TY - JOUR
T1 - An Artificial Intelligence Enabled System for Retinal Nerve Fiber Layer Thickness Damage Severity Staging
AU - Yousefi, Siamak
AU - Huang, Xiaoqin
AU - Poursoroush, Asma
AU - Majoor, Juleke E A
AU - Lemij, Hans
AU - Vermeer, Koen
AU - Elze, Tobias
AU - Wang, Mengyu
AU - Nouri-Mahdavi, Kouros
AU - Mohammadzadeh, Vahid
AU - Brusini, Paolo
AU - Johnson, Chris
N1 - © 2023 by the American Academy of Ophthalmology.
PY - 2024/3
Y1 - 2024/3
N2 - Purpose To develop an objective glaucoma damage severity classification system based on optical coherence tomography (OCT)-derived retinal nerve fiber layer (RNFL) thickness measurements. Design Algorithm development for RNFL damage severity classification based on multicenter OCT data. Subjects and Participants A total of 6561 circumpapillary RNFL profiles from 2269 eyes of 1171 subjects to develop models and 2505 RNFL profiles from 1099 eyes of 900 subjects to validate models. Methods We developed an unsupervised k-means model to identify clusters of eyes with similar RNFL thickness profiles. We annotated the clusters based on their respective global RNFL thickness. We computed the optimal global RNFL thickness thresholds that discriminated different severity levels based on Bayes’ minimum error principle. We validated the proposed pipeline based on an independent validation dataset with 2505 RNFL profiles from 1099 eyes of 900 subjects. Main Outcome Measures Accuracy, area under the receiver operating characteristic curve (AUC), and confusion matrix. Results The k-means clustering discovered four clusters with 1382, 1613, 1727, and 1839 samples with mean global RNFL thickness of 58.3 μm (±8.9: Standard Deviation), 78.9 μm (±6.7), 87.7 μm (±8.2), and 101.5 μm (±7.9). The Bayes’ minimum error classifier identified optimal global RNFL values of >95 μm, 86 – 95 μm, and 70 – 85 μm and
AB - Purpose To develop an objective glaucoma damage severity classification system based on optical coherence tomography (OCT)-derived retinal nerve fiber layer (RNFL) thickness measurements. Design Algorithm development for RNFL damage severity classification based on multicenter OCT data. Subjects and Participants A total of 6561 circumpapillary RNFL profiles from 2269 eyes of 1171 subjects to develop models and 2505 RNFL profiles from 1099 eyes of 900 subjects to validate models. Methods We developed an unsupervised k-means model to identify clusters of eyes with similar RNFL thickness profiles. We annotated the clusters based on their respective global RNFL thickness. We computed the optimal global RNFL thickness thresholds that discriminated different severity levels based on Bayes’ minimum error principle. We validated the proposed pipeline based on an independent validation dataset with 2505 RNFL profiles from 1099 eyes of 900 subjects. Main Outcome Measures Accuracy, area under the receiver operating characteristic curve (AUC), and confusion matrix. Results The k-means clustering discovered four clusters with 1382, 1613, 1727, and 1839 samples with mean global RNFL thickness of 58.3 μm (±8.9: Standard Deviation), 78.9 μm (±6.7), 87.7 μm (±8.2), and 101.5 μm (±7.9). The Bayes’ minimum error classifier identified optimal global RNFL values of >95 μm, 86 – 95 μm, and 70 – 85 μm and
UR - https://www.mendeley.com/catalogue/4f1e851a-0164-3e76-99ce-ac3f4eb5d95d/
U2 - 10.1016/j.xops.2023.100389
DO - 10.1016/j.xops.2023.100389
M3 - Article
C2 - 37868793
SN - 2666-9145
VL - 4
SP - 100389
JO - Ophthalmology Science
JF - Ophthalmology Science
IS - 2
ER -