Volumetric quantitative measurement of hip effusions by manual versus automated artificial intelligence techniques: An OMERACT preliminary validation study

Jacob L Jaremko; Banafshe Felfeliyan; Abhilash Hareendranathan; Bashiar Thejeel; Quinn-Laurin Vanessa; Mikkel Østergaard; Philip G Conaghan; Robert G W Lambert; Janet L Ronsky; Walter P Maksymowych

doi:10.1016/j.semarthrit.2021.03.009

Volumetric quantitative measurement of hip effusions by manual versus automated artificial intelligence techniques: An OMERACT preliminary validation study

Semin Arthritis Rheum. 2021 Jun;51(3):623-626. doi: 10.1016/j.semarthrit.2021.03.009. Epub 2021 Mar 17.

Affiliations

¹ Department of Radiology & Diagnostic Imaging, University of Alberta, 2A2.41 WMC, 8440 - 112 St. NW, Edmonton, AB, Canada. Electronic address: jjaremko@ualberta.ca.
² Biomedical Engineering, Schulich School of Engineering, University of Calgary, Canada.
³ Department of Radiology & Diagnostic Imaging, University of Alberta, 2A2.41 WMC, 8440 - 112 St. NW, Edmonton, AB, Canada.
⁴ Copenhagen Center for Arthritis Research, Center for Rheumatology and Spine Diseases, Rigshospitalet, Copenhagen, Denmark; Department of Clinical Medicine, University of Copenhagen, Copenhagen, Denmark.
⁵ Leeds Institute of Rheumatic and Musculoskeletal Medicine, University of Leeds, and NIHR Leeds Biomedical Research Centre, UK.
⁶ Mechanical & Manufacturing Engineering, Schulich School of Engineering, University of Calgary.
⁷ Department of Medicine, University of Alberta, Canada.

PMID: 33781576
DOI: 10.1016/j.semarthrit.2021.03.009

Abstract

Objective: Preliminary assessment, via OMERACT filter, of manual and automated MRI hip effusion Volumetric Quantitative Measurement (VQM).

Methods: For 358 hips (93 osteoarthritis subjects, bilateral, 2 time points), 2 radiologists performed manual VQM using custom Matlab software. A Mask R-CNN artificial-intelligence (AI) tool was trained to automatically compute joint fluid volumes.

Results: Manual VQM had excellent inter-observer reliability (ICC 0.96). AI predicted hip fluid volumes with ICC 0.86 (status), 0.58 (change) vs. 2 human readers.

Conclusion: Hip joint fluid volumes are reliably assessed by VQM. It is feasible to automate this approach using AI, with promising initial reliability.

Keywords: Artificial intelligence; Effusion; Hip joint; MRI; OMERACT; Osteoarthritis.

Publication types

Research Support, Non-U.S. Gov't

MeSH terms

Artificial Intelligence*
Hip Joint* / diagnostic imaging
Humans
Magnetic Resonance Imaging
Reproducibility of Results
Synovial Fluid