Machine Learning-Guided Prediction of Central Anterior Chamber Depth Using Slit Lamp Images from a Portable Smartphone Device.

Yuki Sasa,Clement Tan,David Chen,Yvonne Ho,Jieying Lee,Ching Chiuan Yen

doi:10.3390/bios11060182

Yuki Sasa, Clement Tan + Show 4 more

Open Access

https://doi.org/10.3390/bios11060182

Copy DOI

Abstract

There is currently no objective portable screening modality for narrow angles in the community. In this prospective, single-centre image validation study, we used machine learning on slit lamp images taken with a portable smartphone device (MIDAS) to predict the central anterior chamber depth (ACD) of phakic patients with undilated pupils. Patients 60 years or older with no history of laser or intraocular surgery were recruited. Slit lamp images were taken with MIDAS, followed by anterior segment optical coherence tomography (ASOCT; Casia SS-1000, Tomey, Nagoya, Japan). After manual annotation of the anatomical landmarks of the slit lamp photos, machine learning was applied after image processing and feature extraction to predict the ACD. These values were then compared with those acquired from the ASOCT. Sixty-six eyes (right = 39, 59.1%) were included for analysis. The predicted ACD values formed a strong positive correlation with the measured ACD values from ASOCT (R2 = 0.91 for training data and R2 = 0.73 for test data). This study suggests the possibility of estimating central ACD using slit lamp images taken from portable devices.

Highlights

Angle closure glaucoma is a cause of major visual impairment in Asia; 87% of patients with angle closure glaucoma globally are Asian [1]
Common parameters used to describe the features of the anterior chamber angle include the angle opening distance (AOD), the anterior chamber area (ACA), and the trabecular-iris space area (TISA), with AOD750
This pilot study demonstrated that using images captured from a portable slit lamp device, it was possible topilot use machine learning tothat predict central

Summary

Introduction

Angle closure glaucoma is a cause of major visual impairment in Asia; 87% of patients with angle closure glaucoma globally are Asian [1]. The prevalence of patients with all occludable angles (including primary angle closure and primary angle closure glaucoma) is estimated to be 6.3% in Singaporeans aged 40 years old and above [2], of whom 79% may not have been diagnosed before [3]. Age and sex standardized incidence of acute angle closure glaucoma (AACG) was reported as being up to 15.5 cases/100,000 person-years among Chinese Singaporeans [4,5]. The advent of anterior segment optical coherence tomography imaging (ASOCT) has made anterior chamber assessment both quantitative and objective. Common parameters used to describe the features of the anterior chamber angle include the angle opening distance (AOD), the anterior chamber area (ACA), and the trabecular-iris space area (TISA), with AOD750

Objectives

Results

Discussion

Conclusion