TY - JOUR
T1 - Monitoring Glaucomatous Functional Loss Using an Artificial Intelligence–Enabled Dashboard
AU - Yousefi, Siamak
AU - Elze, Tobias
AU - Pasquale, Louis R.
AU - Saeedi, Osamah
AU - Wang, Mengyu
AU - Shen, Lucy Q.
AU - Wellik, Sarah R.
AU - De Moraes, Carlos G.
AU - Myers, Jonathan S.
AU - Boland, Michael V.
N1 - Funding Information:
The authors were funded by National Institute of Health , National Eye Institute Grant R21 EY030142 (to S.Y., T.E., M.V.B., L.R.P.), R01 EY015473 (to L.R.P.), and in part by an unrestricted grant from Research to Prevent Blindness (New York, NY) and BrightFocus Foundation (to T.E.), National Eye Institute R01EY030575 (to T.E.), and P30EY003790 (to T.E.). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Data for the second benchmark cohort were acquired from the Assessing the Effectiveness of Imaging Technology to Rapidly Detect Disease Progression in Glaucoma study, and were provided courtesy of Garway-Heath, National Institute for Health Research Biomedical Research Centre at Moorfields Eye Hospital NHS Foundation Trust and UCL Institute of Ophthalmology (London, UK). The views and opinions expressed by authors in this publication are those of the authors and do not necessarily reflect those of the NHS, the National Institute for Health Research Health Technology Assessment Program, or the UK Department of Health.
Funding Information:
The authors were funded by National Institute of Health, National Eye Institute Grant R21 EY030142 (to S.Y., T.E., M.V.B., L.R.P.), R01 EY015473 (to L.R.P.), and in part by an unrestricted grant from Research to Prevent Blindness (New York, NY) and BrightFocus Foundation (to T.E.), National Eye Institute R01EY030575 (to T.E.), and P30EY003790 (to T.E.). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Data for the second benchmark cohort were acquired from the Assessing the Effectiveness of Imaging Technology to Rapidly Detect Disease Progression in Glaucoma study, and were provided courtesy of Garway-Heath, National Institute for Health Research Biomedical Research Centre at Moorfields Eye Hospital NHS Foundation Trust and UCL Institute of Ophthalmology (London, UK). The views and opinions expressed by authors in this publication are those of the authors and do not necessarily reflect those of the NHS, the National Institute for Health Research Health Technology Assessment Program, or the UK Department of Health.
Publisher Copyright:
© 2020 American Academy of Ophthalmology
PY - 2020/9
Y1 - 2020/9
N2 - Purpose: To develop an artificial intelligence (AI) dashboard for monitoring glaucomatous functional loss. Design: Retrospective, cross-sectional, longitudinal cohort study. Participants: Of 31 591 visual fields (VFs) on 8077 subjects, 13 231 VFs from the most recent visit of each patient were included to develop the AI dashboard. Longitudinal VFs from 287 eyes with glaucoma were used to validate the models. Method: We entered VF data from the most recent visit of glaucomatous and nonglaucomatous patients into a “pipeline” that included principal component analysis (PCA), manifold learning, and unsupervised clustering to identify eyes with similar global, hemifield, and local patterns of VF loss. We visualized the results on a map, which we refer to as an “AI-enabled glaucoma dashboard.” We used density-based clustering and the VF decomposition method called “archetypal analysis” to annotate the dashboard. Finally, we used 2 separate benchmark datasets—one representing “likely nonprogression” and the other representing “likely progression”—to validate the dashboard and assess its ability to portray functional change over time in glaucoma. Main Outcome Measures: The severity and extent of functional loss and characteristic patterns of VF loss in patients with glaucoma. Results: After building the dashboard, we identified 32 nonoverlapping clusters. Each cluster on the dashboard corresponded to a particular global functional severity, an extent of VF loss into different hemifields, and characteristic local patterns of VF loss. By using 2 independent benchmark datasets and a definition of stability as trajectories not passing through over 2 clusters in a left or downward direction, the specificity for detecting “likely nonprogression” was 94% and the sensitivity for detecting “likely progression” was 77%. Conclusions: The AI-enabled glaucoma dashboard, developed using a large VF dataset containing a broad spectrum of visual deficit types, has the potential to provide clinicians with a user-friendly tool for determination of the severity of glaucomatous vision deficit, the spatial extent of the damage, and a means for monitoring the disease progression.
AB - Purpose: To develop an artificial intelligence (AI) dashboard for monitoring glaucomatous functional loss. Design: Retrospective, cross-sectional, longitudinal cohort study. Participants: Of 31 591 visual fields (VFs) on 8077 subjects, 13 231 VFs from the most recent visit of each patient were included to develop the AI dashboard. Longitudinal VFs from 287 eyes with glaucoma were used to validate the models. Method: We entered VF data from the most recent visit of glaucomatous and nonglaucomatous patients into a “pipeline” that included principal component analysis (PCA), manifold learning, and unsupervised clustering to identify eyes with similar global, hemifield, and local patterns of VF loss. We visualized the results on a map, which we refer to as an “AI-enabled glaucoma dashboard.” We used density-based clustering and the VF decomposition method called “archetypal analysis” to annotate the dashboard. Finally, we used 2 separate benchmark datasets—one representing “likely nonprogression” and the other representing “likely progression”—to validate the dashboard and assess its ability to portray functional change over time in glaucoma. Main Outcome Measures: The severity and extent of functional loss and characteristic patterns of VF loss in patients with glaucoma. Results: After building the dashboard, we identified 32 nonoverlapping clusters. Each cluster on the dashboard corresponded to a particular global functional severity, an extent of VF loss into different hemifields, and characteristic local patterns of VF loss. By using 2 independent benchmark datasets and a definition of stability as trajectories not passing through over 2 clusters in a left or downward direction, the specificity for detecting “likely nonprogression” was 94% and the sensitivity for detecting “likely progression” was 77%. Conclusions: The AI-enabled glaucoma dashboard, developed using a large VF dataset containing a broad spectrum of visual deficit types, has the potential to provide clinicians with a user-friendly tool for determination of the severity of glaucomatous vision deficit, the spatial extent of the damage, and a means for monitoring the disease progression.
UR - http://www.scopus.com/inward/record.url?scp=85083302769&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85083302769&partnerID=8YFLogxK
U2 - 10.1016/j.ophtha.2020.03.008
DO - 10.1016/j.ophtha.2020.03.008
M3 - Article
C2 - 32317176
AN - SCOPUS:85083302769
SN - 0161-6420
VL - 127
SP - 1170
EP - 1178
JO - Ophthalmology
JF - Ophthalmology
IS - 9
ER -