Contrastive Learning Approach to Predict Radiologist's Error Based on Gaze Data
Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review
Standard
Contrastive Learning Approach to Predict Radiologist's Error Based on Gaze Data. / Pershin, Ilya; Mustafaev, Tamerlan; Ibragimov, Bulat.
2023 IEEE Congress on Evolutionary Computation, CEC 2023. IEEE, 2023.Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review
Harvard
APA
Vancouver
Author
Bibtex
}
RIS
TY - GEN
T1 - Contrastive Learning Approach to Predict Radiologist's Error Based on Gaze Data
AU - Pershin, Ilya
AU - Mustafaev, Tamerlan
AU - Ibragimov, Bulat
N1 - Publisher Copyright: © 2023 IEEE.
PY - 2023
Y1 - 2023
N2 - The increase in medical imaging and, consequently, the growing workload of radiologists requires workflow optimization. The interaction between radiologists and artificial intelligence is a promising area because it can improve radiologist productivity and reduce diagnostic errors. In this study, we describe an eye-tracking experiment with four practicing radiologists who cumulatively examined 1,000 unique chest X-ray images that included samples with multiple pathologies. We demonstrate the feasibility of a self-supervised approach for gazebased data for problems with a small amount of labeled data. We achieve an accuracy of 64.7% for binary classification of radiologist diagnostic error with 218 samples per label. The proposed approach can be used in the system for assessing the need for a second opinion to improve the quality of diagnostics.
AB - The increase in medical imaging and, consequently, the growing workload of radiologists requires workflow optimization. The interaction between radiologists and artificial intelligence is a promising area because it can improve radiologist productivity and reduce diagnostic errors. In this study, we describe an eye-tracking experiment with four practicing radiologists who cumulatively examined 1,000 unique chest X-ray images that included samples with multiple pathologies. We demonstrate the feasibility of a self-supervised approach for gazebased data for problems with a small amount of labeled data. We achieve an accuracy of 64.7% for binary classification of radiologist diagnostic error with 218 samples per label. The proposed approach can be used in the system for assessing the need for a second opinion to improve the quality of diagnostics.
KW - deep learning
KW - eye-tracking
KW - human-AI interaction
KW - self-supervised
UR - http://www.scopus.com/inward/record.url?scp=85174491896&partnerID=8YFLogxK
U2 - 10.1109/CEC53210.2023.10254056
DO - 10.1109/CEC53210.2023.10254056
M3 - Article in proceedings
AN - SCOPUS:85174491896
BT - 2023 IEEE Congress on Evolutionary Computation, CEC 2023
PB - IEEE
T2 - 2023 IEEE Congress on Evolutionary Computation, CEC 2023
Y2 - 1 July 2023 through 5 July 2023
ER -
ID: 372615699