Certain investigations on bone age assessment using geometric features

Abstract

The thesis reports four new computerized methods for Bone Age Assessment to estimate the bone age of a child from his her hand wrist radiograph BAA is very significant in the area of pediatrics especially in relation to endocrino logical problems and growth disorders Based on the skeletal development of the left hand wrist bone age is assessed and compared with the chronological age A remarkable difference between the two is an indication of growth abnormalities This practice is used in the management and diagnosis of endocrine disorders and also in monitoring the therapeutic effect of treatment The main clinical methods for skeletal bone age estimation are the Greulich and Pyle method and the Tanner and Whitehouse method GP is an atlas matching method while TW is a score assigning method TW method is more accurate of the two It is technically feasible to automate BAA and much research has been undertaken to computerize the TW method This research is undertaken to develop new promising techniques for computerized BAA with improved accuracy specificity precision and recall The objectives of the research are to study the existing techniques available for newlineBAA to develop new convex hull approach for BAA from carpal bones to propose newlinenew feature ratio approach for BAA from carpal and radius bones, to develop new decision tree approach for BAA from radius and ulna bones to propose new Hausdorff distance approach for BAA from the Epiphysis Metaphysis Region of Interest of phalangeal bones to analyze and compare the performance of the proposed four BAA methods using Partitioning technique and to optimize the performance of BAA by the proposed four BAA approaches newline newline