Clinical Grade Prediction of Therapeutic Dosage for Electroconvulsive Therapy (ECT) Based on Patient’s Pre-Ictal EEG Using Fuzzy Causal Transformers

Depression and Major Depressive Disorder (MDD) can hinder individuals from a normal personal and social life. That is, it can cause suicide, drug abuse, and divorce. MDD also imposes direct and indirect economic costs on families and governments. However, there is no methodology to know how and how long a healthcare professional needs to administer an electroconvulsive pulse to treat patients with MDD to achieve an adequate seizure response. In this study, for the first time using MDD patients’ pre-ictal EEG data, age, and sex with Causal Deep Learning algorithms, we successfully predict the individualized therapeutic ECT dose and its duration with 93% accuracy. Compared to other previous methods which used Magnetic resonance imaging (MRI) with machine learning algorithms, our method is faster, cheaper, and more concise.