A Hierarchical Prototype-Based Approach for Osteoporosis Classification

Abstract

The widespread use of medical imaging by doctors has made computers a necessity for imaging and analysis. Several clinical decision-making tools have been developed to assess osteoporosis risk through bone density measurement and/or bone image analysis. The purpose of this study is to develop a hierarchical prototype-based approach for classification of osteoporosis. This approach allows you to interactively recognize the multimodal distribution and data space of different levels of data streaming, where you also identify meaningful prototypes for self-organization and self-development hierarchies. The decision-making process is based on the "closest prototype" principle and is clear. The proposed method allows users to present the information retrieved from the data in a hierarchical and easily interpretable form based on a prototype, and is an attractive tool for solving large and complex problems in the real world.