Title: Application of the item response theory in assessing the scale for clinical classification of breast cancer types

Authors: Donald Douglas Atsa'am; Terlumun Gbaden; Ruth Wario

Addresses: Department of Computer Science and Informatics, Faculty of Natural and Agricultural Sciences, University of the Free State, Qwaqwa Campus, South Africa ' Department of Computer Science, College of Applied Sciences, Joseph Sarwuan Tarka University, Makurdi, Nigeria ' Department of Computer Science and Informatics, Faculty of Natural and Agricultural Sciences, University of the Free State, Qwaqwa Campus, South Africa

Abstract: This study assessed the validity and reliability of an existing scale for classifying breast cancer types. The scale consists of nine variables that assess biopsies on a scale of 1 to 10 to determine if a patient's breast cancer is benign or malignant. The item response theory (IRT) was employed for item and scale analysis of the breast cancer dataset consisting of 699 records with known breast cancer types. IRT indexes for assessing model and item fits fell within acceptable ranges. Slope parameters and factor analysis produced results which proved that each item had a reasonable association with breast cancer categories. Each item contained some amount of statistical information that could lead to the discrimination of breast cancer types. The scale reliability evaluated to the value, 0.91. The overall scale is sound and appropriate for breast cancer categorisation and thus recommended for use in research and practice.

Keywords: item response theory; clinical classification; breast cancer; benign; malignant.

DOI: 10.1504/IJMEI.2025.145081

International Journal of Medical Engineering and Informatics, 2025 Vol.17 No.2, pp.173 - 183

Received: 07 Jun 2022
Accepted: 11 Aug 2022
Published online: 18 Mar 2025 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article