pV-Diagrams
Photoreceptors and Visual Pathways
You might also read
Articles linked to this work by shared authors, journal, and citation graph.
Updated: Aug 11, 2025

Assessing Binocular Central Visual Field and Binocular Eye Movements in a Dichoptic Viewing Condition
Published on: July 21, 2020
Mohammad Eslami1, Saber Kazeminasab1, Vishal Sharma1
1Harvard Ophthalmology AI Lab, Schepens Eye Research Institute of Massachusetts Eye and Ear, Harvard Medical School, Boston, MA, USA.
This article introduces a new open-source software package that allows researchers to perform advanced visual field analysis directly within the Python programming environment, bridging a significant gap in available ophthalmic diagnostic tools.
Area of Science:
Background:
Visual field assessment remains a cornerstone for monitoring vision loss progression in clinical practice. Recent advancements in machine learning have transformed how researchers interpret complex ophthalmic data sets. Most modern artificial intelligence frameworks rely heavily on the Python programming ecosystem for development. However, existing statistical tools for analyzing visual field data are primarily restricted to the R language environment. This creates a significant barrier for developers aiming to integrate advanced diagnostic algorithms into Python-based clinical applications. No prior work had resolved the incompatibility between these two dominant computational platforms for vision research. That uncertainty drove the creation of a specialized bridge to unify these analytical capabilities. This paper addresses the technical divide by providing a dedicated library for visual field processing.
Purpose Of The Study:
The primary aim of this study is to introduce a new Python package designed to support visual field analysis. This project seeks to address the current limitation where most diagnostic algorithms reside in the R language. The authors intend to provide a bridge for researchers who prefer using Python for machine learning applications. By translating established statistical functions, the team hopes to facilitate more efficient ophthalmic research. The study addresses the need for accessible tools that can handle data normalization and progression analysis. The researchers aim to simplify the integration of clinical diagnostic tools into modern computational workflows. This work motivates the transition toward more versatile and open-source ophthalmic software solutions. The project ultimately strives to empower developers to create advanced clinical applications for vision care.
Main Methods:
The research team conducted a systematic comparison of existing R libraries to identify core functional requirements. They analyzed the vfprogression and visualFields packages to determine necessary overlaps and distinct analytical features. The developers then translated these identified functions into the Python programming language. They employed the rpy2 wrapper library to maintain operational continuity with the original R-based statistical methods. The team established a roadmap for future versions to ensure long-term software maintenance and independence. They created several Jupyter notebooks to document and illustrate the package capabilities for potential users. The software was packaged for distribution via the Python Package Index to ensure ease of installation. Finally, the developers made the entire source code available through a public GitHub repository for community review.
Main Results:
The primary outcome is the successful development of a Python-based library for comprehensive visual field analysis. The package provides a suite of tools for data normalization, deviation analysis, and plotting. It includes specific functions for calculating and predicting vision loss progression. The researchers demonstrated that these tools effectively replicate the functionality of established R-based diagnostic packages. The software is currently available for installation via the Python Package Index for immediate research use. The team provided interactive notebooks that describe the application of these tools in clinical settings. The current version relies on the rpy2 wrapper to bridge the gap between different programming ecosystems. The developers established a clear milestone for the third version to function independently of external R libraries.
Conclusions:
The authors successfully created a functional software package to facilitate visual field research within the Python ecosystem. This tool enables investigators to utilize existing statistical methods while leveraging modern machine learning frameworks. Future development milestones aim to eliminate reliance on external R libraries entirely. The current version provides essential functions for data normalization, visualization, and progression analysis. Researchers can now access these capabilities through standard open-source repositories for immediate implementation. This software supports the rapid development of clinical applications for detecting vision loss. The team provides comprehensive documentation through interactive notebooks to guide new users. This work represents a shift toward more accessible and integrated ophthalmic diagnostic programming.
The package enables visual field analysis by translating established R-based statistical functions into Python. It utilizes the rpy2 wrapper library to maintain compatibility with existing diagnostic algorithms while facilitating the development of new machine learning models for predicting vision loss.
The researchers utilize the rpy2 wrapper library to bridge the gap between R and Python environments. This tool allows the software to execute functions from established packages like vfprogression and visualFields, ensuring that existing statistical methodologies remain accessible to Python-based developers.
The authors state that the rpy2 wrapper is necessary to maintain functionality during the initial development phase. This technical requirement ensures that the package can immediately perform complex statistical tasks while the team works toward creating a fully independent, native Python version in future releases.
The package uses Jupyter notebooks as a key component to demonstrate its capabilities. These files provide practical examples for data presentation, normalization, deviation analysis, and plotting, serving as a tutorial for researchers to implement the software in their own clinical studies.
The software facilitates progression analysis, which is the measurement of how vision loss changes over time. By providing tools for this specific phenomenon, the package allows clinicians and researchers to better predict the trajectory of ocular diseases using advanced statistical and machine learning techniques.
The researchers propose that this package will accelerate the creation of clinical applications. By providing a Python-native environment, they suggest that investigators can more efficiently develop and deploy cutting-edge artificial intelligence models for real-world diagnostic use.