LOCALISATION OF RACIAL INFORMATION IN CHEST X-RAY FOR DEEP LEARNING DIAGNOSIS

被引：2

作者：

Salvado, Olivier ^{[1
,2
]}

Konate, Salamata ^{[1
,2
]}

Cruz, Rodrigo Santa ^{[1
,2
]}

Bdadley, Andrew ^{[2
]}

Gichoya, Judy Wawira ^{[3
]}

Seyyed-Kalantari, Laleh ^{[4
]}

Price, Brandon ^{[5
]}

Fookes, Clinton ^{[2
]}

Lebrat, Leo ^{[1
,2
]}

机构：

[1] CSIRO Data61, Imaging & Comp Vis Grp, Eveleigh, Australia

[2] Queensland Univ Technol, SAIVT, Brisbane, Australia

[3] Emory Univ, Dept Radiol & Imaging Sci, Atlanta, GA USA

[4] York Univ, Elect Engn & Comp Sci, York, N Yorkshire, England

[5] Florida State Univ, Coll Med, Tallahassee, FL USA

来源：

IEEE INTERNATIONAL SYMPOSIUM ON BIOMEDICAL IMAGING, ISBI 2024 | 2024年

关键词：

saliency maps; eXplainable AI (XAI); chest; x-ray; deep learning; atlas-based registration;

D O I：

10.1109/ISBI56570.2024.10635445

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

Deep learning-based classification of diseases from Chest X-ray has been shown to use implicit information about the subjects' self-reported race, which could result in diagnostic bias. In this paper, we describe and compare two approaches to investigate where racial information is located in the image: first leveraging non-linear registration and computing atlas differences and second using saliency maps. We compute a map visualising the racial information between black and white subjects and discuss whether those maps are consistent with the model explanation.

引用

页数：4

共 11 条

[1]

[Anonymous], 2018, RSNA pneumonia detection challenge

[2]

Ashburner J., 2009, ENCY NEUROSCIENCE, P471, DOI [DOI 10.1016/B978-008045046-9.00306-5, 10.1016/B978-008045046-9.00306-5]

[3]

Avants BB., 2009, Insight J, V2, P1, DOI DOI 10.54294/UVNHIN

[4] BOLD5000, a public fMRI dataset while viewing 5000 visual images [J].

Chang, Nadine ;

Pyles, John A. ;

Marcus, Austin ;

Gupta, Abhinav ;

Tarr, Michael J. ;

Aminoff, Elissa M. .

SCIENTIFIC DATA, 2019, 6 (1)

[5] Understanding Deep Networks via Extremal Perturbations and Smooth Masks [J].

Fong, Ruth ;

Patrick, Mandela ;

Vedaldi, Andrea .

2019 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV 2019), 2019, :2950-2958

[6]

Gichoya Judy Wawira, 2022, LANCET DIGITAL HLTH

[7]

Konate Salamata, 2023, ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), P1, DOI 10.1109/ICASSP49357.2023.10097093

[8]

Petsiuk V., 2018, arXiv

[9] Effect of a comprehensive deep-learning model on the accuracy of chest x-ray interpretation by radiologists: a retrospective, multireader multicase study [J].

Seah, Jarrel C. Y. ;

Tang, Cyril H. M. ;

Buchlak, Quinlan D. ;

Holt, Xavier G. ;

Wardman, Jeffrey B. ;

Aimoldin, Anuar ;

Esmaili, Nazanin ;

Ahmad, Hassan ;

Hung Pham ;

Lambert, John F. ;

Hachey, Ben ;

Hogg, Stephen J. F. ;

Johnston, Benjamin P. ;

Bennett, Christine ;

Oakden-Rayner, Luke ;

Brotchie, Peter ;

Jones, Catherine M. .

LANCET DIGITAL HEALTH, 2021, 3 (08) :E496-E506

[10]

Selvaraju RR, 2020, INT J COMPUT VISION, V128, P336, DOI [10.1109/ICCV.2017.74, 10.1007/s11263-019-01228-7]

← 1 2 →