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

Published on: March 13, 2024

Impact Score: 7.8

Arxiv code: Arxiv:2403.08417

Summary

• What is new: The development of a machine-learning model for classifying five penile diseases with a high accuracy rate, leveraging a novel clinical image dataset.
• Why this is important: Disparities in access to sexual health services, particularly in low-cost, user-guided visual diagnostics.
• What the research proposes: A machine-learning algorithm leveraging U-net architecture and Inception-ResNet v2 for semantic pixel segmentation and classification of penile diseases.
• Results: The model demonstrated an overall accuracy of 0.944 for correctly classifying diseased images across a validation dataset of 239 images.

Technical Details

Technological frameworks used: U-net, Inception-ResNet version 2, GradCAM++

Models used: Semantic pixel segmentation, pixel classification

Data used: Original and augmented images for five penile diseases

Potential Impact

Healthcare providers, telehealth platforms, app developers focusing on sexual health diagnostics

Want to implement this idea in a business?

We have generated a startup concept here: HealthLens.