A Review on Automated Cancer Detection in Medical Images using Machine Learning and Deep Learning based Computational Techniques: Challenges and Opportunities

Abstract

Cancer is one of the most deadly diseases diagnosed among the population across the globe so far. The number of cases is increasing at a high pace each year that subsequently leads to the advancement in different diagnosis tools and technologies to handle this pandemic. Significant increase in the mortality rate worldwide leads tremendous scope to device and implement latest computer aided diagnostic systems for its early detection. The one among such techniques is machine learning coupled with medical imaging modalities. This combination has proven to be efficient in diagnosing various medical conditions in cancer diagnosis. Current study presents a review of different machine learning techniques applied on imaging modalities for cancer diagnosis from 2008 to 2019. This study focuses on diagnosis of five most prevalent and deadly cancers i.e., cervical cancer, oral cancer, breast cancer, brain cancer and skin cancer. Extensive and exhaustive review was carried out after going through different research papers, research articles and book chapters published by reputed international and national publishers such as Springer Link, Science Direct, IEEE Xplore Digital library and PubMed. A number of conference proceedings have also been included subject to the fulfilling of our quality evaluation criteria. This review article provides a comprehensive overview of machine learning approaches using image modalities for cancer detection and diagnosis with main focus on challenges being faced during their research. Majority of the challenges are identified based on the use of potential machine learning based approaches, image modalities, features and evaluation metrics. This review not only identified challenges but also ear mark and present the new research opportunities for researchers working in this field. It has been widely observed that traditional machine learning algorithms Like SVM, GMM performed excellent in classification whereas the deep learning has dominated the field of medical image analysis to a greater extent. It is evident from the literature survey that the researchers have achieved the accuracies of 100% in classification of cancerous and normal tissue images using different machine learning techniques. This article will provide an insight to the researchers working in this domain to identify which machine learning technique work best on what type of data set, selection of features, various challenges and their proposed solutions in solving this complex problem. Limitations and future research opportunities in the field of implementing different machine learning techniques in cancer diagnosis and classification is also presented at the end of this review article.

Computer Based Diagnosis of Some Chronic Diseases: A Medical Journey of the Last Two Decades

“…As a result, voluminous research articles are present in the literature. Accordingly, several review articles [5,57,59,78,91,92,92,93,103,134,177,201] are also present in the literature on chronic disease prediction. We have summarized the types of methods these review articles considered in Table 1.…”

“…Described only 17 previously published research works 3. Results on only the Surveillance, Epidemiology and End Results (SEER) dataset were investigated to suggest a suitable solution Manhas et al [134] 1. Reviewed image based cancer detection techniques 2.…”

PAC-Net: Multi-pathway FPN with position attention guided connections and vertex distance IoU for 3D medical image detection

“…Automatic medical image detection aims to utilize artificial intelligence techniques to detect lesions in medical images accurately and efficiently, which is one of the most important tasks in computer-aided diagnosis (CAD) systems, and can be embedded into portable imaging devices for intelligent Point of Care (PoC) Diagnostics (Manhas et al, 2021). Recently, with the continuous development of artificial intelligence, deep-learning-based methods have started to play an important part in automatic medical image detection (Wang et al, 2018;Wang et al, 2017;Lee et al, 2018).…”

Research on Key Algorithms of the Lung CAD System Based on Deep Feature Fusion and MKL-SVM-IPSO Recognition

“…In recent years, deep learning has become the mainstream method for medical image feature extraction 10 . Deep learning can capture every detail in the image.…”