NIMHANS AI Depression Screening Project: HEADS Details

The National Institute of Mental Health and Neuro Sciences (NIMHANS), IIT Kharagpur and Lokopriya Gopinath Bordoloi Regional Institute of Mental Health (LGBRIMH), Tezpur have launched a two-year research project to develop an AI-assisted depression screening system for Indian languages. The project, named Human-in-the-loop Evaluation of Assisted Depression Screening (HEADS), will focus on Kannada, Hindi, Bengali, Assamese and English.

The AI tool for depression screening is being designed to assist clinicians by analysing clinical conversations, rather than independently diagnosing patients. The project is funded by the Wellcome Charitable Foundation and will involve about 4,500 participants, including 4,000 patients and 500 healthy volunteers.

HEADS Project to Develop AI Depression Screening System

The HEADS project aims to address a specific challenge in mental health screening in India: patients may describe psychological distress using regional languages, idioms or mixed-language expressions that existing AI systems may not interpret accurately.

The proposed system will listen to clinical interviews, convert speech into text, translate conversations where required and generate a clinical summary along with a depression assessment and severity estimate. It will also provide reasoning and a confidence score for clinicians to review.

The project is being developed as a human-in-the-loop AI system, meaning the final clinical decision will remain with a qualified clinician.

AI Depression Screening Project: Languages and Participants

The multilingual design is one of the key features of the NIMHANS AI depression screening project. Clinical data will be collected by NIMHANS and LGBRIMH across five languages.

The following table summarises the major details of the HEADS project:

NIMHANS will collect clinical data in Kannada, Hindi and English, while LGBRIMH will collect data in Assamese, Bengali and English. IIT Kharagpur will lead the AI engineering, model development and testing.

IIT Kharagpur to Lead AI Engineering and Safety Testing

Under the HEADS AI project, IIT Kharagpur will be responsible for developing and testing the artificial intelligence models. The institute will also examine potential AI bias and safety concerns.

The research team plans to assess whether the AI system's depression screening results are affected by factors such as language, gender, educational background or cultural context. This is particularly important for an AI model being trained on multilingual clinical conversations.

The Wellcome project description says the system will use separate modules for automatic speech recognition, Indic-to-English translation and clinical inference. It is intended to support depression assessment using clinical frameworks including DSM-5 and PHQ-9.

NIMHANS AI Depression Screening Tool Will Not Replace Doctors

A key aspect of the AI-based depression screening tool is that it will remain clinician-facing. The AI-generated assessment will be reviewed and corrected by clinicians, who will retain final decision-making authority.

This human oversight is intended to reduce the risk of overdiagnosis and ensure that AI-generated outputs are considered alongside clinical judgment. The system is being developed to assist screening and earlier recognition of depression rather than provide an independent diagnosis directly to the public.

The project could eventually support screening by non-specialist health workers, particularly where access to specialist mental-health professionals is limited. However, the current research does not establish that the tool will be deployed as a public diagnostic service.

Why Indian-Language AI Matters for Depression Screening

The institutions said existing speech recognition systems, language models and privacy tools have limitations when applied to clinical use in Indian languages. Preliminary work reported by the researchers also identified problems with transcription accuracy and protection of personal identifiers.

The HEADS project therefore focuses not only on AI for mental health, but also on how language and cultural context affect clinical interpretation. Its five-language study will provide data for evaluating whether AI can recognise expressions of psychological distress without losing important linguistic or cultural meaning.

Two MoUs were signed as part of the project launch—one between NIMHANS and LGBRIMH and another between NIMHANS and IIT Kharagpur.

NIMHANS AI Depression Project FAQs