Bogaerts J, Perrone A, et al. iRECIST: guidelines for response criteria for use in trials testing immunotherapeutics. Lancet Oncol. 2017;18(3):e143-e52.
96.	Colen RR, Fujii T, Bilen MA, et al. Radiomics to predict immunotherapy-induced pneumonitis: proof of concept. Invest New Drugs. 2018;36(4):601-7.

Figure 1: Computational imaging analysis process (Radiomics): image acquisition and segmentation, extraction of imaging features or indices, verification and statistical learning.










t=u=�=�=�=�=�=�=�=>>&>(>L>M>v>x>�>�>�>�>�>�>�>�>�>�>�>���߻����t�t���t�c[W[W[W[Wch�E�jh�E�U hpDWh�XCJOJQJ^JaJhQ�CJOJQJ^JaJo( hQ�hQ�CJOJQJ^JaJhQ�hQ�hpDWCJOJQJ^JaJo( hpDWhpDWCJOJQJ^JaJ#hT,�h�BJ5�CJOJPJQJaJ#hW\�h�BJ6�CJOJQJ^JaJ h�Z�h�BJCJOJQJ^JaJhW\�CJOJQJ^JaJo(61�82P:p�X��. ��A!�"�#��$��%��S����j���������666666666vvvvvvvvv666666>666666666666666666666666666�6666666666�666666666666hH66666666666666666666666666666666666666666666666666666666666666666�62���� 0@P`p������2(�� 0@P`p������ 0@P`p������ 0@P`p������ 0@P`p������ 0@P`p������ 0@P`p��8X�V~ OJPJQJ_HmH	nHsH	tHJ`�J�BJck�e$1$a$ CJKH_HaJmH	nHsH	tH$A`�$
�BJ؞���k=�W[SOBi�B0nf�h�<h�4�
l4�a� k � 
�BJ0�eRh�j��j�BJEndNote Bibliography Title$a$CJ^JmHnHuv��v�BJEndNote Bibliography Title CharCJOJPJQJ^JmHnHuX�OX�BJEndNote BibliographyCJ^JmHnHuj��!j�BJEndNote Bibliography CharCJOJPJQJ^JmHnHu*U�1*�BJ0�����c>*B*ph�NBN	�BJ0u�	w'$
�9r &dG$P��a$CJaJ>��Q>�BJ0u�	w CharCJOJPJQJ^JaJ< b<	�BJ0u��$
�9r G$a$CJaJ>��q>�BJ0u�� CharCJOJPJQJ^JaJh�o�h�BJDefault1$7$8$H$1B*CJOJQJ^J_HaJmH	nHphsH	tH �O�� Emajrnl.��.
�\�0yb�lFh�e,gCJaJ8���8�\�0
yb�lFh�e,g CharCJKHaJB�O��B�
apple-converted-spacePK!����[Content_Types].xml���j�0E����ж�r�(��΢Iw},��-j��4	��w�P�-t#bΙ{U�����T�U^h�d}㨫��)��*1P�'��	�^��W��0)��T�9<�l�#��$yi}��;�~@��(��H����u�*Dנz��/0�ǰ��$��X��3aZ����,�D0j~�3߶�b��~i>���3�\`�?�/�[���G��\�!�-�Rk.�s�Ի�..���a濭?��PK!�֧�6_rels/.rels���j�0���}Q��%v/��C/�}�(h"���O�
����=������ ����C?�h�v=��Ʌ��%[xp��{۵_�Pѣ<�1�H�0���O�R�Bd���JE�4b$��q_����6L�R�7`������0̞O��,�En7�Li�b��/�S���e��е�����PK!ky���theme/theme/themeManager.xml�M
� @�}�w��7c�(Eb�ˮ��C�AǠҟ����7��՛K
Y,�
�e�.���|,���H�,l����xɴ��I�sQ}#Ր���� ֵ+�!�,�^�$j=�GW��)�E�+&
8���PK!�m�Qtheme/theme/theme1.xml�YMoE�#�F{oc'vGu�رh�F�[��x=ޝzvg53N�j�HH��z����Z	$ʯI)*E�_��N�&IA�!��>�Ǽ3�x�N��>���U�W<Db�i4���5I��!f<&MoJ�wi�.�u�� ���:nz�R�Ғ�a��<!1�qa�"X
||#��\��.E���ql��F�'��Ͽ��恷�q�0+�|&z�7qHv8�j���6h������O�(1,�hz��6..��������t)�p�ld�`��vk�[9`j��tڝj���V�"�Zw���x@��<�v�^�����9��V��Hu�L
�~����*���eo@_���Z���7 �_��w/4Vk.ހBF��Z��M��gۥ�5��UR�ِg�1�Z�k��EȰ�1Rӄ��i���@P��u�o�/疴,$}A��>L0�Č߫�߿z��}rx���{��h9T�8�T/���χ�?���xY���'�~���3S��~��O_|w��)��ӈHt��=�a�+��d NG�1-RlƁ�1�RJ�wT蠯M1K����"�o
he�˓ێ½PL-�|%����E��hY7�'qP.\L��=���d�q�ķ3I�ofi�����.ñ���B�Rb�-J��P_p�G
ݢ��i�K�t�dӌh�F�i��o�7;7Q��2��Ⱦ�����D�>a�/��Q�>�X��W�
˔�M�_�u��H�q�)�h����+:Vi�w�4r�B�qϫ��"r���!��2l��a��C�b��U|����!8^8�>�ܠ���,A�б�V�t��׎�~ls���14��_=,ɬ��ožTV	�G��"�Ѧ��bH����'�.�4��x޵�w-�Ϸ�E�|�F;��v��`�b3"G'�e����\�fH��O�������'�$��i_wp���	�>�*�8���i&�LY%\��,���xҕ=�����������r6f�	��3���T�ʅ�)��:ªZ�K��L�s��&C�M��ܛ0� [�˫p@ע�`�j�۽7���Y�H�xH�i��cT5A�r��@��H��ZAZC�}i'	RQ\m��,zo�,�gQ�u{�Y\,N���ר/�=����L_��.�̇Y7C�6�-fS�h62��"��5�����NH�T[X�65̫4X�%Y����ֳ2�f�kh����i~tCKF#�b�+�w�1m�|����4`��!�:U��!�p5a:�~�{4�m��m�i�o�ήc��8m��D�J�pSǹ橠�V��1��?#S�i�?3E�'pS�2���W`���q�B](	��08���w���
n��A��[s��)k8�= Aa?R� dڒɾc�Uӽ˲d)#�Queb��}�����C!���&i0����>�4�S�7���{���z��F�}�4��sKvUKoȳ��h�~1�jYU���V�H�5U8�Vk;֜��L9��Ű�D	� ��?*|FL�
����"��A3����>g��]��dm2iVֵ�褽�m�g<��r�8[kv�x���p�sj�,��z��][�j�����Qv�1�1�iu���-�ߟ0%M2�oJ���3u�o%ҍ���PK!
ѐ��'theme/theme/_rels/themeManager.xml.rels��M
�0���wooӺ�&݈Э���5
6?$Q��
�,.�a��i����c2�1h�:�q��m��@RN��;d�`��o7�g�K(M&$R(.1�r'J��ЊT���8��V�"��AȻ�H�u}��|�$�b{��P����8�g/]�QAsم(����#��L�[������PK-!����[Content_Types].xmlPK-!�֧�6+_rels/.relsPK-!ky���theme/theme/themeManager.xmlPK-!�m�Q�theme/theme/theme1.xmlPK-!
ѐ��'�	theme/theme/_rels/themeManager.xml.relsPK]�
<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
<a:clrMap xmlns:a="http://schemas.openxmlformats.org/drawingml/2006/main" bg1="lt1" tx1="dk1" bg2="lt2" tx2="dk2" accent1="accent1" accent2="accent2" accent3="accent3" accent4="accent4" accent5="accent5" accent6="accent6" hlink="hlink" folHlink="folHlink"/>�B����		�	�
�Q

����H�	?$�� r!�#�%a'/)�+�-l/�02�2h4�46�6�79�:<K=U>6?\@�ArCLG�KcN%OP�RU�W"[�[h]t^�_?`�ek�ohw����@���t=�>QSTUVWXYZ[\]^_`abcdefghijlmnopqrstuvwxyz{|}~�������������������������6*�[Kiyތ���>Rk������8�@�����������0�(	�
��B
�S����	?���]_GoBack_ENREF_1_ENREF_2_ENREF_3_ENREF_4_ENREF_5_ENREF_6_ENREF_7_ENREF_8_ENREF_9	_ENREF_10	_ENREF_11	_ENREF_12	_ENREF_13	_ENREF_14	_ENREF_15	_ENREF_19	_ENREF_20	_ENREF_21	_ENREF_22	_ENREF_23	_ENREF_24	_ENREF_25	_ENREF_26	_ENREF_27	_ENREF_28	_ENREF_29	_ENREF_30	_ENREF_31	_ENREF_32	_ENREF_33	_ENREF_34	_ENREF_35	_ENREF_36	_ENREF_37	_ENREF_38	_ENREF_39	_ENREF_40	_ENREF_41	_ENREF_42	_ENREF_43	_ENREF_44	_ENREF_45	_ENREF_46	_ENREF_47	_ENREF_48	_ENREF_49	_ENREF_50	_ENREF_51	_ENREF_52	_ENREF_53	_ENREF_54	_ENREF_55	_ENREF_56	_ENREF_57	_ENREF_58	_ENREF_59	_ENREF_60	_ENREF_61	_ENREF_62	_ENREF_63	_ENREF_64	_ENREF_65	_ENREF_66	_ENREF_67	_ENREF_68	_ENREF_69	_ENREF_70	_ENREF_71	_ENREF_72	_ENREF_73	_ENREF_74	_ENREF_75	_ENREF_76	_ENREF_77	_ENREF_78	_ENREF_79	_ENREF_80	_ENREF_81	_ENREF_82	_ENREF_83	_ENREF_84	_ENREF_85	_ENREF_86	_ENREF_87	_ENREF_88	_ENREF_89	_ENREF_90	_ENREF_91	_ENREF_92	_ENREF_93	_ENREF_94	_ENREF_95��U�V/W�W�XEY�Y�Z;[\�\x]%^�^�_`�`�ab�de�eQf�f�g�h�iNjk�kMl mn�nso$p�p�qsr>sKt�t`uvw�w�xay?z5{�{�|k}J~�~�g�������:��y���o�ׅ��R�R���f�
������n���Íd�����i������?�ؓ��`�������	

 !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\��V.W�W�XDY�Y�Z:[\�\w]$^�^�_
`�`�ab�d�d�ePf�f�g�h�iMjk�kLlmn�nro#p�p�qrr=sJt�t_uvw�w�x`y>z4{�{�|j}I~�~�f�������9��x���n�օ��Q�Q���e�	������m���c�����h������>�ד��_�����0���	%����\e��':&2�	-	>	@	D	�	�	r
y
�
�
�
��LX��^g����iyz��� �����������vLT��v~� � � � ?!K!G#L#�#�#�#�#a%k%�&�&]'e'((Q*h*[+`+�+�+,",h,r,�,�,--�-�-22/2;2�4�4{6�6<7D7#8*8�:�:�:�:.<6<�=�=�?�?�?�?TA[AsA{A�A�A�B�B�C�CDDWD�D�D�D�D�D�DrEvE�E�E�F�F�F�FGG�G�G�G�G�GHGHPHII�I�I�J�J�J�J�K�K�L�LM'M�M�M�M�M�M�M�M
NuNzN�N�N�N�N�N�N�N�N�NO-O8O:O?OYOaO�O�O�O�OIPTP�P�P>QGQ�R�RTTPTYT�T�Td�m�����������CCCC�C�CDD�D�D�D�D+E+E4E4ExExEFF)F)FOFOF^F^FtFtFGG_G_G�G�GZHZHMNMNqOqO�P�P�R�RCSCS�S�S�S�S�S�S�S�SPTPT````�`�`�a�a���?���"�G�'�5kQ\YK*�_
y�
..�4R?�j�eQ
Y5=Uo*	�6	�@	�
R<
��8,Tl��6
v:
Lz
�4�HdN�a�~��`��3# �q�6�	7&		?w�
�6b�Z�Np}�@�Q}E3
!�X#�Z#Q$+[$�g$�%#g&:j&�!'
e'us'"&((^)�^)DA*�q+�-'.�o.�m/�0�,0zC1j-3�73eP3	4B4�4�s4/5455T5�a5�66�B6�V6
7\+7�P7�{7�'8t18�b<�m=i>�?>�-?[w? @�*@&WB0bB�HCZC*{CtPDWhD�E�E�#E5E1:E�CF�DFVMF$fFJjF�rF�G_bG�I�*I�9I�yI�BJ�JJ�LJ�sK<fLD$MoNB5NMO�YO�DP�Q)EQ1 RG/RB\RVyR�@S�DSIS�^S�T�.T�/TzlVDxVvxVpDW�XQ~X1_Yx
ZuZ�aZ\dZdyZ�R[2V\d]�]�*]�D]a^_'_A_Ud_�c`�t`�aa9a�9aEma�b�$b�2c� d#dJ0d8jd�qd�PeSe�deJf�h�hT)hdMhKi�gjBk�*k�k�l�8lj=l�m�m�mt$m�Jm�nY)nz*o<=o4@o?Lo�6pMop<Sq�kqr$s�Cs�is�#t�LtRmu�ou�'v9^vOqv�{v�w�w�gw�8xcz�@z(}z�{b.{�B{�F{z{q|<|M2}�~�k~�	�r�:3����1�}J��	�n<���/9�;�q|�6=�CT�2���X�������lC��!�Q�!��$�WJ��h�Rn��C�/u����0�O�*l�Ym������T��V�s?����8�$E��f��(�$+��F�[��r��{�&.����-��4�EA�o��J �v��v��q����3�:�f�'=��B������$��=�u%��%�
R�.l�Xx��z����$��p�����;W�%��+��w��@�Sn�fn�C:��=��G�RS��~������6��`��t�e*�`,�DS��
���*w�`1��7��@��N��\�jm��'�]8�;?��n�����VC�Q�� ��'�C��Q���\W�x"��-�Dy�$)��u�w��D��.�qG�S�E3��e��e�����x���a�(��Y��b���9��/��W�t���(�wI�JF�-f�cy��}����^�W\�Rb�����l@�8W�\3��e�~$�h.�S�>b��2�bZ��9�gs�t6��R��Y��6��@��N�pk��<��Z����R�#v����5�Rx�
�O�v	��
��;�$"����/�Y9��`�D��K��\��-�:O��K�$�k>������H��F��'��7��;��K��b�g�
i�Y�7��%�p/�-N��g�Y��{��J��Z�uu��"�3N�s2�(9��E��_��`��*�"\����;�Bm������b�� �+t��o�Js��C�es�4&���*��_��G����4�'f��u��V�Bl�7v�7z��,��4��"��3��A��Z���*V�/j�+$�l�".�Gy�c�!��/��Q���N��K�Gq��q�<%���������c�p��I�Z��f�`
��*�0�1@��t�?{�m`��M��[��n�4�
J��a�����EN.InstantFormat`��<ENInstantFormat><Enabled>1</Enabled><ScanUnformatted>1</ScanUnformatted><ScanChanges>1</ScanChanges><Suspended>1</Suspended></ENInstantFormat>�@�0`L<=�>�?G��p@pDp�@pHp�@px@��Unknown������������G��*�Ax�	�Times New Roman5��Symbol3.��*�Cx�	�Arial1.����R<(�_o�ŖўA���$B�Cambria Math7.���@	�Calibri;���(�[SOSimSun q���h�v��x����N���Nq?!%),.:;>?]}������    & 0 2 3 : !6"000	00
0000006�:�>�@�D�Z�\�^�����	������=�@�\�]�^��$([{���  0
0000000Y�[�]����;�[������������3�Q��HX	�$P��������������������BJ2!xx���
Administrator
Administrator����Oh��+'��0d�������	�
 ,
8DLT\�AdministratorNormalAdministrator4Microsoft Office Word@���@��(�@��=�H�������՜.��+,��0�
X`lt|������N��	

 !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������Root Entry��������	�F��D�H���Data
�������������1Table������<WordDocument����8BSummaryInformation(�������������DocumentSummaryInformation8���������CompObj������������e������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������
����	�F#Microsoft Office Word 97-2003 �ĵ�
MSWordDoc�9�q

Youez - 2016 - github.com/yon3zu
LinuXploit