Treatment of depression using machine learning

A computer-implemented method, including:

• training, based at least on training data, a machine learning model,
• the training data including a first data corresponding to one or more brain signals of a subject, the training data further including a treatment outcome associated with the subject,
• the machine learning model including a filter configured to generate, based at least on the first data, a first plurality of latent signals, the machine learning model further including a regression model configured to generate, based at least on a feature of each of the first plurality of latent signals, a first treatment outcome prediction for the subject,
• the training of the machine learning model including optimizing the filter to generate a reduced quantity of latent signals whose feature minimizes an error in the first treatment outcome prediction generated by the regression model, the error comprising a difference between the treatment outcome of the subject and the first treatment outcome prediction generated by the regression model;
• and
• deploying the trained machine learning model.

A non-transitory computer readable medium storing instructions, which, when executed by at least one data processor, result in performance of operations including:

• training, based at least on training data, a machine learning model,
• the training data including a first data corresponding to one or more brain signals of a subject, the training data further including a treatment outcome associated with the subject,
• the machine learning model including a filter configured to generate, based at least on the first data, a first plurality of latent signals, the machine learning model further including a regression model configured to generate, based at least on a feature of each of the first plurality of latent signals, a first treatment outcome prediction for the subject,
• the training of the machine learning model including optimizing the filter to generate a reduced quantity of latent signals whose feature minimizes an error in the first treatment outcome prediction generated by the regression model, the error comprising a difference between the treatment outcome of the subject and the first treatment outcome prediction generated by the regression model;
• and
• deploying the trained machine learning model.

A system comprising:

• one or more processors;
• and
• one or more computer readable media comprising computer readable code executable by the one or more processors to perform operations including:
• training, based at least on training data, a machine learning model,
• the training data including a first data corresponding to one or more brain signals of a subject, the training data further including a treatment outcome associated with the subject,
• the machine learning model including a filter configured to generate, based at least on the first data, a first plurality of latent signals, the machine learning model further including a regression model configured to generate, based at least on a feature of each of the first plurality of latent signals, a first treatment outcome prediction for the subject,
• the training of the machine learning model including optimizing the filter to generate a reduced quantity of latent signals whose feature minimizes an error in the first treatment outcome prediction generated by the regression model, the error comprising a difference between the treatment outcome of the subject and the first treatment outcome prediction generated by the regression model;
• and
• deploying the trained machine learning model.