White matter diffusion estimates in obsessive-compulsive disorder across 1653 individuals: machine learning findings from the ENIGMA OCD Working Group.
Kim B-G., Kim G., Abe Y., Alonso P., Ameis S., Anticevic A., Arnold PD., Balachander S., Banaj N., Bargalló N., Batistuzzo MC., Benedetti F., Bertolín S., Beucke JC., Bollettini I., Brem S., Brennan BP., Buitelaar JK., Calvo R., Castelo-Branco M., Cheng Y., Chhatkuli RB., Ciullo V., Coelho A., Couto B., Dallaspezia S., Ely BA., Ferreira S., Fontaine M., Fouche J-P., Grazioplene R., Gruner P., Hagen K., Hansen B., Hanna GL., Hirano Y., Höxter MQ., Hough M., Hu H., Huyser C., Ikuta T., Jahanshad N., James A., Jaspers-Fayer F., Kasprzak S., Kathmann N., Kaufmann C., Kim M., Koch K., Kvale G., Kwon JS., Lazaro L., Lee J., Lochner C., Lu J., Manrique DR., Martínez-Zalacaín I., Masuda Y., Matsumoto K., Maziero MP., Menchón JM., Minuzzi L., Moreira PS., Morgado P., Narayanaswamy JC., Narumoto J., Ortiz AE., Ota J., Pariente JC., Perriello C., Picó-Pérez M., Pittenger C., Poletti S., Real E., Reddy YCJ., van Rooij D., Sakai Y., Sato JR., Segalas C., Shavitt RG., Shen Z., Shimizu E., Shivakumar V., Soriano-Mas C., Sousa N., Sousa MM., Spalletta G., Stern ER., Stewart SE., Szeszko PR., Thomas R., Thomopoulos SI., Vecchio D., Venkatasubramanian G., Vriend C., Walitza S., Wang Z., Watanabe A., Wolters L., Xu J., Yamada K., Yun J-Y., Zarei M., Zhao Q., Zhu X., ENIGMA-OCD Working Group None., Thompson PM., Bruin WB., van Wingen GA., Piras F., Piras F., Stein DJ., van den Heuvel OA., Simpson HB., Marsh R., Cha J.
White matter pathways, typically studied with diffusion tensor imaging (DTI), have been implicated in the neurobiology of obsessive-compulsive disorder (OCD). However, due to limited sample sizes and the predominance of single-site studies, the generalizability of OCD classification based on diffusion white matter estimates remains unclear. Here, we tested classification accuracy using the largest OCD DTI dataset to date, involving 1336 adult participants (690 OCD patients and 646 healthy controls) and 317 pediatric participants (175 OCD patients and 142 healthy controls) from 18 international sites within the ENIGMA OCD Working Group. We used an automatic machine learning pipeline (with feature engineering and selection, and model optimization) and examined the cross-site generalizability of the OCD classification models using leave-one-site-out cross-validation. Our models showed low-to-moderate accuracy in classifying (1) "OCD vs. healthy controls" (Adults, receiver operator characteristic-area under the curve = 57.19 ± 3.47 in the replication set; Children, 59.8 ± 7.39), (2) "unmedicated OCD vs. healthy controls" (Adults, 62.67 ± 3.84; Children, 48.51 ± 10.14), and (3) "medicated OCD vs. unmedicated OCD" (Adults, 76.72 ± 3.97; Children, 72.45 ± 8.87). There was significant site variability in model performance (cross-validated ROC AUC ranges 51.6-79.1 in adults; 35.9-63.2 in children). Machine learning interpretation showed that diffusivity measures of the corpus callosum, internal capsule, and posterior thalamic radiation contributed to the classification of OCD from HC. The classification performance appeared greater than the model trained on grey matter morphometry in the prior ENIGMA OCD study (our study includes subsamples from the morphometry study). Taken together, this study points to the meaningful multivariate patterns of white matter features relevant to the neurobiology of OCD, but with low-to-moderate classification accuracy. The OCD classification performance may be constrained by site variability and medication effects on the white matter integrity, indicating room for improvement for future research.