Scientists in London have created new artificial intelligence (AI) technologies inspired by video games that will revolutionize the way we diagnose and manage depression. Patients and professionals are actively testing Thymia, a platform that aims to make depression and other mental health diseases as measurable as physical ailments.
Thymia, the brainchild of neuroscience Dr. Emilia Molimpakis and theoretical physicist Dr. Stefano Goria, will make mental health examinations objective. Patients will play specially created video games on the platform, which employs linguistics, neuropsychology, and machine learning to detect signs of depression and monitor whether patients are responding to treatment, rather than filling out questionnaires.
By having a mental illness as objectively measurable as physical health disorders, the scientists behind the technology want to empower physicians to make faster and more precise treatment judgments.
Dr Emilia Molimpakis’ close friend had depression, prompting the creation of Thymia. Traditional depression diagnostic procedures failed to adequately communicate the degree of her pain to her clinicians, leading to a suicide attempt. Dr. Molimpakis used her knowledge of linguistics, cognitive neuroscience, and experimental psychology to create a platform that could potentially enhance and replace the extremely subjective questionnaire-based approach to diagnosing depression.
Thymia gives clinicians a more efficient and accurate way to identify and monitor serious depression.
Dr. Emilia Molimpakis and Dr. Stefano Goria, an expert inexplainable, multi-modal AI, have devised smart video game-style activities for patients to play with that test for a variety of depressed symptoms. Verbally describing animated sights and memorizing moving items such as bees are among the activities. While consumers are playing the games, the Thymia program analyzes three major data streams anonymously:
- Voice: both the manner in which a person talks and the content of their words (to pick up acoustic and linguistic depressive cues)
- Video: eye-gaze and micro expressions (which can help track current mood)
- Reaction times, memory, and error rates are examples of behavioral metrics (which can help detect depression severity).
To help determine a diagnosis more quickly and correctly, the software identifies data patterns suggestive of depression. It will demonstrate whether or not any treatments (therapies or drugs) are effective over time.
In the weeks between in-person appointments, practitioners can use the platform to continually and remotely monitor patients at home. Doctors and patients will be able to have a better grasp of their condition over time as a result of this.
To ethically and inclusively train its AI, Thymia is collecting data from thousands of people with severe depressive illness and a normative control group. To make AI models as accurate and impartial as possible is reducing the racial, gender, and age biases that are commonly associated with them.
