External site

By following this link, you are leaving Medically and entering a website that is not owned or controlled by Roche. Roche does not take any responsibility for access to or use of this website, nor for any content therein.

Leave site

Geographic Atrophy

What's new

Sign up or login to unlock the full suite of MEDICALLY features

Personalised Healthcare
May 6 / Roche and Genentech
Predicting Geographic Atrophy Growth Rate With Clinical and Derived Imaging Features
Geographic atrophy progression prediction models have many use cases in clinical trials, including covariate adjustment. In the future, these models may also be useful for patient counseling and treatment decisions in clinical practice. End-to-end deep learning (DL) models with fundus autofluorescence (FAF) baseline images have outperformed feature-based models trained on standard clinical and imaging features, including those with run-in data. However, the imaging features that drive the success of these DL models remain unknown. This poster described the development of imaging-feature-based models that aim to achieve similar performance to FAF-DL models to help derive prognostic imaging biomarkers and create more interpretable models.

Sign up or login to unlock the full suite of MEDICALLY features

Personalised Healthcare
May 6 / Roche and Genentech
Generalizability of foundation models for geographic atrophy (GA) lesion segmentation in fundus autofluorescence (FAF)
High-capacity, transformer-based, deep-learning models trained in a self-supervised manner on large datasets are referred to as foundation models (FMs). These models may offer generalizability to unseen domains/tasks suggesting the potential to accelerate algorithm development, particularly with limited data. This poster explores the generalizability of vision foundation models (FMs), DINOv21 and SAM2, to an unseen ophthalmology task, segmentation of geographic atrophy (GA) lesions in fundus autofluorescence (FAF), and characterizes performance as a function of dataset size.

Sign up or login to unlock the full suite of MEDICALLY features

Oct 9 / Roche and Genentech
Visual Loss in Geographic Atrophy: A Comprehensive Exploratory Analysis of Function and Morphology From the Lampalizumab Trial Programme
Geographic atrophy (GA) is an advanced form of age-related macular degeneration and a major cause of vision impairment worldwide. There is large variability in GA growth rate between individuals and is highly influenced by morphological factors. Interventional clinical trials have shown reductions in GA growth as measured on fundus autofluorescence images with no corresponding reduction in any other clinical measure of visual function such as Best Corrected Visual Acuity (BCVA). However, there is major need for analyses of change in function by morphological characteristics and by growth rate. The aim of this study is to explore how baseline factors and GA expansion rates influence the rate of BCVA loss.

Sign up or login to unlock the full suite of MEDICALLY features

Sep 17 / Roche and Genentech
Educational video on BCVA & visual function
This video discusses visual function assessments for patients with geography atrophy, describing the limitations of best-corrected visual acuity.

Sign up or login to unlock the full suite of MEDICALLY features

Sep 13 / Roche and Genentech
Through the eyes of a person with geographic atrophy - An advanced form of age-related macular degeneration
This video helps to better understand what it feels like to have Geographic Atrophy (GA) by seeing from the eyes of an affected patient. The visual impairment defect simulated in this video is based on microperimatry data from a patient with GA.
Upcoming congresses
Access to Roche and Genentech’s latest medical information
View related congresses

Ask a question or share feedback