Archives

  • 2019-10
  • 2019-11
  • 2020-03
  • 2020-07
  • 2020-08
  • 2021-03
  • br Introduction br The cancer tumor is developed through the

    2020-08-12


    1. Introduction
    The cancer tumor is developed through the abnormal growth of cells, which invades the surrounding tissues in the human body. There are two classes of tumors, i.e., malignant and benign. Malignant tumors spread to other parts of the body while benign tumors do not spread to the surrounding 36913-39-0 [1]. In women, breast cancer is one of the most common cancers and a severe public health issue all over the world. It is a more widespread disease in developed countries [2]. According to the report of International Agency for Research on Cancer (IARC) presented
    ∗ Corresponding author. E-mail address: [email protected] (I.U. Din).
    by the World Health Organization (WHO), in the year 2012, approximately 8.2 million deaths are caused by cancer, which further states that in 2030 the death ratio from cancer is expected to increase up to 27 million [3]. Therefore, on-time detection, early diagnosis, and active prevention can reduce the possibility of deaths.
    The use of image processing and computer vision based tech-niques are commonly used in cancer screening for the last three decades. In this regards, techniques such as diagnoses of Mammo-grams (X-rays), Magnetic Resonance Imaging (MRI), Ultrasound (sonography) and Thermography are generally used for the de-tection and diagnosis of breast cancer [4]. However, the most reliable diagnoses with reassurance are achieved through biopsy. The common biopsy techniques are Fine Needle Aspiration (FNA),
    Core Needle Biopsy (CNB), Vacuum-Assisted, and Surgical Open Biopsy (SOB) [5]. All biopsy procedures consist of collecting sam-ples of cells or tissues fixed on a glass slide for further staining, which is finally examined by microscopes. Similarly, Histopathol-ogy images are used for diagnosing all types of cancer including breast cancer [6]. The presence of breast lump and its classifica-tion into benign and malignant requires a pathological diagnosis for treatment and cure [7]. The Fine Needle Aspiration Cytology (FNAC) has become a popular approach in the preoperative of breast masses assessment. The major goal of FNAC is to differ-entiate benign from malignant lesions. However, the accuracy of FNAC as compared to visual analysis range from 65% to 98% on the basis of the doctor’s knowledge and experience [8]. Human error can be the cause of incorrect diagnoses or may delay the correct diagnoses that can eventually result in fatalities. In order to han-dle such a situation, Computer Aided Diagnosis (CAD) technology has been extensively applied to minimize false-negative rates and to increase the true positive rate of breast tumor [9].
    Numerous researchers have worked on the detection and clas-sification of breast cancer cells and proposed different automated solutions which are based on artificial intelligence based machine learning techniques like artificial neural networks (ANN), Sup-port Vector Machine (SVM), Naïve Bayesian (NB), Random Forest (RF), Decision Trees, Vector Quantization, etc [10–13]. The au-thors [14–16] used Particle Swarm Optimization along with Ker-nel Density Estimation (PSO-KDE), fuzzy cerebellar model neural network and decision tree methods for the detection and classi-fication of breast cancer. In probabilistic approaches, a compar-ative study shows that weighted Naïve Bayes algorithm has the best performance as compared to Naïve Bayes [12]. Moreover, a method known as sparse component analysis uses a combina-tion of acriflavine staining and fluorescence microscopy for the segmentation of nuclei for breast cancer detection [17].
    Internet of Things (IoT) is one of the fast-growing phenom-ena in information and communication technologies [18]. It has gained significant achievements in numerous application do-mains, such as e-Health, businesses, home services, military, and automation, etc. In application domains, IoT is aimed to intercon-nect communicating devices and gather data to provide efficient and robust services to users. In the domain of e-Health care appli-cations, remotely connecting devices can exchange information among patients, expert physicians and/or Artificial Intelligence (AI) based health care experts. The exchange of information through medical devices that are connected via the Internet to the cloud platform for quality services is also referred to as Internet of Medical Things (IoMT). IoMT is a newly emerging technology to interconnect medical devices and software applications to increase the productivity, reliability, and accuracy of medical devices in the health care industry. Recently, approximately 3.7 million medical devices are in operation, and it is estimated that the IoMT trade will touch 136.8 billion/year up to the year 2021 globally [19]. Nowadays, the concept of IoMT is mostly utilized for activities such as remote patient monitoring, tracking of med-ication orders, and wearable mobile health devices [20,21]. IoMT also provides a wide range of services to medical experts, such as sending feedback to medical staff, device information and its configurations according to the need of patients and experts [22]. IoMT provides a quick and ultimate access to different reports that assist surgeons in operation theaters during surgeries.