Our aim is to improve the efficiency, accuracy and cost-effectiveness of the current workflow employed in
skin lesion diagnostics. We will do so by introducing AI as part of the diagnostic process, initially providing
clinicians with AI enhanced feedback and personalized learning materials. The generated dataset will be used
to develop and continuously improve an embedded neural network (ENN) for skin lesion classification, with
an unprecedented range and accuracy.
To accomplish this, we will develop DermLoop, an educational platform that couples images of skin lesions
taken by clinicians at initial evaluation with a final histopathological diagnosis and provides automated AI
augmented feedback, based on individual clinician data. Thus, DermLoop creates a powerful incentive
through direct personal feedback for procurement of high quality images. The generated dataset will be used
to develop and continuously improve an ENN for skin lesion classification.
DermLoop offers a paradigm shift in medical education and a bridge between clinical practice and AI powered
decisional support. Our solution will benefit patients and doctors alike whilst reducing societal costs associated
with delayed skin cancer diagnosis as well as unnecessary surveillance and removal of benign skin
lesions.