Center for Information and Neural Networks (CiNet), Advanced ICT Research Institute, National Institute of Information and Communications Technology (NICT), Suita, Osaka, Japan; Japan Society for the Promotion of Science, Chiyoda-ku, Tokyo, Japan; Human Health Sciences, Graduate School of Medicine, Kyoto University, Sakyo-ku, Kyoto, Japan.
Center for Information and Neural Networks (CiNet), Advanced ICT Research Institute, National Institute of Information and Communications Technology (NICT), Suita, Osaka, Japan.
Ultrasound Med Biol. 2022 Sep;48(9):1966-1976. doi: 10.1016/j.ultrasmedbio.2022.06.001. Epub 2022 Jul 10.
Freehand 3-D ultrasound (3DUS) system is a promising technique for accurately assessing muscle morphology. However, its accuracy has been validated mainly in terms of volume by examining lower limb muscles. This study was aimed at validating 3DUS in the measurements of 3-D surface shape and volume by comparing them with magnetic resonance imaging (MRI) measurements while ensuring the reproducibility of participant posture by focusing on the shoulder muscles. The supraspinatus, infraspinatus and posterior deltoid muscles of 10 healthy men were scanned using 3DUS and MRI while secured by an immobilization support customized for each participant. A 3-D surface model of each muscle was created from the 3DUS and MRI methods, and the agreement between them was assessed. For the muscle volume, the mean difference between the two models was within -0.51 cm. For the 3-D surface shape, the distances between the closest points of the two models and the Dice similarity coefficient were calculated. The results indicated that the median surface distance was less than 1.12 mm and the Dice similarity coefficient was larger than 0.85. These results suggest that, given the aforementioned error is permitted, 3DUS can be used as an alternative to MRI in measuring volume and surface shape, even for the shoulder muscles.
自由手三维超声（3DUS）系统是一种很有前途的技术，可以准确评估肌肉形态。然而，其准确性主要通过检查下肢肌肉的体积来验证。本研究旨在通过比较 3DUS 与磁共振成像（MRI）的测量结果，同时通过关注肩部肌肉来确保参与者姿势的可重复性，从而验证 3DUS 在三维表面形状和体积测量方面的准确性。使用为每个参与者定制的固定支架，对 10 名健康男性的冈上肌、冈下肌和后三角肌进行了 3DUS 和 MRI 扫描。从 3DUS 和 MRI 方法创建了每个肌肉的三维表面模型，并评估了它们之间的一致性。对于肌肉体积，两个模型之间的平均差异在-0.51cm 以内。对于三维表面形状，计算了两个模型的最近点之间的距离和 Dice 相似系数。结果表明，中位数表面距离小于 1.12mm，Dice 相似系数大于 0.85。这些结果表明，在允许上述误差的情况下，3DUS 可以替代 MRI 测量体积和表面形状，即使是肩部肌肉。
