Breast ultrasound tumour classification: A Machine Learning—Radiomics based approach

Abstract

Prediction of breast tumour malignancy using ultrasound imaging, is an important step for early detection of breast cancer. An efficient prediction system can be a great help to improve the survival chances of the involved patients. In this work, a machine learning (ML)—radiomics based classification pipeline is proposed, to perform this predictive modelling task, in a much more efficient manner. Multiple different types of image features of the region of interests are considered in this work, followed by a recursive feature elimination based feature selection step. Furthermore, a synthetic minority oversampling technique based step is also included in the pipeline, to deal with the class imbalance problem, that is often present in medical imaging datasets. Various ML models are considered in the subsequent model training phase, on a publicly available breast ultrasound image dataset (BUSI). From experimental analysis it has been observed that, shape, texture and histogram oriented gradients related features are the most informative, with respect to the predictive modelling task. Furthermore, it was observed that ensemble learners such as random forest, gradient boosting and AdaBoost classifiers are able to achieve significant results with respect to multiple evaluation metrics. The proposed approach achieved the state‐of‐the‐art accuracy, area under the curve, F1‐score and Mathews correlation coefficient values of 0.974, 0.97, 0.94 and 0.959, respectively, on the BUSI dataset. Such kind of impressive results suggest that the proposed approach can have a very high practical utility, in real medical diagnostic settings.

Breast Cancer Classification from Ultrasound Images Using Probability-Based Optimal Deep Learning Feature Fusion

“…Mishra et al [ 49 ] introduced a machine learning (ML) radiomics-based classification pipeline. The region of interest (ROI) was separated, and useful features were extracted.…”

Semi-supervised vision transformer with adaptive token sampling for breast cancer classification

“…Such worrying numbers highlight the significance of properly using present technological advancements to undertake efficient BC detection in its early stage. In particular, a recent development in artificial intelligence (AI) that explores the usage of deep learning models in a wide spectrum of health care applications presents a promising direction toward building a more effective computer-aided diagnosis (CAD) system for BC detection (Hu et al, 2020;Mewada et al, 2020;Moon et al, 2020;Boumaraf et al, 2021;Eroğlu et al, 2021;Mishra et al, 2021).…”

“…A variety of imaging techniques can be used for BC detection and diagnosis, including X-rays (mammograms) (Abdelrahman et al, 2021), ultrasound (sonography) (Moon et al, 2020;Mishra et al, 2021), thermography (Singh and Singh, 2020), magnetic resonance imaging (MRI) (Mann et al, 2019), and histopathology imaging (Benhammou et al, 2020). Ultrasound has been a widely adopted, low-cost, non-invasive, and nonradioactive imaging modality in the procedure of BC diagnosis and is usually followed by histopathological analysis.…”

CR‐SSL: A closely related self‐supervised learning based approach for improving breast ultrasound tumor segmentation

“…This is where the recent advancements in the fields of artificial intelligence (AI) and machine learning (ML) can be of great help since they enable the development of effective data-driven techniques for disease detection. [5][6][7][8][9][10][11][12] Such automatic techniques can help provide crucial real-time decision support to the involved diagnostic radiologists. 13,14 However, these approaches require access to large-sized human-annotated datasets to perform effective model training, which is usually hard to acquire in the medical domain.…”

“…As shown by several studies, 2–4 detecting tumors in breast ultrasound (BUS) images requires a high level of radiological expertise, as this process often suffers from a high rate of interobserver variability. This is where the recent advancements in the fields of artificial intelligence (AI) and machine learning (ML) can be of great help since they enable the development of effective data‐driven techniques for disease detection 5–12 . Such automatic techniques can help provide crucial real‐time decision support to the involved diagnostic radiologists 13,14 .…”