Automated noninvasive detection of idiopathic scoliosis in children and adolescents: A principle validation study

被引：27

作者：

Sudo, Hideki ^{[1
,2
]}

Kokabu, Terufumi ^{[2
]}

Abe, Yuichiro ^{[3
]}

Iwata, Akira ^{[2
]}

Yamada, Katsuhisa ^{[2
]}

Ito, Yoichi M. ^{[4
]}

Iwasaki, Norimasa ^{[2
]}

Kanai, Satoshi ^{[5
]}

机构：

[1] Hokkaido Univ, Fac Med & Grad Med, Dept Adv Med Spine & Spinal Cord Disorders, N15W7, Sapporo, Hokkaido 0608638, Japan

[2] Hokkaido Univ Hosp, Dept Orthopaed Surg, N14W5, Sapporo, Hokkaido 0608648, Japan

[3] Eniwa Hosp, Dept Orthopaed Surg, Koganechuo 2-1-1, Eniwa, Hokkaido 0611449, Japan

[4] Hokkaido Univ, Grad Sch Med, Dept Biostat, Sapporo, Hokkaido, Japan

[5] Hokkaido Univ, Grad Sch Informat Sci & Technol, Div Syst Sci & Informat, N14W9, Sapporo, Hokkaido 0600814, Japan

来源：

SCIENTIFIC REPORTS | 2018年 / 8卷

关键词：

TOPOGRAPHY; RELIABILITY; IMPROVEMENT; ACCURACY;

D O I：

10.1038/s41598-018-36360-w

中图分类号：

O [数理科学和化学]; P [天文学、地球科学]; Q [生物科学]; N [自然科学总论];

学科分类号：

070301 [无机化学]; 070403 [天体物理学]; 070507 [自然资源与国土空间规划学]; 090105 [作物生产系统与生态工程];

摘要：

Idiopathic scoliosis is the most common pediatric musculoskeletal disorder that causes a three-dimensional deformity of the spine. Early detection of this progressive aliment is essential. The aim of this study is to determine outcomes using a newly developed automated asymmetry-evaluation system for the surface of the human back using a three-dimensional depth sensor. Seventy-six human subjects suspected to have idiopathic scoliosis were included in this study. Outcome measures include patient demographics, radiographic measurements, and asymmetry indexes defined in the automated asymmetry-recognition system. The mean time from scanning to analysis was 1.5 seconds. For predicting idiopathic scoliosis of greater than 25 degrees, the area under the curve was 0.96, sensitivity was 0.97, and specificity was 0.88. The coefficient of variation for repeatability analyses using phantom models was 1-4%. The intraclass correlation coefficient obtained for intra-observer repeatability for human subjects was 0.995. The system three-dimensionally scans multiple points on the back, enabling an automated evaluation of the back's asymmetry in a few seconds. This study demonstrated discriminative ability in determining whether an examinee requires an additional x-ray to confirm diagnosis.

引用

页数：10

共 37 条

[1]

Computing and rendering point set surfaces [J].