Skin Lesion Detection Algorithms in Whole Body Images

Michał H Strzelecki,Marcin Kociołek,Tomasz Urbańczyk,Maria Strąkowska,Dorota Wielowieyska-Szybińska,Michał Kozłowski

doi:10.3390/s21196639

Michał H Strzelecki, Marcin Kociołek + Show 4 more

Open Access

https://doi.org/10.3390/s21196639

Copy DOI

Abstract

Melanoma is one of the most lethal and rapidly growing cancers, causing many deaths each year. This cancer can be treated effectively if it is detected quickly. For this reason, many algorithms and systems have been developed to support automatic or semiautomatic detection of neoplastic skin lesions based on the analysis of optical images of individual moles. Recently, full-body systems have gained attention because they enable the analysis of the patient’s entire body based on a set of photos. This paper presents a prototype of such a system, focusing mainly on assessing the effectiveness of algorithms developed for the detection and segmentation of lesions. Three detection algorithms (and their fusion) were analyzed, one implementing deep learning methods and two classic approaches, using local brightness distribution and a correlation method. For fusion of algorithms, detection sensitivity = 0.95 and precision = 0.94 were obtained. Moreover, the values of the selected geometric parameters of segmented lesions were calculated and compared for all algorithms. The obtained results showed a high accuracy of the evaluated parameters (error of area estimation <10%), especially for lesions with dimensions greater than 3 mm, which are the most suspected of being neoplastic lesions.

Highlights

Melanoma is a malignant skin tumor and is considered the deadliest of skin cancers
Such an analysis, carried out based on a photo taken with a smartphone, allows obtaining information about the potential risk of detecting a neoplastic lesion, which must be verified by a dermatologist
Mobile applications developed for skin cancer detection recently gained popularity; the thorough analysis provided in [1] reports 43 such applications

Summary

Introduction

Melanoma is a malignant skin tumor and is considered the deadliest of skin cancers The reason for this is its rapid growth and ease of metastasis. Melanoma is almost 100% curable if detected quickly For this reason, it is very important to increase preventive measures aimed at promoting behaviors that prevent the formation of this cancer and awareness of the importance of screening. It is very important to popularize mobile applications that allow independent tests to be conducted of a suspected naevi detected on the body. Such an analysis, carried out based on a photo taken with a smartphone, allows obtaining information about the potential risk of detecting a neoplastic lesion, which must be verified by a dermatologist. The first implements conditional generative adversarial neural network segmentation and SVM for mole classification, resulting in sensitivity in the range of 57–72% and specificity of

Objectives

Methods

Results

Conclusion