Research Article
An Optimized KNN Classification Approach for Brain Tumor Prediction Using Order Statistic Techniques
Issue:
Volume 14, Issue 3, June 2026
Pages:
108-120
Received:
5 April 2026
Accepted:
20 April 2026
Published:
13 May 2026
DOI:
10.11648/j.ijsts.20261403.11
Downloads:
Views:
Abstract: A tumor, otherwise called a neoplasm, are abnormal accumulations of tissues that can either be solid or filled with fluid. Lumps or growths known as tumors constitute clusters of abnormal cells, originating from any of the trillions of cells present throughout the body. The growth patterns and behaviors of tumors vary significantly, depending on whether their classification is cancerous (malignant), non-cancerous (benign) or precancerous. Medical imaging is a process that is widely used to produce images of the human body for both medical and research purposes. A significant focus in clinical research is the automated detection of tumors in the brain using Magnetic Resonance Imaging scans. Magnetic Resonance Imaging (MRI) is a sophisticated medical-imaging technology. It enables non-invasive visualization of the internal anatomy of the human body. Segmenting MRI images is vital for the detection of brain tumor. Complexities of tumor characteristics, such as shape and size, tumor-location, gray?level intensity, makes it challenging to classify segmented MRI scans into normal versus abnormal. Histogram Segmentation method, proposed in this study, is based on differentiating with order static filter. In post-processing, morphological techniques are practically employed to enhance the visibility of brain tumors. The K-Nearest Neighbors (KNN) classification method is then utilized to categorize tumor values into their respective categories, such as benign or malignant, based on the tumor's size. This approach helps patients and students to understand the tumor and helps physicians decide treatment based on the tumor’s size, shape, location and the type.
Abstract: A tumor, otherwise called a neoplasm, are abnormal accumulations of tissues that can either be solid or filled with fluid. Lumps or growths known as tumors constitute clusters of abnormal cells, originating from any of the trillions of cells present throughout the body. The growth patterns and behaviors of tumors vary significantly, depending on wh...
Show More
,
Swathi Pothala
