FemoraLyze: A Modular Framework for Proximal Femur Analysis

Marten Johannes Finck, Niklas Christoph Koser, Jan-Bernd Hövener, Claus-C. Glüer, Soren Pirk

Keywords: Femur Assessment, Deep Learning, Bone Microstructure, Bone Geometry

TL;DR: FemoraLyze is a deep-learning based framework for the automated assessment of the proximal femur.

Abstract:

The proximal femur is exposed to an increased risk of fracture, particularly in the context of osteoporosis. As the prevalence increases with age, it is expected that the incidence of osteoporotic fractures will likely continue to rise in terms of demographic trends.
Early, guideline-based therapy offers strong prospects of success, but requires precise and reliable diagnostic and prognostic procedures. Automated bone metrics are suitable for this purpose and could also be used for other applications such as preoperative planning of total hip arthroplasty in patients affected by arthritis.
In this paper we present FemoraLyze, which is a modular deep-learning-based Python framework that combines the automated and differentiated calculation of segmentation masks, bone structure and geometry parameters of the proximal femur based on a computed tomography (CT) image. The code is available at https://github.com/martenfi/FemoraLyze.

Submission Number: 128