ƒ(Cell): Software for Reproducible Analysis of Optoretinograms.

Area of Science:

• Ophthalmology and Vision Science
• Biomedical Imaging
• Computational Biology

Background:

• Optoretinography (ORG) is a noninvasive functional imaging technique with increasing use.
• Standardized methodologies are crucial for repeatable ORG analyses, especially for clinical applications.
• Existing ORG data analysis workflows can vary, impacting consistency.

Purpose of the Study:

• To introduce ƒ(Cell), an open-source Python software library for reproducible and repeatable optoretinography data analysis.
• To provide tools for co-registration, analysis, and process auditing of ORG datasets.
• To identify key analysis parameters influencing the repeatability of ORG data.

Main Methods:

• Developed ƒ(Cell) as a Python library for ORG data analysis.
• Validated the software using normative ORG datasets and data from six external research groups.
• Performed variance decomposition analysis on ƒ(Cell) parameters across multiple datasets and individuals.

Main Results:

• ƒ(Cell) successfully processed diverse ORG datasets regardless of format or protocol.
• A full factorial analysis identified segmentation radius, normalization, and relativization method as significant parameters (P < 0.001).
• Mean subtraction relativization, cell-specific segmentation radius, and score-based normalization yielded the lowest coefficient of variance.

Conclusions:

• ƒ(Cell) enables consistent, reproducible, and auditable analyses of intensity-based ORG (iORG) data.
• The software facilitates the standardization of ORG analysis for improved reliability.
• This work supports the use of ORG in clinical trials through enhanced data analysis consistency.