Abstract
This chapter presents an overview of methods that have been proposed for analysis of histopathological images. Diagnosing and detecting abnormalities in medical images helps the pathologist in making better decisions. Different machine learning algorithms such as k-nearest neighbor, random forest, support vector machine, ensemble learning, multilayer perceptron, and convolutional neural network are incorporated for carrying out the analysis process. Further, multi-criteria decision-making (MCDM) methods such as SAW, WPM, and TOPSIS are used to improve the efficiency of the decision-making process.
TopIntroduction
Over past decades, determination of diseased histopathological image has become a difficult task for radiologists. Image analysis with computer aided analytical approaches for diagnosis play a major rule in present era (Barik, Dubey, Misra, Borthakur, Constant, Sasane, & Mankodiya, 2018). Histopathology refers a minute identification of a tissue in order to learn the symptom of a disease which is being carried out from biopsy examination or by a surgical specimen. The process highly depends on radiologist or pathologist’s opinion which becomes a hectic issue in some preliminary cases. To add up with manual result, a computational environment is recommended for working in digitized tissue histopathological images. As microscopic images have a momentous role for visualizing and interpreting the cells and tissues, image analysis is opted in this chapter with new machine learning techniques and decision making algorithms incorporated together for disease detection, diagnosis and prognosis prediction which can aid the physician’s opinion. In addition, this chapter also deals quantitative characterization and analysis of pathological imagery which is important for clinical and research application to reduce variations in diagnosis and to comprehend the biological mechanism of disease. This chapter mainly focuses on method to identify the categories by classification into favorable or unfavorable tissue morphology and multicriteria decision making method assists with it to correctly diagnose the diseased tissue for further treatment planning. This chapter also includes experimental analysis results of various classification algorithms with multicriteria decision making techniques to determine class labels with high accuracy (Blessa Binolin Pepsi & Bargavi, 2017). Overall design of this chapter is presented in Figure 1.
TopMain Focus Of The Chapter
This chapter presents machine learning techniques for image analysis and multicriteria decision making techniques for quantitative characterization and analysis of classification algorithms (Thanki, Rohit, Surekha, Nilanjan, & Amira, 2018).
Issues, Controversies, Problems
Conventional histopathological image analysis includes biopsy examination or by means of surgical specimen. Accuracy of this method is in the hands of radiologist or pathologist decision, it may be error prone. Moreover, this method is invasive; it incurs time, pain and discomfort.
Key Terms in this Chapter
Machine Learning: A field of information technology that has the ability to learn data insights by using statistical techniques.
Image Analysis: Process of extracting meaningful information from images.
Digital Pathology: Provides an image-based environment which includes acquisition, diagnosis, and interpretation from the generated digitized data.
Histopathology: Includes study of changes in diseased tissues.
Computer-Aided Diagnosis: Use of computer systems to aid in medical image diagnosis to improve the quality of treatment.
Multi-Criteria Decision Making: A field of operation research that evaluates multiple conflicting criteria in decision making.
Root Mean Squared Error: A measure to evaluate the differences between predicted model values to the observed values.