Bulletin of the American Physical Society
APS March Meeting 2024
Monday–Friday, March 4–8, 2024; Minneapolis & Virtual
Session Z18: Data science, AI, and machine learning in physics II
11:30 AM–2:18 PM,
Friday, March 8, 2024
Room: M100I
Sponsoring
Units:
GDS GMED
Chair: Neha Goswami, University of Illinois Urbana-Champaign
Abstract: Z18.00002 : Longitudinal Interpretability of Deep-Learning based Breast Cancer Risk Prediction Model*
11:42 AM–11:54 AM
Presenter:
Brayden Schott
(University of Wisconsin - Madison)
Authors:
Zan Klanecek
(University of Ljubljana, Faculty of Mathematics and Physics, Medical Physics, Ljubljana, Slovenia)
Yao Kuan Wang
(KU Leuven, Department of Imaging and Pathology, Division of Medical Physics & Quality Assessment, Leuven, Belgium)
Tobias Wagner
(KU Leuven, Department of Imaging and Pathology, Division of Medical Physics & Quality Assessment, Leuven, Belgium)
Lesley Cockmartin
(UZ Leuven, Department of Radiology, Leuven, Belgium)
Nicholas Marshall
(KU Leuven, Department of Imaging and Pathology, Division of Medical Physics & Quality Assessment, Leuven, Belgium and UZ Leuven, Department of Radiology, Leuven, Belgium)
Brayden Schott
(University of Wisconsin - Madison)
Alison Deatsch
(University of Wisconsin - Madison)
Andrej Studen
(University of Ljubljana, Faculty of Mathematics and Physics, Medical Physics, Ljubljana, Slovenia and Jožef Stefan Institute, Ljubljana, Slovenia)
Miloš Vrhovec
(Institute of Oncology Ljubljana, Ljubljana, Slovenia)
Hilde Bosmans
(KU Leuven, Department of Imaging and Pathology, Division of Medical Physics & Quality Assessment, Leuven, Belgium and UZ Leuven, Department of Radiology, Leuven, Belgium)
Robert Jeraj
(University of Ljubljana, Faculty of Mathematics and Physics, Slovenia and Jožef Stefan Institute, Slovenia and University of Wisconsin-Madison, Madison, U.S.A.)
The Mirai DL model was employed for BCR predictions using a dataset containing 1210 mammography studies of patients who developed cancer within six years. Each study included four images (two views per breast). For each BCR prediction, gradients with respect to each pixel of the four input images were calculated. The attribution (importance to model prediction) across both views for each breast was summed. It was assumed that the cancer would form in the breast with higher relative attribution. ROC-AUC analysis was performed for six-time points (during cancer and each year before cancer from 2y-6y).
Results showed that the model heavily relies on the attribution from the breast where cancer is already detectable (AUC=0.92±0.02). For the studies where the cancer was not yet present, the high AUC for 2y (0.91±0.07) decreased monotonically to 0.51±0.15 for 6y before cancer.
This interpretability research suggests that the model's predictions for time points closer to cancer occurrence primarily rely on the signal derived from the breast where cancer will form. In contrast, the relative difference is random for more distant time points.
*The authors acknowledge the financial support from the Slovenian Research Agency (research core funding P1-0389) and the Research Foundation – Flanders (research core funding G0A7121N).
Follow Us |
Engage
Become an APS Member |
My APS
Renew Membership |
Information for |
About APSThe American Physical Society (APS) is a non-profit membership organization working to advance the knowledge of physics. |
© 2024 American Physical Society
| All rights reserved | Terms of Use
| Contact Us
Headquarters
1 Physics Ellipse, College Park, MD 20740-3844
(301) 209-3200
Editorial Office
100 Motor Pkwy, Suite 110, Hauppauge, NY 11788
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700