Improving Automatic Melanoma Diagnosis Using Deep Learning-Based Segmentation of Irregular Networks

Abstract

Deep learning has achieved significant success in malignant melanoma diagnosis. These diagnostic models are undergoing a transition into clinical use. However, with melanoma diagnostic accuracy in the range of ninety percent, a significant minority of melanomas are missed by deep learning. Many of the melanomas missed have irregular pigment networks visible using dermoscopy. This research presents an annotated irregular network database and develops a classification pipeline that fuses deep learning image-level results with conventional hand-crafted features from irregular pigment networks. We identified and annotated 487 unique dermoscopic melanoma lesions from images in the ISIC 2019 dermoscopic dataset to create a ground-truth irregular pigment network dataset. We trained multiple transfer learned segmentation models to detect irregular networks in this training set. A separate, mutually exclusive subset of the International Skin Imaging Collaboration (ISIC) 2019 dataset with 500 melanomas and 500 benign lesions was used for training and testing deep learning models for the binary classification of melanoma versus benign. The best segmentation model, U-Net++, generated irregular network masks on the 1000-image dataset. Other classical color, texture, and shape features were calculated for the irregular network areas. We achieved an increase in the recall of melanoma versus benign of 11% and in accuracy of 2% over DL-only models using conventional classifiers in a sequential pipeline based on the cascade generalization framework, with the highest increase in recall accompanying the use of the random forest algorithm. The proposed approach facilitates leveraging the strengths of both deep learning and conventional image processing techniques to improve the accuracy of melanoma diagnosis. Further research combining deep learning with conventional image processing on automatically detected dermoscopic features is warranted.

A novel deep learning framework for accurate melanoma diagnosis integrating imaging and genomic data for improved patient outcomes

“…In addition, recent research has pioneered deep learning techniques in conjunction with conventional image processing methods to detect significant abnormalities in pigment patterns. As a result, it has become possible to diagnose melanoma with higher accuracy rates (Nambisan et al., 2023) 17 …”

“…As a result, it has become possible to diagnose melanoma with higher accuracy rates (Nambisan et al., 2023). 17 …”

Leveraging Machine Learning for Accurate Detection and Diagnosis of Melanoma and Nevi: An Interdisciplinary Study in Dermatology

“…Emerging technologies like machine learning and deep learning algorithms offer possibilities for improving the diagnosis of melanoma and nevi. By employing artificial intelligence (AI) techniques to analyze imaging data, these algorithms can learn features and patterns that enable accurate differentiation between benign and malignant lesions [ 12 ]. The implementation of AI-based approaches holds the potential to enhance accuracy and improve decision-making for melanoma management.…”

Improved Tuna Swarm-based U-EfficientNet: Skin Lesion Image Segmentation by Improved Tuna Swarm Optimization

“…The separation of an image into useful areas is known as segmentation. In instance, segmentation classifies the essential area with the proper classes and labels [16]. Here, the approach utilized for skin lesions is binary task, i.e., differentiating the tumor from the skin surrounding.…”