A first explainable-AI-based workflow integrating forward-forward and backpropagation-trained networks of label-free multiphoton microscopy images to assess human biopsies of rare neuromuscular disease
Our workflow, using clinically routine biopsies and transparent diagnostic features, demonstrates the unique potential of forward-forward in providing novel, reliable, interpretable results from biomedical images. This paves the way for its dual use with backpropagation, shifting benchmarks by enabling the discovery of features potentially overlooked by backpropagation across biomedical datasets.
Authors
- Frauke Alves (ORCID: https://orcid.org/0000-0002-6258-5409)
- Riccardo Scodellaro (ORCID: https://orcid.org/0000-0001-8889-1124)
- Jana Zschüntzsch (ORCID: https://orcid.org/0000-0002-7062-5319)
- Anna Kathrin Hell (ORCID: https://orcid.org/0000-0002-6655-7130)
Institutions
- Universitätsmedizin Göttingen (DE)
- Max Planck Institute for Multidisciplinary Sciences (DE)
- University of Göttingen (DE)
Publication Details
- Journal
- Computer Methods and Programs in Biomedicine
- Published
- 2025-03-22
- DOI
- https://doi.org/10.1016/j.cmpb.2025.108733
- Citations
- 5
- Primary Topic
- Cell Image Analysis Techniques
- Type
- article
- Field-Weighted Citation Impact
- 8.17
Funders
- European Federation of Pharmaceutical Industries and Associations
- Innovative Health Initiative
- European Commission
- Innovative Medicines Initiative