Fundus Image-based Visual Acuity Assessment with PAC-Guarantees

Sooyong Jang, Kuk Jin Jang, Hyonyoung Choi, Yong-Seop Han, Seongjin Lee, Jin-hyun Kim, Insup Lee
Proceedings of the 4th Machine Learning for Health Symposium, PMLR 259:535-549, 2025.

Abstract

Timely detection and treatment are essential for maintaining eye health. Visual acuity (VA), which measures the clarity of vision at a distance, is a crucial metric for managing eye health. Machine learning (ML) techniques have been introduced to assist in VA measurement, potentially alleviating clinicians’ workloads. However, the inherent uncertainties in ML models make relying solely on them for VA prediction less than ideal. The VA prediction task involves multiple sources of uncertainty, requiring more robust approaches. A promising method is to build prediction sets or intervals rather than point estimates, offering coverage guarantees through techniques like conformal prediction and Probably Approximately Correct (PAC) prediction sets. Despite the potential, to date, these approaches have not been applied to the VA prediction task.To address this, we propose a method for deriving prediction intervals for estimating visual acuity from fundus images with a PAC guarantee. Our experimental results demonstrate that the PAC guarantees are upheld, with performance comparable to or better than that of two prior works that do not provide such guarantees.

Cite this Paper


BibTeX
@InProceedings{pmlr-v259-jang25a, title = {Fundus Image-based Visual Acuity Assessment with PAC-Guarantees}, author = {Jang, Sooyong and Jang, Kuk Jin and Choi, Hyonyoung and Han, Yong-Seop and Lee, Seongjin and Kim, Jin-hyun and Lee, Insup}, booktitle = {Proceedings of the 4th Machine Learning for Health Symposium}, pages = {535--549}, year = {2025}, editor = {Hegselmann, Stefan and Zhou, Helen and Healey, Elizabeth and Chang, Trenton and Ellington, Caleb and Mhasawade, Vishwali and Tonekaboni, Sana and Argaw, Peniel and Zhang, Haoran}, volume = {259}, series = {Proceedings of Machine Learning Research}, month = {15--16 Dec}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v259/main/assets/jang25a/jang25a.pdf}, url = {https://proceedings.mlr.press/v259/jang25a.html}, abstract = {Timely detection and treatment are essential for maintaining eye health. Visual acuity (VA), which measures the clarity of vision at a distance, is a crucial metric for managing eye health. Machine learning (ML) techniques have been introduced to assist in VA measurement, potentially alleviating clinicians’ workloads. However, the inherent uncertainties in ML models make relying solely on them for VA prediction less than ideal. The VA prediction task involves multiple sources of uncertainty, requiring more robust approaches. A promising method is to build prediction sets or intervals rather than point estimates, offering coverage guarantees through techniques like conformal prediction and Probably Approximately Correct (PAC) prediction sets. Despite the potential, to date, these approaches have not been applied to the VA prediction task.To address this, we propose a method for deriving prediction intervals for estimating visual acuity from fundus images with a PAC guarantee. Our experimental results demonstrate that the PAC guarantees are upheld, with performance comparable to or better than that of two prior works that do not provide such guarantees.} }
Endnote
%0 Conference Paper %T Fundus Image-based Visual Acuity Assessment with PAC-Guarantees %A Sooyong Jang %A Kuk Jin Jang %A Hyonyoung Choi %A Yong-Seop Han %A Seongjin Lee %A Jin-hyun Kim %A Insup Lee %B Proceedings of the 4th Machine Learning for Health Symposium %C Proceedings of Machine Learning Research %D 2025 %E Stefan Hegselmann %E Helen Zhou %E Elizabeth Healey %E Trenton Chang %E Caleb Ellington %E Vishwali Mhasawade %E Sana Tonekaboni %E Peniel Argaw %E Haoran Zhang %F pmlr-v259-jang25a %I PMLR %P 535--549 %U https://proceedings.mlr.press/v259/jang25a.html %V 259 %X Timely detection and treatment are essential for maintaining eye health. Visual acuity (VA), which measures the clarity of vision at a distance, is a crucial metric for managing eye health. Machine learning (ML) techniques have been introduced to assist in VA measurement, potentially alleviating clinicians’ workloads. However, the inherent uncertainties in ML models make relying solely on them for VA prediction less than ideal. The VA prediction task involves multiple sources of uncertainty, requiring more robust approaches. A promising method is to build prediction sets or intervals rather than point estimates, offering coverage guarantees through techniques like conformal prediction and Probably Approximately Correct (PAC) prediction sets. Despite the potential, to date, these approaches have not been applied to the VA prediction task.To address this, we propose a method for deriving prediction intervals for estimating visual acuity from fundus images with a PAC guarantee. Our experimental results demonstrate that the PAC guarantees are upheld, with performance comparable to or better than that of two prior works that do not provide such guarantees.
APA
Jang, S., Jang, K.J., Choi, H., Han, Y., Lee, S., Kim, J. & Lee, I.. (2025). Fundus Image-based Visual Acuity Assessment with PAC-Guarantees. Proceedings of the 4th Machine Learning for Health Symposium, in Proceedings of Machine Learning Research 259:535-549 Available from https://proceedings.mlr.press/v259/jang25a.html.

Related Material