AI Model MANAS-1 Detects Mental Health Disorders via Brain Waves

Introduction

Indian researchers have developed an innovative AI system named MANAS-1 that interprets brain waves to identify neurological and psychiatric disorders. This model utilizes standard electroencephalogram (EEG) equipment to analyze electrical signals from the brain.

How MANAS-1 Works

The MANAS-1 model processes brain wave data as a structured language that artificial intelligence can learn to interpret. According to Dr. Puneet Agarwal, the neurologist behind this technology and a professor at Max Super Speciality Hospital in New Delhi, the system was trained on an extensive dataset comprising 60,000 hours of brain wave data from over 25,000 patients globally.

EEG Technology

MANAS-1 employs EEG technology to capture the electrical activity of the brain. The AI analyzes these brain wave patterns to identify distinct signatures associated with various neurological and psychiatric conditions. Currently, the system operates with 400 million parameters—variables it utilizes to make decisions based on training data. Plans are in place to expand this to 2 billion parameters to enhance accuracy, with a second version, MANAS-2, expected to be released in the coming weeks.

Dr. Agarwal emphasizes that the model can interpret the connections and communications between different brain regions, enhancing its diagnostic capabilities.

Importance of the Technology

This technology has the potential to provide faster and more cost-effective diagnoses of brain disorders. Traditional neurological testing often requires expensive equipment and specialist interpretation, whereas MANAS-1 can function with standard EEG machines commonly found in hospitals.

Dr. Agarwal reports that the model achieves an impressive 95% accuracy in identifying computational markers for epilepsy and mental health disorders. If validated independently, this could significantly improve access to brain health screenings in areas with a shortage of specialist doctors.

Broader Applications

Beyond diagnosis, the MANAS-1 system holds promise for applications in cognitive assessment and the development of brain-computer interfaces. Such interfaces could enable paralyzed patients to control devices using their thoughts, showcasing the model's versatility.

Context in AI Development

MANAS-1 is part of a growing field of AI tools designed to analyze brain activity. For instance, researchers in Singapore developed Brain-JEPA, which also identifies brain regions and their functions to assist in diagnosing disorders and predicting disease progression. Additionally, South Korean scientists have created an AI model capable of predicting dementia risk based on voice recordings from elderly patients.

In India, Google has been actively supporting healthcare AI development, partnering with institutions like the All India Institute of Medical Sciences and the Indian Institute of Science to create AI tools for diagnosing skin conditions and sorting hospital patients. Earlier this year, Google also released MedGemma 1.5, an open-source medical AI model in India.

Conclusion

The development of the MANAS-1 AI model represents a significant advancement in the field of mental health diagnostics, potentially transforming how brain disorders are identified and treated.

Source: Indian AI model reads brain waves to spot mental health disorders - DHArab