Artificial intelligence-based image classification methods for diagnosis of skin cancer: Challenges and opportunities

Abstract

Recently, there has been great interest in developing Artificial Intelligence (AI) enabled computer-aided diagnostics solutions for the diagnosis of skin cancer. With the increasing incidence of skin cancers, low awareness among a growing population, and a lack of adequate clinical expertise and services, there is an immediate need for AI systems to assist clinicians in this domain. A large number of skin lesion datasets are available publicly, and researchers have developed AI solutions, particularly deep learning algorithms, to distinguish malignant skin lesions from benign lesions in different image modalities such as dermoscopic, clinical, and histopathology images. Despite the various claims of AI systems achieving higher accuracy than dermatologists in the classification of different skin lesions, these AI systems are still in the very early stages of clinical application in terms of being ready to aid clinicians in the diagnosis of skin cancers. In this review, we discuss advancements in the digital image-based AI solutions for the diagnosis of skin cancer, along with some challenges and future opportunities to improve these AI systems to support dermatologists and enhance their ability to diagnose skin cancer.

Keywords: Artificial intelligence; Computer-aided diagnostics; Deep learning; Dermatologists; Digital dermatology; Skin cancer.

Fig. 1.

Illustration of different types of dermoscopic skin lesions where (a) Nevi (b) Melanoma (c) Basal Cell Carcinoma (d) Actinic Keratosis (e) Benign Keratosis (f) Dermatofibroma (g) Vascular Lesion (h) Squamous Cell Carcinoma [22]

Fig. 2.

Illustration of different types of clinical skin lesions where (a) Benign Keratosis (b) Melanoma (c) BCC (d) SCC [19]

Fig. 3.

Illustration of different types of histopathology images where (a) Nevi (b) Melanoma (c) Basal Cell Carcinoma (d) Squamous Cell Carcinoma [19]

Fig. 4.

Ensemble CNN approach for skin lesion classification

Fig. 5.

Illustration of intra-class dissimilarities in BCC (a) Nodular BCC (b) Superficial BCC (c) Morphoeic BCC (d) Basosquamous BCC [32]

Fig. 6.

Examples of pre-processing with a Shades of Gray algorithm. (a) Original images with different background colors; and (b) Pre-processed images with more consistent background colors.