Jadwal Sholat

Memuat jadwal sholatโ€ฆ

Ilmu Komputer & AI editorial

Open AccessOA2025

Diagnosis and Prognosis of Melanoma from Dermoscopy Images Using Machine Learning and Deep Learning: A Systematic Literature Review

This systematic review analyzes 34 studies on machine learning and deep learning for melanoma diagnosis and prognosis using dermoscopy images, finding that architectures like DenseNet and DCNN achieve over 95% accuracy on benchmark datasets. Key challenges remain in data diversity, model interpretability, and computational requirements for clinical adoption.
Hoda Naseri; A. Safaeiยท BMC Cancerยท 2025ยท DOI 10.1186/s12885-024-13423-y

The core problem

Melanoma is a highly aggressive form of skin cancer, and early, accurate diagnosis is critical for improving patient outcomes. Dermoscopy, a non-invasive imaging technique, enhances melanoma detection by visualizing subsurface skin structures, but its interpretation remains subjective and varies among clinicians. Recent advances in machine learning (ML) and deep learning (DL) offer a promising solution by automating the analysis of dermoscopy images to increase diagnostic precision. This systematic review aims to consolidate recent evidence on ML and DL applications for melanoma diagnosis and prognosis using dermoscopy images, covering studies published between 2016 and 2024. The review addresses the need for a comprehensive overview of model architectures, datasets, methodologies, and evaluation metrics, and identifies gaps for future research.

Innovation

A thorough literature search was conducted across multiple databases, yielding 34 studies that met the inclusion criteria. The review focused on studies applying ML and DL techniques to dermoscopy images for melanoma diagnosis or prognosis. Data extracted from each study included model architectures (e.g., DenseNet, ResNet, DCNN), dataset characteristics (e.g., HAM10000, ISIC), training methodologies, and performance metrics such as accuracy, sensitivity, and specificity. The quality of studies was assessed, and findings were synthesized to identify trends, strengths, limitations, and future directions. The review protocol followed systematic review guidelines to ensure reproducibility and comprehensiveness.
Introduction
Melanoma is a highly aggressive form of skin cancer, and early, accurate diagnosis is critical for improving patient outcomes. Dermoscopy, a non-invasive imaging technique, enhances melanoma detection by visualizing subsurface skin structures, but its interpretation remains subjective and varies among clinicians. Recent advances in machine learning (ML) and deep learning (DL) offer a promising solution by automating the analysis of dermoscopy images to increase diagnostic precision. This systematic review aims to consolidate recent evidence on ML and DL applications for melanoma diagnosis and prognosis using dermoscopy images, covering studies published between 2016 and 2024. The review addresses the need for a comprehensive overview of model architectures, datasets, methodologies, and evaluation metrics, and identifies gaps for future research.

Why it matters

The review highlights the transformative potential of ML and DL in melanoma diagnosis, yet several challenges must be addressed before widespread clinical adoption. Data diversity is a major concern; most studies use a limited number of public datasets that may not represent the global population, leading to potential biases. Model interpretability is another critical issue, as deep learning models often operate as 'black boxes,' making it difficult for clinicians to trust their predictions. Computational resource requirements also pose a barrier, especially for deployment in resource-constrained settings. Future research should focus on creating accessible, large, and diverse datasets, developing interpretable models (e.g., through attention mechanisms or explainable AI), and optimizing computational efficiency. Collaborative efforts between clinicians, data scientists, and policymakers are essential to translate these promising technologies into routine clinical practice. Ultimately, ML and DL models could play a central role in advancing melanoma diagnosis and improving patient care.

Who should read this

Pembaca Research Digest dan praktisi rumpun ini.

Opening member contentโ€ฆ