EUVIMEDEUVIMED entdecken
Open Access

Topo-CNN: Retinal Image Analysis with Topological Deep Learning

Faisal Ahmed, Mohammad Alfrad Nobel Bhuiyan, Baris Coskunuzer

Journal of Imaging Informatics in Medicine · 2025

DOI: 10.1007/s10278-025-01575-7 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Imaging Informatics in Medicine, 2025. Verfasst von Faisal Ahmed und 2 weiteren Autor:innen. Beteiligte Einrichtung: Department of Data Science and Mathematics, Embry-Riddle Aeronautical University, 3700 Willow Creek Rd, 86301, Prescott, AZ, USA. Hauptthemen laut MEDLINE: Deep Learning, Retina, Retinal Diseases, Image Interpretation, Computer-Assisted. Die Arbeit wurde bisher 7-mal zitiert (Crossref) und nennt 78 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar. Gefördert von: NIGMS NIH HHS, NIH HHS, NIH.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The analysis of fundus images is vital for early detection of retinal diseases such as diabetic retinopathy (DR), glaucoma, and age-related macular degeneration (AMD), but traditional methods are resource-intensive. We propose an automated and interpretable diagnostic framework that leverages novel feature representations to improve performance. Our main contribution is a topological feature extraction technique based on Topological Data Analysis (TDA), which captures geometric and structural patterns in fundus images. These features are computationally efficient and interpretable. We integrate them with pretrained CNN features (e.g., ResNet-50) into a hybrid deep model, Topo-CNN, combining global image context with topological structure. We evaluate Topo-CNN on three benchmarks: APTOS (binary and five-class DR), ORIGA (Glaucoma), and IChallenge-AMD. Our model achieves 98.7% accuracy/98.9 AUC on binary DR, 95.5 AUC on five-class DR, 93.8% accuracy/93.6 AUC on AMD, and 82.3% accuracy/95.8 specificity on glaucoma. Ablation studies confirm the added value of topological features, and our Topo-CNN consistently outperforms existing methods across tasks.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Deep LearningRetinaRetinal DiseasesImage Interpretation, Computer-AssistedHumansDiabetic RetinopathyFundus OculiMacular Degeneration

Stichwörter der Autor:innen

Deep learningOphthalmologyRetinal disease diagnosisTopological data analysis

Autor:innen und Institutionen

  • Department of Data Science and Mathematics, Embry-Riddle Aeronautical University, 3700 Willow Creek Rd, 86301, Prescott, AZ, USA.
  • Department of Medicine, Louisiana State University Health Sciences Center at Shreveport, 1501 Kings Highway, 71103, Shreveport, LA, USA. nobel.bhuiyan@lsuhs.edu.
  • Department of Mathematical Sciences, The University of Texas at Dallas, 800 W Campbell Rd, 75080, Richardson, TX, USA. coskunuz@utdallas.edu.

Förderung

  • NIGMS NIH HHS
  • NIH HHS
  • NIH

Zitieren

APA

Ahmed, F., Bhuiyan, M. A. N., & Coskunuzer, B. (2025). Topo-CNN: Retinal Image Analysis with Topological Deep Learning. Journal of Imaging Informatics in Medicine. https://doi.org/10.1007/s10278-025-01575-7

Vancouver

Ahmed F, Bhuiyan MAN, Coskunuzer B. Topo-CNN: Retinal Image Analysis with Topological Deep Learning. Journal of Imaging Informatics in Medicine. 2025. doi:10.1007/s10278-025-01575-7

BibTeX
@article{Ahmed2025,
  author = {Ahmed, Faisal and Bhuiyan, Mohammad Alfrad Nobel and Coskunuzer, Baris},
  title = {Topo-CNN: Retinal Image Analysis with Topological Deep Learning},
  journal = {Journal of Imaging Informatics in Medicine},
  year = {2025},
  doi = {10.1007/s10278-025-01575-7}
}
Kostenlos starten →