Deep Convolution Neural Networks (CNN) enable advanced methods to predict the skin cancer classes through the automatic analysis of digital dermoscopic images. However, in despite of the good performance provided by the Deep Learning Models, the difficult interpretability of the corresponding results does often not facilitate a massive adoption in clinical practice by dermatologists. To overcome the problem, this paper discloses the adoption of different CNNs as Semantic Segmentation technique able to detect and measure skin lesions atypical criteria according to the well-known diagnostic method 7-Point Check List. The experimental results show that the Artificial Intelligence-based model can suitably manage the classification uncertainty of the CNNs and finally distinguish melanomas from benignant nevi. Diagnostic performance is very promising towards a decision-supporting system to be used by a dermatologist with low experience during clinical practice.
Automatic Classification of Skin Lesions based on Semantic Segmentation
Sommella P.;Ferro M.;Di Leo G.;Gallo V.;Carratu' M.
2025
Abstract
Deep Convolution Neural Networks (CNN) enable advanced methods to predict the skin cancer classes through the automatic analysis of digital dermoscopic images. However, in despite of the good performance provided by the Deep Learning Models, the difficult interpretability of the corresponding results does often not facilitate a massive adoption in clinical practice by dermatologists. To overcome the problem, this paper discloses the adoption of different CNNs as Semantic Segmentation technique able to detect and measure skin lesions atypical criteria according to the well-known diagnostic method 7-Point Check List. The experimental results show that the Artificial Intelligence-based model can suitably manage the classification uncertainty of the CNNs and finally distinguish melanomas from benignant nevi. Diagnostic performance is very promising towards a decision-supporting system to be used by a dermatologist with low experience during clinical practice.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.