Machine learning techniques in a structural and functional MRI diagnostic approach in schizophrenia: a systematic review

• Published in: Neuropsychiatric disease and treatment Vol. 15; pp. 1605 - 1627
• Main Authors: de Filippis, Renato, Carbone, Elvira Anna, Gaetano, Raffaele, Bruni, Antonella, Pugliese, Valentina, Segura-Garcia, Cristina, De Fazio, Pasquale
• Format: Journal Article
• Language: English
• Published: New Zealand Dove Medical Press Limited 01-06-2019; Taylor & Francis Ltd; Dove; Dove Medical Press
• Subjects: Algorithms; Artificial intelligence; Biological markers; Brain; Machine learning; Magnetic resonance imaging; Medical diagnosis; Medical research; Mental disorders; multivariate pattern analysis; NMR; Nuclear magnetic resonance; resting-state fMRI; Review; Schizophrenia; sMRI; support vector machine; Systematic review; Italy; support vector machine; schizophrenia; sMRI; machine learning; resting-state fMRI; multivariate pattern analysis
• ISSN: 1176-6328; 1178-2021; 1178-2021
• DOI: 10.2147/NDT.S202418

Diagnosis of schizophrenia (SCZ) is made exclusively clinically, since specific biomarkers that can predict the disease accurately remain unknown. Machine learning (ML) represents a promising approach that could support clinicians in the diagnosis of mental disorders.
A systematic review, according to the PRISMA statement, was conducted to evaluate its accuracy to distinguish SCZ patients from healthy controls.
We systematically searched PubMed, Embase, MEDLINE, PsychINFO and the Cochrane Library through December 2018 using generic terms for ML techniques and SCZ without language or time restriction. Thirty-five studies were included in this review: eight of them used structural neuroimaging, twenty-six used functional neuroimaging and one both, with a minimum accuracy >60% (most of them 75-90%). Sensitivity, Specificity and accuracy were extracted from each publication or obtained directly from authors.
Support vector machine, the most frequent technique, if associated with other ML techniques achieved accuracy close to 100%. The prefrontal and temporal cortices appeared to be the most useful brain regions for the diagnosis of SCZ. ML analysis can efficiently detect significantly altered brain connectivity in patients with SCZ (eg, default mode network, visual network, sensorimotor network, frontoparietal network and salience network).
The greater accuracy demonstrated by these predictive models and the new models resulting from the integration of multiple ML techniques will be increasingly decisive for early diagnosis and evaluation of the treatment response and to establish the prognosis of patients with SCZ. To achieve a real benefit for patients, the future challenge is to reach an accurate diagnosis not only through clinical evaluation but also with the aid of ML algorithms.