A vertebral compression fracture score based on deep generative contextual modeling

Michel Botros, Matthieu Rutten, Twan van Laarhoven, Nikolas Lessmann

Keywords: spine, compression fractures, generative model, inpainting, anomaly detection

TL;DR: This paper proposes a novel automatic method to express the degree of abnormality of the three-dimensional shape of the vertebral body.

Abstract: Grading of vertebral compression fractures most commonly relies on a semi-quantitative grading scale that defines fractures as height loss of the vertebral body. This paper presents an alternative approach that instead considers the three-dimensional shape of the vertebral body and expresses the abnormality of the shape on a scale from 0 to 100. The abnormality is expressed relative to the expected healthy shape, which is predicted by a deep generative model that is provided with contextual information, namely shape and orientation of surrounding vertebrae.

Registration: I acknowledge that acceptance of this work at MIDL requires at least one of the authors to register and present the work during the conference.

Authorship: I confirm that I am the author of this work and that it has not been submitted to another publication before.

Paper Type: novel methodological ideas without extensive validation

Primary Subject Area: Detection and Diagnosis

Secondary Subject Area: Segmentation

Confidentiality And Author Instructions: I read the call for papers and author instructions. I acknowledge that exceeding the page limit and/or altering the latex template can result in desk rejection.