Unknown

Dataset Information

0

Multi-site benchmark classification of major depressive disorder using machine learning on cortical and subcortical measures.


ABSTRACT: Machine learning (ML) techniques have gained popularity in the neuroimaging field due to their potential for classifying neuropsychiatric disorders. However, the diagnostic predictive power of the existing algorithms has been limited by small sample sizes, lack of representativeness, data leakage, and/or overfitting. Here, we overcome these limitations with the largest multi-site sample size to date (N = 5365) to provide a generalizable ML classification benchmark of major depressive disorder (MDD) using shallow linear and non-linear models. Leveraging brain measures from standardized ENIGMA analysis pipelines in FreeSurfer, we were able to classify MDD versus healthy controls (HC) with a balanced accuracy of around 62%. But after harmonizing the data, e.g., using ComBat, the balanced accuracy dropped to approximately 52%. Accuracy results close to random chance levels were also observed in stratified groups according to age of onset, antidepressant use, number of episodes and sex. Future studies incorporating higher dimensional brain imaging/phenotype features, and/or using more advanced machine and deep learning methods may yield more encouraging prospects.

SUBMITTER: Belov V 

PROVIDER: S-EPMC10784593 | biostudies-literature | 2024 Jan

REPOSITORIES: biostudies-literature

altmetric image

Publications

Multi-site benchmark classification of major depressive disorder using machine learning on cortical and subcortical measures.

Belov Vladimir V   Erwin-Grabner Tracy T   Aghajani Moji M   Aleman Andre A   Amod Alyssa R AR   Basgoze Zeynep Z   Benedetti Francesco F   Besteher Bianca B   Bülow Robin R   Ching Christopher R K CRK   Connolly Colm G CG   Cullen Kathryn K   Davey Christopher G CG   Dima Danai D   Dols Annemiek A   Evans Jennifer W JW   Fu Cynthia H Y CHY   Gonul Ali Saffet AS   Gotlib Ian H IH   Grabe Hans J HJ   Groenewold Nynke N   Hamilton J Paul JP   Harrison Ben J BJ   Ho Tiffany C TC   Mwangi Benson B   Jaworska Natalia N   Jahanshad Neda N   Klimes-Dougan Bonnie B   Koopowitz Sheri-Michelle SM   Lancaster Thomas T   Li Meng M   Linden David E J DEJ   MacMaster Frank P FP   Mehler David M A DMA   Melloni Elisa E   Mueller Bryon A BA   Ojha Amar A   Oudega Mardien L ML   Penninx Brenda W J H BWJH   Poletti Sara S   Pomarol-Clotet Edith E   Portella Maria J MJ   Pozzi Elena E   Reneman Liesbeth L   Sacchet Matthew D MD   Sämann Philipp G PG   Schrantee Anouk A   Sim Kang K   Soares Jair C JC   Stein Dan J DJ   Thomopoulos Sophia I SI   Uyar-Demir Aslihan A   van der Wee Nic J A NJA   van der Werff Steven J A SJA   Völzke Henry H   Whittle Sarah S   Wittfeld Katharina K   Wright Margaret J MJ   Wu Mon-Ju MJ   Yang Tony T TT   Zarate Carlos C   Veltman Dick J DJ   Schmaal Lianne L   Thompson Paul M PM   Goya-Maldonado Roberto R  

Scientific reports 20240111 1


Machine learning (ML) techniques have gained popularity in the neuroimaging field due to their potential for classifying neuropsychiatric disorders. However, the diagnostic predictive power of the existing algorithms has been limited by small sample sizes, lack of representativeness, data leakage, and/or overfitting. Here, we overcome these limitations with the largest multi-site sample size to date (N = 5365) to provide a generalizable ML classification benchmark of major depressive disorder (M  ...[more]

Similar Datasets

| S-EPMC11887997 | biostudies-literature
| S-EPMC4879183 | biostudies-literature
| S-EPMC8675412 | biostudies-literature
| S-EPMC9486057 | biostudies-literature
| S-EPMC6812119 | biostudies-literature
| S-EPMC9062667 | biostudies-literature
2011-10-04 | E-GEOD-19738 | biostudies-arrayexpress
| S-EPMC11442283 | biostudies-literature
| S-EPMC4047134 | biostudies-other
| S-EPMC2993928 | biostudies-literature