Conference proceeding
The relationship between eye tracking features and transfer learning in modeling decision prediction of radiologists reading mammograms
Vol.13409, pp.1340906-1340906-7
Progress in Biomedical Optics and Imaging
04/10/2025
DOI: 10.1117/12.3048803
Abstract
Predicting radiologists’ decisions when reading mammograms is a novel way to reduce the number of false positives and false negatives made at breast cancer screening. In this study, we aimed to enhance the accuracy of predicting radiologists’ decisions in mammography by leveraging transfer learning. Our dataset comprised 120 digital mammogram cases, each annotated with radiologists’ decisions categorized as true positive (TP), false positive (FP), or false negative (FN). We adopted the ResNet50 convolutional neural network (CNN) for our modeling approach, developing two different models. In the first model, ResNet50 was pretrained on the ImageNet dataset, with the initial layers frozen and the remaining layers fine-tuned to adapt to our mammography data. The second model was initialized with ImageNet weights obtained in the first model and further pretrained using the VinDr-Mammo dataset, an open-access large-scale Vietnamese dataset of full-field digital mammograms (FFDM) consisting of 5,000 four-view exams with breast-level assessments and extensive lesion-level annotations. Our transfer learning method improved the decision prediction accuracy by leveraging features from the VinDr-Mammo models.
Details
- Title: Subtitle
- The relationship between eye tracking features and transfer learning in modeling decision prediction of radiologists reading mammograms
- Creators
- Karthika Kelat - University of IowaSarah E. Gerard - University of IowaBulat Ibragimov - University of CopenhagenClaudia Mello-Thoms - University of Iowa
- Contributors
- Mark A. Anastasio (Editor) - University of Illinois Urbana-ChampaignJovan G. Brankov (Editor) - Illinois Institute of Technology
- Resource Type
- Conference proceeding
- Publication Details
- Vol.13409, pp.1340906-1340906-7
- Publisher
- SPIE
- Series
- Progress in Biomedical Optics and Imaging
- DOI
- 10.1117/12.3048803
- ISSN
- 1605-7422
- Grant note
- NIH/NCI: 1 R01 CA259048-01
We would like to thank the radiologists that participated in our experiment. We also would like to thank our funding sponsors, NIH/NCI grant 1 R01 CA259048-01.
- Language
- English
- Date published
- 04/10/2025
- Academic Unit
- Roy J. Carver Department of Biomedical Engineering; Radiology
- Record Identifier
- 9984816012202771
Metrics
4 Record Views