bli 9qt gx s 6iit i rmrk sr mspsk ct xi
play

BLI9QTGXS6IITIRMRK SR@MSPSK )(-CTXI - PowerPoint PPT Presentation

Sep 02, 2023 •154 likes •585 views

BLI9QTGXS6IITIRMRK SR@MSPSK )(-CTXI @SRP%AQQI%6%L%6% AIRMS9RIXMKXS Imaging'Biomarkers'and'CAD'Laboratory

  1. BLI�9QT�GX�S��6IIT��I��RMRK SR�@��MSPSK��� )�(-�CT��XI @SR�P���%�A�QQI�����%6%���L%6% AIRMS��9R�I�XMK�XS� Imaging'Biomarkers'and'CAD'Laboratory Radiology'and'Imaging'Sciences NIH'Clinical'Center Bethesda,'MD www.cc.nih.gov/drd/summers.html

  2. Disclosure • Patent'royalties'from'iCAD • Research'support'from'Ping'An'&'NVidia • Software'licenses'to'Imbio &'Zebra'Med. Disclaimer • Opinions'discussed'are'mine'alone'and'do'not' necessarily'represent'those'of'NIH'or'DHHS.

  4. Overview • Background • Radiology'imaging'applications • Data'mining'radiology'reports'and'images

  5. 6IIT��I��RMRK � 5SR�SP�XMSR�P�RI���P�RIX�S�����5SR�=IX� S��5==�� � 3R�MQT�S�IQIRX�XS�RI���P�RIX�S��� � �S�I�P��I���TI�QMX�LMKLI��PI�IP��S���F�X��GXMSR � AMQMP��MXMI��XS�PS��PI�IP��M�MSR�T�SGI��MRK�MR��RMQ�P� � ����I��MQT�S�IQIRX��MR��SP�MRK�L����T�SFPIQ��PM�I� SFNIGX��IGSKRMXMSR�MR�TMGX��I�

  6. PubMed1Articles 700 600 500 400 300 200 100 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Deep1Learning Deep1Learning1Radiology

  7. Deep'Learning'Improves'CAD Summers'et'al.'Gastroenterology'2005P'Roth'et'al.'IEEE'TMI'2015'

  8. Deep'Learning'Improves'CAD Hua,'Liu,'Summers et'al.'ARRS'2012P'Roth'et'al.'IEEE'TMI'2015'

  9. • 90'CTs'with' 388'mediastinal' LNs • 86'CTs'with' 595'abdominal' LNs • Sensitivities' 70%/83%'at'3' FP/vol.'and' 84%/90%'at'6' FP/vol.,' respectively H'Roth'et'al.,'MICCAI'2014'

  10. • Deeper'CNN'model'performed'best • GoogLeNet'for'mediastinal'LNs • Sensitivity'85%'at'3'FP/vol.' HC'Shin'et'al.,'IEEE'TMI'2016

  11. Lymph'Node'Segmentation I'Nogues et'al.'RSNA'2016'

  12. Lymph'Node'CT'Dataset � �SM%S�K&(�%-/�-&:/&B593%)�(�% 3�9965=� � B593�5B���QTL =S�I � (-,��G�R����.�8� � 3RRSX�XMSR���G�R�M��XI���Q����

  13. ��RG�I���536���MRK�5== H'Roth'et'al.,'MICCAI'2016

  14. Pancreas'CT'Dataset � �SM%S�K&(�%-/�-&:/&B593%) �(,%X=�(�U�C � B593�5B���RG�I�� � .)��G�R���(��8�

  15. 5SPMXM��536 Wei'et'al.'SPIE,'ISBI'2013

  16. Colitis'CAD J'Liu'et'al.'SPIE'Med'Imaging'2016

  17. Colitis'CAD ),�5B��G�R��S��T�XMIRX���MXL�GSPMXM� • ),��MQ�KI� • .����IR�MXM�MX���X�(�7�&MQ�KI • J'Liu'et'al.'SPIE'Med'Imaging'and'ISBI'2016

  18. Colitis'CAD 80 patients 80 controls 93.7% Sensitivity 95.0% Specificity 0.986 AUC J'Liu'et'al.'Medical'Physics'2017

  19. Prostate T2WI ADC B2000 CAD DL Kwak et.1al.1 T2WI ADC B2000 Tsehay'et'al.'SPIE'MI'2017

  20. Prostate Cheng'et'al.'SPIE'MI'2017

  21. Prostate Cheng'et'al.'JMI'2017

  22. ATMRI��IX��X��M��536 J'Burns,'J'Yao'et'al.'RSNA'2011P'Radiology'2013

  23. Deep'Learning'Improves'CAD Roth'et'al.'IEEE'TMI'2015'

  24. AIKQIRX�XMSR���FIP���ST�K�XMSR Gao'et'al.'IEEE'ISBI'2016

  25. AIKQIRX�XMSR���FIP���ST�K�XMSR Gao'et'al.'IEEE'ISBI'2016

  26. A'Harrison'et'al.'MICCAI'2017

  27. J'Yao'et'al.'MICCAI'2017

  28. J'Yao'et'al.'MICCAI'2017

  29. L'Zhang'et'al.'MICCAI'2017

  30. 6�X���MRMRK�@ITS�X�� �9Q�KI� HC'Shin'et'al.'CVPR'2015

  31. 6�X���MRMRK�@ITS�X�� �9Q�KI� � B��MRI��SR�)(,������I��MQ�KI���5B���@���� � (,/�����5B�MQ�KI� � ,������T�XMIRX��G�R� � @IG�PP��X�:��:1(��@2(��GS�I��������%�,� HC'Shin'et'al.'CVPR'2015'&'JMLR'2016

  32. 6�X���MRMRK�@ITS�X�� �9Q�KI� HC'Shin'et'al.'CVPR'2015'&'JMLR'2016

  33. BSTMG0��IX��X��I� HC'Shin'et'al.'CVPR'2015'&'JMLR'2016

  34. 6�X���MRMRK�@ITS�X�� �9Q�KI� HC'Shin'et'al.'CVPR'2015'&'JMLR'2016

  35. 6�X���MRMRK�@ITS�X�� �9Q�KI� HC'Shin'et'al.'CVPR'2016

  36. 6�X���MRMRK�@ITS�X�� �9Q�KI� HC'Shin'et'al.'CVPR'2016

  37. 6�X���MRMRK�@ITS�X�� �9Q�KI� X'Wang'et'al.'WACV'2017

  38. 6�X���MRMRK�@ITS�X�� �9Q�KI� X'Wang'et'al.'CVPR'2017

  39. 5LI�XD����. X'Wang'et'al.'CVPR'2017

  40. ChestX`ray8'Dataset • https://nihcc.app.box.com/v/ ChestXray`NIHCC' • “ChestX`ray8'Dataset” • 112,120'frontal`view chest radiographs,'30,805'unique'patients • 42'GB • Metadata for'all'images • Bounding boxes'for'1000'images

  42. Acknowledgments • CC • Le'Lu • Nicholas'Petrick • NIH'Fellowship' • Berkman'Sahiner • Jack'Yao Programs: • NCI • Joseph'Burns • Jiamin Liu Fogarty • NHLBI • • Perry'Pickhardt • Nathan'Lay ISTP • Mingchen Gao • • NIDDK • Hadi Bagheri • Daniel'Mollura IRTA • • Holger'Roth • Baris Turkbey • NIAID BESIP • • Peter'Choyke • Hoo`Chang'Shin • FDA CRTP • • Matthew'Greer • Xiaosong Wang • Brad'Wood • Mayo'Clinic • Adam'Harrison • Jin Tae'Kwak • DOD • Ke Yan • Ruida Cheng • Nvidia for'GPU' • Ling'Zhang • U.'Wisconsin card'donations • Isabella'Nogues

  43. To'Learn'More'… www.cc.nih.gov/drd/summers.html X'Wang'et'al.'RSNA'2016

Recommend

More recommend