Interscan measurement error in measurements of the meniscus, including the meniscal roots, by fully automatic three-dimensional MRI analysis

• Published in: Osteoarthritis imaging Vol. 2; no. 1; p. 100007
• Main Authors: Katano, Hisako, Ozeki, Nobutake, Koga, Hideyuki, Suzuki, Kenji, Masumoto, Jun, Tomita, Makoto, Sekiya, Ichiro
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.03.2022
• Subjects: 3D MRI; Automatic; Interscan error; Knee; Meniscus; Thickness; Knee; 3D MRI; MM; LM; DMOADs; PDW; CNN; Meniscus; MME; MRI; 3.0T; ICC; ROI; SPGR; Thickness; 3D; SD; OA; DSC; Interscan error; Automatic
• ISSN: 2772-6541; 2772-6541
• DOI: 10.1016/j.ostima.2022.100007

The meniscus coverage ratio is an important factor in knee osteoarthritis. The ratio calculated by our software, which automatically segments meniscus and cartilage regions from knee MR images, correlates with cartilage measurements. However, the significance of the changes is questionable due to possible intermeasurement error during sequential observations. The purpose of this study was to show the interscan measurement errors of meniscal measurements, especially the meniscal coverage ratio. Ten healthy volunteers had two 3.0T MRI examinations on the same day. We determined the volume, area, coverage ratio, width, and thickness of the medial (MM) and the lateral (LM) menisci, including their roots, twice per volunteer, using the software with a 3D convolutional neural network algorithm for segmentation. The relative interscan error, expressed as average ± SD, was the absolute value of the difference between the two measurements divided by the average value of all measurements. The relative interscan errors were 1.3 ± 1.1%, 1.7 ± 1.8%, 0.7 ± 0.9%, 1.0 ± 0.9%, 1.9 ± 1.3%, and 0.9 ± 0.9% for MM volume, LM volume, MM area, LM area, MM coverage ratio, and LM coverage ratio, respectively. The relative interscan errors in the middle segment were 0.6 ± 0.3%, 1.3 ± 1.3%, 2.3 ± 1.2%, and 1.6 ± 1.2% for MM width, LM width, MM thickness, and LM thickness, respectively. This study documented the interscan measurement errors in meniscus measurements. These can be indicators to determine the significance of changes observed in longitudinal and interventional studies.