ABSTRACT
Background: Traditional methods for neuroretinal rim width measurement in spectral domain optical coherence tomography (SD-OCT) employs the Bruch‘s membrane opening (BMO) as the anatomical border of the rim, referenced to a BMO horizontal reference plane, termed as “Bruch’s Membrane Opening-Horizontal Rim Width” (BMO-HRW). BMO-HRW is defined as the distance between BMO and internal limiting membrane (ILM) on the horizontal plane. In contrast, the Spectralis OCT (Heidelberg Engineering, Germany) employs a new parameter called “Bruch’s Membrane Opening–Minimum Rim Width” (BMO-MRW) with Glaucoma Module Premium Edition (GMPE). GMPE provides a novel objective method of optic nerve head (ONH) analysis using BMO, but the neuroretinal rim assessment is performed from the BMO to the nearest point on the ILM, rather than on the horizontal reference plane. It is the BMO-MRW and is defined as the minimum distance between the BMO and ILM in the ONH. Purpose: In this video, anatomy of the ONH and GMPE is decoded from a neophyte user’s point of view, as to why BMO-MRW is more important than the traditional BMO-HRW for glaucoma evaluation. Synopsis: The GMPE concepts are depicted in a novel dynamic (Clinical vs OCT Vs Histology) screenplay, detailing the below focal points with 2D & 3D animations: True Margin of ONH, Bruch’s Membrane (BM), Histology Vs OCT, BMO, Bruch’s Membrane Opening-Minimum Rim Width, Bruch’s Membrane Opening-Minimum Rim Width Versus Bruch’s Membrane Opening-Horizontal Rim Width, Alpha, Beta, Gamma Zone of ONH in OCT, Anatomic Positioning System, Impact of Fovea Bruch’s Membrane Opening Centre Axis. Highlights: This video also highlights, how with the advent of Anatomic Positioning System, scans were able to align relative to the individual’s Fovea-to-BMO-center (FoBMOC) axis at every follow-up, for accurately detecting changes, as small as 1 micron in BMO-MRW, thus creating a new world in diagnosing glaucoma and detecting glaucomatous progression with precision.
ABSTRACT
Purpose: For diagnosing glaucomatous damage, we have employed a novel convolutional neural network (CNN) from TrueColor confocal fundus images to conquer the black box dilemma in artificial intelligence (AI). This neural network with CNN architecture with human?in?the?loop (HITL) data annotation helps not only in diagnosing glaucoma but also in predicting and locating detailed signs in the glaucomatous fundus, such as splinter hemorrhages, glaucomatous optic atrophy, vertical glaucomatous cupping, peripapillary atrophy, and retinal nerve fiber layer (RNFL) defect. Methods: The training was done on a well?curated private dataset of 1,400 high?resolution confocal fundus images, out of which 1,120 images (80%) were used exclusively for training and 280 images (20%) were used exclusively for testing. A custom trained You Only Look Once version 5 (YOLOv5)?based object detection methodology was used to identify the underlying conditions precisely. Twenty?six predefined medical conditions were annotated by a team of humans (comprising two glaucoma specialists and two optometrists) by using the Microsoft Visual Object Tagging Tool (VoTT) tool. The 280 testing images were split into three groups (90,100, and 90 images) for three test runs done once every 15 days. Results: Test results showed consistent increments in the accuracy, from 94.44% to 98.89%, in predicting the glaucoma diagnosis along with the detailed signs of the glaucomatous fundus. Conclusion: Utilizing human intelligence in AI for detecting glaucomatous fundus images by using HITL machine learning has never been reported in the literature before. This AI model not only has good sensitivity and specificity in accurate glaucoma predictions but is also an explainable AI, thus overcoming the black box dilemma.