Reproduce “The Development and Performance of a Machine Learning Based Mobile Platform for Visually Determining the Etiology of Penile Pathology” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Lao-Tzu Allan-Blitz, Sithira Ambepitiya, Raghavendra Tirupathi, Jeffrey D. Klausner, Yudara Kularathne

Researchers developed a machine-learning-based mobile platform, HeHealth, to visually diagnose five penile pathologies: herpes eruption, syphilitic chancres, penile candidiasis, penile cancer, and genital warts. The platform achieved an F1-score of 0.944 on a validation dataset and has received over 37,000 unique global submissions.

Key insights

  • Combining semantic segmentation with classification and salience mapping enhances the model's ability to isolate diseased regions and provide interpretable predictions.
  • Layered image augmentation, including manual patterning and automated transformations, effectively addresses data scarcity and class imbalances in sensitive medical datasets.
  • Ubiquitous smartphone technology can be leveraged to deliver complex AI-driven diagnostic tools, potentially bridging significant gaps in healthcare access for stigmatized conditions.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.