AI-Powered App for Managing Atopic Dermatitis; AI-Based App for Classifying Itchy Dermatoses; AI-Enabled Ultrasound Diagnostics App

Context & Background

Skin disorders and ultrasound scanning are highly common in primary care, but rural clinics lack trained dermatologists and radiologists to make quick diagnostic decisions.

Problems to be Addressed

Dermatophytosis is often misdiagnosed as other itchy dermatoses. Additionally, real-time ultrasound analysis requires high training, causing long diagnostic queues in rural centers.

Aims and Objectives

1. Develop an AI app to monitor atopic dermatitis severity.
2. Build image-classification models to differentiate dermatophytosis from eczema.
3. Design real-time ultrasound guidance and diagnostics support tools.

Methodology

The project uses mobile camera images and develops light deep-learning models (MobileNet) optimized to run on edge mobile devices. For ultrasound data, temporal models track anatomical markers on video streams to assist clinicians in localization.

Expected Outcomes

A suite of three mobile medical apps, image classification datasets for Indian skin conditions, and clinical testing reviews.