Occult Spastic Diplegic Cerebral Palsy Recognition Using Efficient Machine Learning for Big Data and Structural Connectivity Abnormalities Analysis
文献类型：期刊论文
| 作者 | Duan SF(段绍峰) ; Huang Q(黄琪); Shan, BC; Ma, Y; Huang, Q; Mu, XT; Duan, SF; Dan BC(单保慈)
|
| 刊名 | JOURNAL OF MEDICAL IMAGING AND HEALTH INFORMATICS
 |
| 出版日期 | 2018 |
| 卷号 | 8期号:2页码:317-324 |
| 关键词 | Machine Learning Structural Connectivity Occult Spastic Diplegic Cerebral Palsy Tractography Graph Theory |
| ISSN号 | 2156-7018 |
| DOI | 10.1166/jmihi.2018.2282 |
| 文献子类 | Article |
| 英文摘要 | Occult spastic diplegic cerebral palsy (occult SDCP) is a special type of SDCP with invisible brain injuries in conventional MRI. Invisible is a subjective perception, so that it is difficult for clinicians to distinguish the occult cases from small brain injured cases. Besides, whether the invisible brain injuries disrupt the structural connectivity is unknown. We aimed to make use of machine learning methods to classify occult SDCP and analyze the integrity of their structural connectivity in the brain networks of occult SDCP patients. Thus, we used k-nearest neighbor algorithm (k-NN) to recognize occult SDCP patients from large datasets, then constructed the structural networks of occult SDCP children and eleven age- and sex-matched healthy subjects, a permutation test was used to evaluate the groupwise difference in their global and local topological features. The study found k-NN provided effective classification results. And in network analysis, global efficiency in the occult SDCP group was significantly lower than that in the normal control group (p < 0.001). On the other hand, the occult SDCP group had larger normalized clustering coefficient (gamma, p < 0.001), normalized characteristic path length (lambda, p = 0.005) and small-worldness (sigma, p = 0.002), when compared to the normal control group. Both the control and occult SDCP groups' networks had small-worldness characteristics (gamma >> 1, lambda approximate to 1, and sigma >> 1). These findings revealed that the brain of occult SDCP patients had abnormal structural connectivity and their brain as a complex system had lower function efficiency, which furthered our understanding of the mechanisms underlying occult SDCP. |
| 电子版国际标准刊号 | 2156-7026 |
| WOS关键词 | GRAPH-THEORETICAL ANALYSIS ; WHITE-MATTER LESIONS ; BRAIN CONNECTIVITY ; NETWORK ANALYSIS ; CHILDREN ; TRACTOGRAPHY ; AREAS |
| WOS研究方向 | Mathematical & Computational Biology ; Radiology, Nuclear Medicine & Medical Imaging |
| 语种 | 英语 |
| WOS记录号 | WOS:000423786700023 |
| 源URL | [http://ir.ihep.ac.cn/handle/311005/285594]  |
| 专题 | 高能物理研究所_核技术应用研究中心 |
| 通讯作者 | Dan BC(单保慈) |
| 作者单位 | 中国科学院高能物理研究所 |
| 推荐引用方式 GB/T 7714 | Duan SF,Huang Q,Shan, BC,et al. Occult Spastic Diplegic Cerebral Palsy Recognition Using Efficient Machine Learning for Big Data and Structural Connectivity Abnormalities Analysis[J]. JOURNAL OF MEDICAL IMAGING AND HEALTH INFORMATICS,2018,8(2):317-324. |
| APA | 段绍峰.,黄琪.,Shan, BC.,Ma, Y.,Huang, Q.,...&单保慈.(2018).Occult Spastic Diplegic Cerebral Palsy Recognition Using Efficient Machine Learning for Big Data and Structural Connectivity Abnormalities Analysis.JOURNAL OF MEDICAL IMAGING AND HEALTH INFORMATICS,8(2),317-324. |
| MLA | 段绍峰,et al."Occult Spastic Diplegic Cerebral Palsy Recognition Using Efficient Machine Learning for Big Data and Structural Connectivity Abnormalities Analysis".JOURNAL OF MEDICAL IMAGING AND HEALTH INFORMATICS 8.2(2018):317-324. |
入库方式： OAI收割
来源：高能物理研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。