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

▶
bli 9qt gx s 6iit i rmrk sr mspsk ct xi
SMART_READER_LITE
LIVE PREVIEW

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

Sep 02, 2023 154 likes •587 views

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

slide-1
SLIDE 1

BLI9QTGXS6IITIRMRK SR@MSPSK )(-CTXI

@SRP%AQQI%6%L%6% AIRMS9RIXMKXS

Imaging'Biomarkers'and'CAD'Laboratory Radiology'and'Imaging'Sciences NIH'Clinical'Center Bethesda,'MD www.cc.nih.gov/drd/summers.html

slide-2
SLIDE 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.

slide-3
SLIDE 3
slide-4
SLIDE 4

Overview

  • Background
  • Radiology'imaging'applications
  • Data'mining'radiology'reports'and'images
slide-5
SLIDE 5

6IITIRMRK

5SRSPXMSRPRIPRIXS5SR=IX S5== 3RMQTSIQIRXXSRIPRIXS SIPITIQMXLMKLIPIIPSFXGXMSR AMQMPMXMIXSPSPIIPMMSRTSGIMRKMRRMQP IMQTSIQIRXMRSPMRKLTSFPIQPMI SFNIGXIGSKRMXMSRMRTMGXI

slide-6
SLIDE 6

100 200 300 400 500 600 700

2017 2016 2015 2014 2013 2012 2011 2010 2009 2008 2007 2006 2005 2004 2003 2002 2001 2000

PubMed1Articles

Deep1Learning Deep1Learning1Radiology

slide-7
SLIDE 7

Deep'Learning'Improves'CAD

Summers'et'al.'Gastroenterology'2005P'Roth'et'al.'IEEE'TMI'2015'

slide-8
SLIDE 8

Deep'Learning'Improves'CAD

Hua,'Liu,'Summers et'al.'ARRS'2012P'Roth'et'al.'IEEE'TMI'2015'

slide-9
SLIDE 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'

slide-10
SLIDE 10
  • Deeper'CNN'model'performed'best
  • GoogLeNet'for'mediastinal'LNs
  • Sensitivity'85%'at'3'FP/vol.'

HC'Shin'et'al.,'IEEE'TMI'2016

slide-11
SLIDE 11

Lymph'Node'Segmentation

I'Nogues et'al.'RSNA'2016'

slide-12
SLIDE 12

Lymph'Node'CT'Dataset

SM%SK&(%-/-&:/&B593%)(% 39965= B5935BQTL =SI (-,GR.8 3RRSXXMSRGRMXIQ

slide-13
SLIDE 13

RGI536MRK5==

H'Roth'et'al.,'MICCAI'2016

slide-14
SLIDE 14

Pancreas'CT'Dataset

SM%SK&(%-/-&:/&B593%) (,%X=(UC B5935BRGI .)GR(8

slide-15
SLIDE 15

5SPMXM536

Wei'et'al.'SPIE,'ISBI'2013

slide-16
SLIDE 16

Colitis'CAD

J'Liu'et'al.'SPIE'Med'Imaging'2016

slide-17
SLIDE 17

Colitis'CAD

J'Liu'et'al.'SPIE'Med'Imaging'and'ISBI'2016

  • ),5BGRSTXMIRXMXLGSPMXM
  • ),MQKI
  • .IRMXMMXX(7&MQKI
slide-18
SLIDE 18

Colitis'CAD

J'Liu'et'al.'Medical'Physics'2017

80 patients 80 controls 93.7% Sensitivity 95.0% Specificity 0.986 AUC

slide-19
SLIDE 19

Prostate

Tsehay'et'al.'SPIE'MI'2017

T2WI T2WI ADC B2000 ADC B2000 Kwak et.1al.1 CADDL

slide-20
SLIDE 20

Prostate

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

slide-21
SLIDE 21

Prostate

Cheng'et'al.'JMI'2017

slide-22
SLIDE 22

ATMRIIXXM536

J'Burns,'J'Yao'et'al.'RSNA'2011P'Radiology'2013

slide-23
SLIDE 23

Deep'Learning'Improves'CAD

Roth'et'al.'IEEE'TMI'2015'

slide-24
SLIDE 24

Gao'et'al.'IEEE'ISBI'2016

AIKQIRXXMSRFIPSTKXMSR

slide-25
SLIDE 25

Gao'et'al.'IEEE'ISBI'2016

AIKQIRXXMSRFIPSTKXMSR

slide-26
SLIDE 26

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

slide-27
SLIDE 27

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

slide-28
SLIDE 28

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

slide-29
SLIDE 29

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

slide-30
SLIDE 30

HC'Shin'et'al.'CVPR'2015

6XMRMRK@ITSX 9QKI

slide-31
SLIDE 31

6XMRMRK@ITSX 9QKI

BMRISR)(,IMQKI5B@ (,/5BMQKI ,TXMIRXGR @IGPPX::1(@2(GSI%,

HC'Shin'et'al.'CVPR'2015'&'JMLR'2016

slide-32
SLIDE 32

HC'Shin'et'al.'CVPR'2015'&'JMLR'2016

6XMRMRK@ITSX 9QKI

slide-33
SLIDE 33

BSTMG0IXXI

HC'Shin'et'al.'CVPR'2015'&'JMLR'2016

slide-34
SLIDE 34

6XMRMRK@ITSX 9QKI

HC'Shin'et'al.'CVPR'2015'&'JMLR'2016

slide-35
SLIDE 35

HC'Shin'et'al.'CVPR'2016

6XMRMRK@ITSX 9QKI

slide-36
SLIDE 36

HC'Shin'et'al.'CVPR'2016

6XMRMRK@ITSX 9QKI

slide-37
SLIDE 37

X'Wang'et'al.'WACV'2017

6XMRMRK@ITSX 9QKI

slide-38
SLIDE 38

X'Wang'et'al.'CVPR'2017

6XMRMRK@ITSX 9QKI

slide-39
SLIDE 39

X'Wang'et'al.'CVPR'2017

5LIXD.

slide-40
SLIDE 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
slide-41
SLIDE 41
slide-42
SLIDE 42

Acknowledgments

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

Programs:

  • Fogarty
  • ISTP
  • IRTA
  • BESIP
  • CRTP
  • Nvidia for'GPU'

card'donations

slide-43
SLIDE 43

To'Learn'More'…

www.cc.nih.gov/drd/summers.html

X'Wang'et'al.'RSNA'2016