Lab Overview
Adam Dubis, PhD, is an enterprise-minded educator and researcher specializing in image and health data analysis. His research team’s work focuses on using the data science and population health methods of large dataset acquisition, curation, and modeling to explain biological processes.
Research interests include ophthalmic imaging analysis, data science, deep-learning and AI, and Safe AI for medicine.
About Our Research
Developing an ethos that bridges clinical, discovery, and translational work, the Dubis Lab is positioned to exploit technologies to improve clinical care. This drive has already led to several commercial spinouts, with more in development.
Having worked in clinical trials for several ophthalmic diseases, the lab also aims to put the patient at the center of care. Core to the group's work is the development of explainable artificial intelligence (AI) and Safe AI technologies, as well as education about the power and limitations of technical developments.
The Dubis Laboratory research group has three primary aims:
- Studying the structural and functional relationships underlying normal vision and retinas affected by inherited retinal diseases. This work has expanded to help develop trials to test several potential therapies to reverse or stop retinal degeneration.
- Using the eye as a model of the brain and investigating retinal vascular function in systemic and neurological diseases. The aim is to develop more sensitive tests of vascular function and neurovascular coupling that may someday be used in treatment trials for vascular disease.
- Applying deep learning techniques to retinal image analysis. The goal is to develop classification tools that may be used to expand ophthalmic expertise to under-resourced countries and increase research efficiency through improved image analysis.
Current Research Projects
This project aims to develop AI pipelines to predict visual function from single and multiple image types. The output would be tools to aid clinical trials with highly robust endpoints in traditionally subjective tasks. The datasets cover a number of diseases.
Current understanding of disease suggests that changes in retinal vessel function occur before structural change and clinically relevant disease. The ability to detect these earlier allows for more opportunity to intervene. Datasets exist for several diseases and clinical trials.
These pipelines are based on proprietary supervised and unsupervised methods to classify patients at high and low risk of disease progression and identify ideal treatment plans. Datasets exist for several diseases.
Research Publications
- Deep Active Learning for Robust Biomedical Segmentation. Arikan M, Willoughby J, Lilaonitkul W, Dubis AM. Nature: Machine Intelligence, 2025 in press.
- OCT5k: A dataset of multi-disease and multi-graded annotations for retinal layers. Arikan M, Willoughby J, Ongun S, Sallo F, Montesel A, Ahmed H, Hagag A, Book M, Faatz H, Cicinelli MV, Fawzi AA, Podkowinski D, Cilkova M, de Almeida D, Zouache M, Ramsamy G, Lilaonitkul W, Dubis AM. Nature: Scientific Data, 2024, In – Press.
- Deep Density Estimation for Cone Counting and Diagnosis of Genetic Eye Diseases From Adaptive Optics Scanning Light Ophthalmoscope Images. Toledo-Cortés S, Dubis AM, González FA, Müller H. Transl Vis Sci Technol. 2023 Nov 1;12(11):25.
- Reduced Cone Density Is Associated with Multiple Sclerosis. McIlwaine G, Csincsik L, Coey R, Wang L, Fitzgerald D, Moffat J, Dubis AM, McDonnell G, Hughes S, Peto T, Lengyel I. Ophthalmol Sci. 2023 Apr 13;3(3):100308.
- Investigating Biomarkers for USH2A Retinopathy Using Multimodal Retinal Imaging. Gill JS, Theofylaktopoulos V, Mitsios A, Houston S, Hagag AM, Dubis AM, Moosajee M. Int J Mol Sci. 2022 Apr 11;23(8):4198.
- Towards standardizing retinal optical coherence tomography angiography: a review. Sampson DM, Dubis AM, Chen FK, Zawadzki RJ, Sampson DD. Light Sci Appl. 2022 Mar 18;11(1):63.
- Clinically applicable deep learning-based decision aids for treatment of neovascular AMD. Gutfleisch M, Ester O, Aydin S, Quassowski M, Spital G, Lommatzsch A, Rothaus K, Dubis AM, Pauleikhoff D. Graefes Arch Clin Exp Ophthalmol. 2022 Jul;260(7):2217-2230.
- Clinical and pre-clinical therapeutic outcome metrics for USH2A-related disease. Dubis AM, Toms M, de Vrieze W, Tracey-White-Dhani, Mitsios A, Hayes M, Broekman A, Baxendale S, Utoomprurkporn N, Bamiou D, Bitner-Glindzicz, Webster AR, Van Wijk EV, Moosajee M. Hum Mol Genet, 2020 Jul 21, 29)11_1882-1899.
How Does Our Research Help Patients?
Since the beginning of his postgraduate work, Dr. Dubis has focused on translating research from in silico to bedside. He has sought to continually build personal and team development, tackling new challenging areas for himself and working in multidisciplinary teams to deliver successful outcomes through the commercialization of several devices and algorithms.
These experiences have taught him about the regulatory frameworks required for commercialization in the medical technology space and, given his international placements, exposed him to how these vary across global markets.
Some of his most recent technologies have been spun out into the founding of deepeye Medical GmbH, a digital health company focused on developing software as a medical device (SaMD) solutions for ophthalmology.
Additionally, Dr. Dubis sits on the board of five other companies in the digital health space, spanning mental health to patient wellness. He also broadly consults in image analysis and biopharmaceuticals and works with several international bodies on the evolution of policy and health technology regulation.
More About Dr. Dubis
Dr. Dubis started his career in the U.S., where he studied biology and chemistry and participated in his university’s health entrepreneurship courses.
Upon entering graduate school at the Medical College of Wisconsin, he drew on his engineering background to commercialize several analytical tools and methods for ophthalmology.
After moving to London in 2013, he has continued his enterprise activities, consulting for several start-ups and commercializing health analytics through patent filings and spin-out companies. He also developed two master’s-level courses aimed at enabling biomedical idea generators to take their ideas to market.
He joined the research faculty at the John A. Moran Eye Center in 2024 and continues teaching activities within the University of Utah Biomedical Engineering BioInnovate Program.
Contact Us
Adam Dubis, PhD