The power of deep learning for intelligent tumor classification systems: A review. (March 2023)

• Record Type: Journal Article
• Authors: Chandni, ; Sachdeva, Monika; Kushwaha, Alok Kumar Singh
• Is Part Of: Computers & electrical engineering. Volume 106(2023)
• Publication Date: 2023-03
• Subjects: Tumor classification -- Machine learning -- Deep learning -- Convolution neural network -- Malignant tumor; Computer engineering -- Periodicals; Electrical engineering -- Periodicals; Electrical engineering -- Data processing -- Periodicals; Ordinateurs -- Conception et construction -- Périodiques; Électrotechnique -- Périodiques; Électrotechnique -- Informatique -- Périodiques; Computer engineering; Electrical engineering; Electrical engineering -- Data processing; Periodicals; Electronic journals; 621.302854
• DOI: 10.1016/j.compeleceng.2023.108586
• Languages: English
• ISSNs: 0045-7906

Abstract

Highlights: Comprehensive review of pioneer techniques for identification and classification of tumor malignancy. It investigates more than 150 recent articles for tumor or cancer classification, covering 5 sites of cancer: breast, brain, lungs, skin and cervical cancer. It establishes a general framework used in automated cancer classification task. It investigates articles on different dimensions like: dataset or input modality, feature extraction technique, segmentation approach, classification technique, performance. Abstract: A tumor is a life-threatening disease that refers to the abnormal growth of cells in any part of the human body. Early detection of this abnormality not only helps with appropriate treatment but also increases life expectancy. In this era of technology, Machine Learning (ML) and Deep Learning (DL) offer reliable and effective techniques for creating intelligent data-driven systems. It thus also serves as a prominent aid in the intelligent diagnosis of various fatal diseases, like tumors or cancer. This study conducts in-depth research on recent deep learning based studies on the classification of tumors at various sites of the human body. Research questions pertinent to the tumor classification are framed, and the literature is explored to answer the questions. After providing insights into the concepts of ML and DL, the study highlights the strengths and limitations of the existing studies. It also provides a qualitative and quantitative comparison of pioneer studies in terms of approach, methods, datasets, and performance metrics. Finally, this study summarizes the challenges confronted by the researchers and makes recommendations for the future application of DL techniques for tumor classification. Graphical abstract: Graphical Abstract Image, graphical abstract