Automated Multimodal Detection and Analysis of Geographic Atrophy
Geographic atrophy (GA) is a form of age-related macular degeneration (AMD), and increasingly the main cause of vision loss in patients. Much of the previous research of GA has focused on individual imaging modalities, utilizing two-dimensional (2D) information alone. However, considering the 3D topology of the disease, utilizing information from all imaging modalities concomitantly could potentially yield a more precise and comprehensive depiction of GA lesions. The overall goal of this project is to develop an automated multimodal GA segmentation system to more precisely quantify GA progression over time in multimodal 2D and 3D images to facilitate the understanding of GA’s relationship to vision loss.
The overall goal of this project is to develop an automated multimodal segmentation system to more precisely quantify the progression of geographic atrophy (GA), a type of advanced-stage eye disease, over time in multimodal 2D and 3D images to facilitate our understanding of GA’s relationship to vision loss.
GA is the late-stage of age-related macular degeneration (AMD) and is increasingly the main cause of vision loss in patients. Research efforts (including ours) to develop methods for the automated identification and quantitative analysis of GA in various eye images have been reported. However, much of this previous research has focused on individual modalities, utilizing 2D information alone. Considering the 3D topology of AMD, an approach that utilizes information from all imaging modalities concomitantly could potentially yield a more precise and comprehensive depiction of GA lesions.
This project includes two major aims. In Aim 1, we develop and validate an automated segmentation system for detecting GA in different 2D and 3D (optical coherence tomography, or OCT) imaging modalities. To do so, we align each individual 2D image to the corresponding OCT image using a feature-based image registration algorithm. We then apply the multimodal GA segmentation by combining image features from different 2D modalities and the 3D OCT images.
In Aim 2, we derive optimal multimodal definitions of GA and identify the most predictive value of subsequent growth of the GA lesions over time. Various multimodal GA descriptors/features are generated from different modality images and are correlated with the microperimetry sensitivity to establish which GA descriptors/features are most predictive of function. We also establish which GA descriptors/features are most predictive of subsequent GA growth over time.
The deliverable from this research program is a fully automated system for the detection of GA lesions and the quantitative analysis of GA progression. We will make the developed system accessible to the broad research community. Furthermore, current advances of multimodal imaging make the translation of the multimodal segmentation practical for routine use in the clinical setting. This proposal is expected to facilitate the understanding of the pathogenesis of GA in research and to facilitate the diagnosis of GA in routine clinic environments.
About the Researcher
Zhihong Hu, PhD, is a scientist/senior research engineer at the Doheny Eye Institute (DEI) – University of California, Los Angeles (UCLA). She is the supervisor of the Doheny Image Analysis Laboratory (DIAL). Dr. Hu obtained her PhD from the University of Iowa (UI), her master of science degree from the Technical University of Denmark (DTU), and her bachelor of science degree from Zhengzhou University (ZZU) in China. Dr. Hu’s overall research interests are in the fields of medical image processing, computer vision, machine learning, video surveillance, and lasers. Her latest research interest focuses on ophthalmic image processing and analysis, including image segmentation and registration specifically in graph-based image segmentation; supervised pixel classification; deep learning, multi-object, multi-modality image segmentation; and classification and image registration in 3D optical coherence tomography (OCT) and various 2D ophthalmic images.
I have been working in different areas in applied engineering for many years. I like working on the engineering field as it is closely related to the needs of our daily life. I got to be involved in the research of ophthalmic image processing during my doctor studies in electrical and computer engineering at the University of Iowa, where I specialized in the detection and analysis of the glaucomatous 3D eye structures involving the neural canal opening (NCO), optic cup, rim, and blood vessels using 2D and 3D (optical coherence tomography, or OCT) imaging. Glaucoma is the second leading cause of blindness in the United States and globally. One crucial contribution I made to the vision community is the detection and quantification analysis of the neural NCO. The NCO is a 3D single anatomic structure in OCT volumes (versus traditional 2D optic disc margin) that I and other groups proved was an effective reference point to derive various glaucomatous measurements.
After I joined in Doheny Eye Institute, I turned my attention to age-related macular degeneration (AMD), which is the leading cause of blindness in the United States and globally. My research specializes in the detection and analysis of various eye structures and lesions in AMD using various 2D and 3D (OCT) images. I developed several automated segmentation algorithms in the macular region to facilitate the understanding of the mechanism of dry AMD. For instance, I developed a graph-based algorithm to identify 11 retinal layers in macular spectral domain (SD)-OCT volumetric scans. The automated multiple layer segmentation facilitates the analysis of the retinal layer thickness and reflectivity changes, which are believed to be the indicators/predictors of various eye diseases (e.g. dry AMD) and their progression.
Most recently, I developed a semi-automated algorithm and a fully automated algorithm for the detection and quantification analysis of geographic atrophy (GA), i.e., the late stage of AMD. Based on the segmented GA lesions, I performed a GA directional growth analysis. The result was submitted to ARVO 2016, and a NEI Travel Grant was rewarded. I felt I was so lucky to be able to contribute in the area of preventing blindness, especially to conduct research in GA detection and analysis. I have been looking for grants to be able to continue this interesting project. Very luckily, we receive the Macular Degeneration Research grant, which is fantastic and comes just in time. With the great support of BrightFocus donors, we expect to provide a more robust GA segmentation system and perform more interesting analysis. I am very excited about this opportunity and look forward to contributing more to the amazing research area that is directed to the prevention of blindness.
First published on: July 19, 2016
Last modified on: July 1, 2018