Explainable AI Model for Skin Lesion Image Quality Assessment

Overview

Dermatology services face growing pressures from long waiting times, increased demand, staff shortages, and rising costs. Teledermatology offers a way to ease this burden by enabling remote assessments, but only if the images are of sufficient quality. In practice, many patient-captured images are poorly lit, out of focus, or incorrectly framed, making them unsuitable for accurate analysis.
This PhD project will design explainable AI (XAI) models that not only assess skin lesion image quality but also provide transparent, actionable feedback.

Why it matters:

•  Improves Access: Supports safe and effective teledermatology, reducing waiting times.
• Empowers Patients: Offers clear, real-time guidance to capture clinically usable images.
• Builds Trust: Ensures AI outputs are interpretable, giving clinicians confidence in automated assessments.

Key objectives:

1.  Develop deep learning models for automated skin lesion image quality assessment.
2. Integrate explainable AI techniques to deliver reliable, interpretable feedback.
3. Collaborate with dermatologists to validate models in real-world teledermatology practice.

Why Choose a PhD at Dundee?

•  Clinical Collaboration: Work directly with dermatologists and healthcare researchers at Ninewells Hospital.
• Cutting-Edge Research: Combine computer vision, explainable AI, and clinical validation.
• Societal Impact: Contribute to improving cancer care pathways and reducing diagnostic delays.
• Career Development: Gain expertise in both technical AI innovation and clinically applied research.

Diversity statement

Our research community thrives on the diversity of students and staff which helps to make the University of Dundee a UK university of choice for postgraduate research. We welcome applications from all talented individuals and are committed to widening access to those who have the ability and potential to benefit from higher education.