[Paper Review] An Overview of Melanoma Detection in Dermoscopy Images Using Image Processing and Machine Learning

Nabin K. Mishra, M. Emre Celebi|arXiv (Cornell University)|Jan 28, 2016

Cutaneous Melanoma Detection and Management107 citations

TL;DR

A survey of automatic melanoma detection in dermoscopy images, outlining lesion segmentation, feature extraction, and classification with machine learning, plus challenges toward clinical adoption.

ABSTRACT

The incidence of malignant melanoma continues to increase worldwide. This cancer can strike at any age; it is one of the leading causes of loss of life in young persons. Since this cancer is visible on the skin, it is potentially detectable at a very early stage when it is curable. New developments have converged to make fully automatic early melanoma detection a real possibility. First, the advent of dermoscopy has enabled a dramatic boost in clinical diagnostic ability to the point that melanoma can be detected in the clinic at the very earliest stages. The global adoption of this technology has allowed accumulation of large collections of dermoscopy images of melanomas and benign lesions validated by histopathology. The development of advanced technologies in the areas of image processing and machine learning have given us the ability to allow distinction of malignant melanoma from the many benign mimics that require no biopsy. These new technologies should allow not only earlier detection of melanoma, but also reduction of the large number of needless and costly biopsy procedures. Although some of the new systems reported for these technologies have shown promise in preliminary trials, widespread implementation must await further technical progress in accuracy and reproducibility. In this paper, we provide an overview of computerized detection of melanoma in dermoscopy images. First, we discuss the various aspects of lesion segmentation. Then, we provide a brief overview of clinical feature segmentation. Finally, we discuss the classification stage where machine learning algorithms are applied to the attributes generated from the segmented features to predict the existence of melanoma.

Motivation & Objective

• Motivate the use of dermoscopy for early melanoma detection and discuss its impact on clinical diagnosis.
• Summarize image processing steps for analyzing dermoscopy images, especially lesion segmentation.
• Review how clinical features are segmented and converted into machine-learnable attributes.
• Discuss machine learning classifiers applied to segmented features for melanoma prediction.
• Highlight current limitations and the need for accuracy, reproducibility, and broader implementation.

Proposed method

• Describe the role of dermoscopy in enabling early melanoma detection.
• Outline lesion segmentation approaches in dermoscopy images.
• Provide an overview of clinical feature segmentation for dermoscopy.
• Discuss the pipeline from segmented features to machine learning-based classification.
• Review machine learning algorithms used to predict melanoma from extracted features.
• Address challenges related to accuracy, reproducibility, and clinical adoption.

Experimental results

Research questions

• RQ1 What are the key segmentation methods for dermoscopy lesions and their effectiveness?
• RQ2 What clinical features are extracted from dermoscopy images and how are they transformed into attributes for ML?
• RQ3 Which machine learning classifiers are suitable for melanoma prediction from dermoscopy-derived features?
• RQ4 What barriers exist to widespread clinical adoption regarding accuracy and reproducibility?
• RQ5 How can automatic melanoma detection in dermoscopy images be further improved to reduce unnecessary biopsies?

Key findings

• Dermoscopy has driven improvements in early melanoma detection and diagnostic ability.
• Advances in image processing enable automatic distinction between melanoma and benign mimics.
• Current systems show promise in preliminary trials but require improvements in accuracy and reproducibility for widespread use.
• Development of fully automatic detection is feasible but not yet ready for universal clinical implementation.
• The pipeline from segmentation to classification is central to melanoma prediction in dermoscopy images.
• There is a need to reduce unnecessary and costly biopsies through reliable automation.