review · Diagnostics
This systematic review evaluates the diagnostic accuracy of computer-aided systems for skin lesion diagnosis, a growing area of research. It synthesises and assesses evidence from 102 articles published over the last five years, sourced from ScienceDirect, IEEE, and SpringerLink databases. The review included 53 articles on traditional machine learning and 49 on deep learning methods. The studies were compared based on their contributions, the methodologies employed, and the results achieved. A key outcome was the identification of significant challenges in evaluating these systems, such as the use of small datasets, ad hoc image selection practices, and the presence of racial bias in the data.
Accurate and early diagnosis of skin lesions is crucial for effective treatment. This review highlights the current state and challenges of using artificial intelligence for this purpose, informing researchers and developers about critical issues like data bias and dataset limitations that need addressing for reliable diagnostic tools.
This research provides a foundational overview for developers of computer-aided diagnostic tools for skin lesions. By identifying challenges such as small datasets and racial bias, it guides future research and development towards more robust and equitable systems. Addressing these issues is essential for creating reliable, commercially viable diagnostic software that can be adopted by healthcare providers and clinics, though this review itself is not a product.
AI-generated from the published abstract. Always read the original work before citing.
Computer-aided systems for skin lesion diagnosis is a growing area of research. Recently, researchers have shown an increasing interest in developing computer-aided diagnosis systems. This paper aims to review, synthesize and evaluate the quality of evidence for the diagnostic accuracy of computer-aided systems. This study discusses the papers published in the last five years in ScienceDirect, IEEE, and SpringerLink databases. It includes 53 articles using traditional machine learning methods and 49 articles using deep learning methods. The studies are compared based on their contributions, the methods used and the achieved results. The work identified the main challenges of evaluating skin lesion segmentation and classification methods such as small datasets, ad hoc image selection and racial bias.
This page summarises published work. The authoritative version sits with the publisher.
DOI: 10.3390/diagnostics11081390
Is something wrong with this record? Report it or request removal.
Discussion
Have you built on this work, tried to replicate it, or seen it applied in practice? Share what you know. Verified researchers and MARATTO™ domain experts can open a discussion, and any member can reply. Contributions are reviewed before they appear.
No discussion yet. Open the first thread.
New to MARATTO™? Create a free account.
