 朗读者：夏爽教授  Psychoradiologic Utility of MR Imaging for Diagnosis of Attention Deficit Hyperactivity Disorder: A Radiomics Analysis Huaiqiang Sun, Ying Chen, Qiang Huang, Su Lui, Xiaoqi Huang, Yan Shi, Xin Xu, John A. Sweeney1, Qiyong Gong Address correspondence to Q.G. (e-mail: qiyonggong@hmrrc.org.cn). Vol. 287: , Issue. 2, : Pages. 620-630 (Issue publication date: May 2018) https:///10.1148/radiol.2017170226 By integrating imaging features extracted from both gray and white matter and by using an all-relevant feature selection algorithm and random forest classifier, the authors of this study established a cerebral radiomics analysis framework for assisting diagnosis and subtyping of attention deficit hyperactivity disorder. Purpose To identify cerebral radiomic features related to diagnosis and subtyping of attention deficit hyperactivity disorder (ADHD) and to build and evaluate classification models for ADHD diagnosis and subtyping on the basis of the identified features. Materials and Methods A consecutive cohort of 83 age- and sex-matched children with newly diagnosed and never-treated ADHD (mean age 10.83 years ± 2.30; range, 7–14 years; 71 boys, 40 with ADHD-inattentive [ADHD-I] and 43 with ADHD-combined [ADHD-C, or inattentive and hyperactive]) and 87 healthy control subjects (mean age, 11.21 years ± 2.51; range, 7–15 years; 72 boys) underwent anatomic and diffusion-tensor magnetic resonance (MR) imaging. Features representing the shape properties of gray matter and diffusion properties of white matter were extracted for each participant. The initial feature set was input into an all-relevant feature selection procedure within cross-validation loops to identify features with significant discriminative power for diagnosis and subtyping. Random forest classifiers were constructed and evaluated on the basis of identified features. Results No overall difference was found between children with ADHD and control subjects in total brain volume (1069830.00 mm3 ± 90743.36 vs 1079 213.00 mm3 ± 92742.25, respectively; P = .51) or total gray and white matter volume (611978.10 mm3 ± 51622.81 vs 616960.20 mm3 ± 51872.93, respectively; P = .53; 413532.00 mm3 ± 41 114.33 vs 418173.60 mm3 ± 42395.48, respectively; P = .47). The mean classification accuracy achieved with classifiers to discriminate patients with ADHD from control subjects was 73.7%. Alteration in cortical shape in the left temporal lobe, bilateral cuneus, and regions around the left central sulcus contributed significantly to group discrimination. The mean classification accuracy with classifiers to discriminate ADHD-I from ADHD-C was 80.1%, with significant discriminating features located in the default mode network and insular cortex. Conclusion The results of this study provide preliminary evidence that cerebral morphometric alterations can allow discrimination between patients with ADHD and control subjects and also between the most common ADHD subtypes. By identifying features relevant for diagnosis and subtyping, these findings may advance the understanding of neurodevelopmental alterations related to ADHD. © RSNA, 2017 Notes: 1. Attention Deficit Hyperactivity Disorder 注意缺陷多动障碍 2. algorithm [ˈælɡəˌrɪðəm] n. 演算法; 运算法则; 计算程序 3. cerebral [səˈri:brəl] adj. 脑的; 大脑的; 理智的; 非感情方面的 4. radiomics 影像组学 5.temporal lobe [ˈtɛmpərəl lob] 颞叶 6. cuneus ['kju:nɪrs] n. 楔片 7. sulcus ['sʌlkəs] n. 脑沟 8. insular cortex [ˈɪnsəlɚ ˈkɔrˌteks] 岛叶皮质 9. morphometric 形态测定的 10. neurodevelopmental [nəroʊdi:'veləpmentl] 神经发育的 加减号读法：plus and minus Random Forest随机森林算法 Random Forest是加州大学伯克利分校的Breiman Leo和Adele Cutler于2001年发表的论文中提到的新的机器学习算法，可以用来做分类，聚类，回归，这里只简单介绍该算法在分类上的应用。 Random Forest（随机森林）算法是通过训练多个决策树，生成模型，然后综合利用多个决策树的分类结果进行投票，从而实现分类。随机森林算法只需要两个参数：构建的决策树的个数t，在决策树的每个节点进行分裂时需要考虑的输入特征的个数m。 来源：每天朗读一段医学影像学英语文章 |
|
|
