Article Preview
TopIntroduction
Machine learning (ML) as a field in computer science gives computers the ability to learn without being explicitly programmed . Algorithms that can learn from a large body of data and make predictions on the data result from machine learning, which evolved from pattern recognition and computational learning theory in artificial intelligence (Copeland, 2016). Machine learning is applied to a wide range of computing tasks, such as email filtering, detection of network intruders or malicious insiders working towards a data breach, optical character recognition, learning to rank, and computer vision. Researchers are discovering innovative ways to apply machine learning to improve healthcare systems (Liu et al., 2017). In these systems, biomedical engineering as a discipline which advances knowledge in engineering, biology, and medicine to improve human health through cross-disciplinary activities becomes useful (Frosini et al., 2015). This includes applying traditional engineering expertise to analyze and solve biology and medicine problems, providing an overall enhancement of health care. Recently, there have been rapid developments in advanced computing and imaging systems in biomedical engineering, giving rise to a new research dimension. Iqbal, Khan, Saba, & Rehman (2018) highlights a few of the practical applications of medical image analysis. Among which we have: locating abnormal region and other pathologies, measuring tissue sizes, computer-guided surgery, computer-aided diagnosis, radiotherapy, treatment planning, the study of anatomical structure and identification of malignant parts within tumor area in order to minimize the risk of sampling errors in the biopsy. Recently, the number of deaths caused by brain tumors have significantly increased by over 300%. Therefore, researchers have focused on computerized brain tumor diagnosis to obtain important clinical information regarding tumor presence, location, and type.
The human body is made up of different types of cells with each playing a peculiar function. Hence, the human body has to be kept in good condition. To achieve that, the cells in the body multiply and differentiate in a structured way. Although only a few cells fail to regulate their growth, they grow improperly and shape a mass of the tissue known as a tumor that results in extra cells' formation. Brain tumors are a solid neoplasm within the skull that typically occurs in the brain or elsewhere, like in lymphatic tissue, in blood vessels, in cranial nerves, and in brain envelopes. Brain tumors can be categorized by tumor location, tissue type, malignant or benign. The tumors may be either malignant or benign. Malignant tumors lead to cancer, while benign tumors are not cancerous, as shown in Figure 1. In most cases, tumors in the kidney, luminous tissue, breast, or skin melanoma spread to the brain to cause secondary brain damage.
Figure 1. An Image Describing Benign and Malignant Tumour (Doru & Fayed, 2019)
One of the most common medical problems in this field of study is tumor detection as illustrated in Figure 2. One approach to its detection is segmentation.
Figure 2. An MRI Image of the Brain with a Tumour (National Cancer Institute, 2018)
Segmentation is the process of dividing an image into regions of interest (ROIs) to facilitate the characterization, delineation, and visualization of the data. The goal of this process is to increase the significance and accessibility of the interpretation of CT or MRI images in terms of the tumour's location and constraints. Segmentation separates the tumor tissues, such as necrotic and edema, from normal tissues, such as white matter (WM) and gray matter (GM), as seen in Figure 3. Once segmentation is achieved, classification can be carried out.
Figure 3. (a) General brain MR image, (b) Gray Matter, (c) White Matter (Tiwari, Srivastava, & Pant, 2019)