An interpretable multiparametric radiomics model for the diagnosis of schizophrenia using magnetic resonance imaging of the corpus callosum.
An interpretable multiparametric radiomics model for the diagnosis of schizophrenia using magnetic resonance imaging of the corpus callosum.
复制标题
可解释的多参数放射组学模型用于精神分裂症的诊断，使用的是胼胝体的磁共振成像。
DOI：
10.1038/s41398-021-01586-2
复制
发表时间：
2021-09-06
影响因子：
6.8
通讯作者：
Lee SH
中科院分区：
文献类型：
--
作者：
Bang M;Eom J;An C;Kim S;Park YW;Ahn SS;Kim J;Lee SK;Lee SH
有一个日益增长的需求，以发展新的战略，精神分裂症的诊断使用神经影像学生物标志物。我们研究了使用胼胝体（CC）的T1加权和扩散张量图像的放射组学特征的精神分裂症诊断模型的鲁棒性。共有165名参与者[86名精神分裂症患者和79名健康对照（HC）]被分配到训练组（N = 115）和测试组（N = 50）。从T1加权、表观扩散系数（ADC）和各向异性分数（FA）图像（N = 1605）中提取CC亚区的放射组学特征。在特征选择之后，训练各种组合的分类器，并在性能最好的分类器中采用贝叶斯优化。对该模型的识别、校准和临床实用性进行了评估。一个在线计算器被构建来提供患精神分裂症的概率。应用SHapley加性解释（SHAP）对模型的可解释性进行了探讨。我们确定了30个放射组学特征来区分精神分裂症参与者与HC。贝叶斯优化模型实现了最高性能，在测试集中，曲线下面积（AUC）、准确度、灵敏度和特异性分别为0.89（95%置信区间：0.81-0.98）、80.0、83.3和76.9%。最终模型在在线计算器中提供临床概率。SHAP的模型解释表明，来自后CC的二阶特征与精神分裂症的风险高度相关。集中在CC的多参数放射组学模型显示其对精神分裂症诊断的鲁棒性。放射组学特征可能是支持基于生物标志物的精神分裂症诊断的生物标志物的潜在来源，并提高对其神经生物学的理解。
There is a growing need to develop novel strategies for the diagnosis of schizophrenia using neuroimaging biomarkers. We investigated the robustness of the diagnostic model for schizophrenia using radiomic features from T1-weighted and diffusion tensor images of the corpus callosum (CC). A total of 165 participants [86 schizophrenia and 79 healthy controls (HCs)] were allocated to training (N = 115) and test (N = 50) sets. Radiomic features of the CC subregions were extracted from T1-weighted, apparent diffusion coefficient (ADC), and fractional anisotropy (FA) images (N = 1605). Following feature selection, various combinations of classifiers were trained, and Bayesian optimization was adopted in the best performing classifier. Discrimination, calibration, and clinical utility of the model were assessed. An online calculator was constructed to offer the probability of having schizophrenia. SHapley Additive exPlanations (SHAP) was applied to explore the interpretability of the model. We identified 30 radiomic features to differentiate participants with schizophrenia from HCs. The Bayesian optimized model achieved the highest performance, with an area under the curve (AUC), accuracy, sensitivity, and specificity of 0.89 (95% confidence interval: 0.81–0.98), 80.0, 83.3, and 76.9%, respectively, in the test set. The final model offers clinical probability in an online calculator. The model explanation by SHAP suggested that second-order features from the posterior CC were highly associated with the risk of schizophrenia. The multiparametric radiomics model focusing on the CC shows its robustness for the diagnosis of schizophrenia. Radiomic features could be a potential source of biomarkers that support the biomarker-based diagnosis of schizophrenia and improve the understanding of its neurobiology.
登录
查看更多内容
影响因子：
4.5
作者：
Goghari, VM;Lang, DJ;Honer, WG
通讯作者：
Honer, WG
影响因子：
11
作者：
Kelly S;Jahanshad N;Zalesky A;Kochunov P;Agartz I;Alloza C;Andreassen OA;Arango C;Banaj N;Bouix S;Bousman CA;Brouwer RM;Bruggemann J;Bustillo J;Cahn W;Calhoun V;Cannon D;Carr V;Catts S;Chen J;Chen JX;Chen X;Chiapponi C;Cho KK;Ciullo V;Corvin AS;Crespo-Facorro B;Cropley V;De Rossi P;Diaz-Caneja CM;Dickie EW;Ehrlich S;Fan FM;Faskowitz J;Fatouros-Bergman H;Flyckt L;Ford JM;Fouche JP;Fukunaga M;Gill M;Glahn DC;Gollub R;Goudzwaard ED;Guo H;Gur RE;Gur RC;Gurholt TP;Hashimoto R;Hatton SN;Henskens FA;Hibar DP;Hickie IB;Hong LE;Horacek J;Howells FM;Hulshoff Pol HE;Hyde CL;Isaev D;Jablensky A;Jansen PR;Janssen J;Jönsson EG;Jung LA;Kahn RS;Kikinis Z;Liu K;Klauser P;Knöchel C;Kubicki M;Lagopoulos J;Langen C;Lawrie S;Lenroot RK;Lim KO;Lopez-Jaramillo C;Lyall A;Magnotta V;Mandl RCW;Mathalon DH;McCarley RW;McCarthy-Jones S;McDonald C;McEwen S;McIntosh A;Melicher T;Mesholam-Gately RI;Michie PT;Mowry B;Mueller BA;Newell DT;O'Donnell P;Oertel-Knöchel V;Oestreich L;Paciga SA;Pantelis C;Pasternak O;Pearlson G;Pellicano GR;Pereira A;Pineda Zapata J;Piras F;Potkin SG;Preda A;Rasser PE;Roalf DR;Roiz R;Roos A;Rotenberg D;Satterthwaite TD;Savadjiev P;Schall U;Scott RJ;Seal ML;Seidman LJ;Shannon Weickert C;Whelan CD;Shenton ME;Kwon JS;Spalletta G;Spaniel F;Sprooten E;Stäblein M;Stein DJ;Sundram S;Tan Y;Tan S;Tang S;Temmingh HS;Westlye LT;Tønnesen S;Tordesillas-Gutierrez D;Doan NT;Vaidya J;van Haren NEM;Vargas CD;Vecchio D;Velakoulis D;Voineskos A;Voyvodic JQ;Wang Z;Wan P;Wei D;Weickert TW;Whalley H;White T;Whitford TJ;Wojcik JD;Xiang H;Xie Z;Yamamori H;Yang F;Yao N;Zhang G;Zhao J;van Erp TGM;Turner J;Thompson PM;Donohoe G
通讯作者：
Donohoe G
影响因子：
5.7
作者：
Alexander, Andrew L.;Lee, Jee Eun;Lainhart, Janet E.
通讯作者：
Lainhart, Janet E.
影响因子：
--
作者：
Pierri, JN;Volk, CLE;Lewis, DA
通讯作者：
Lewis, DA
影响因子：
11
作者：
Flynn, SW;Lang, DJ;Honer, WG
通讯作者：
Honer, WG