Smartwatches, Earbuds, and AI: How Technology Could Change Mental Health Support
A virtual assistant capable of understanding when a person is going through a moment of emotional difficulty without waiting for them to ask for support. This is the goal of UbiMyTherapist, a project developed by researchers at the University of Ottawa that aims to leverage multimodal artificial intelligence and data collected from wearable devices to offer personalized psychological support even outside clinical environments.
Unlike traditional mental health chatbots that require explicit user interaction, UbiMyTherapist is designed to operate in both reactive mode, responding to user requests, and proactive mode, detecting signals of possible distress and suggesting interventions before a help request is formulated.
The system is compatible with consumer devices such as smartwatches, smartphones, and earbuds, from which it gathers useful information to estimate the emotional state of the individual. The parameters taken into consideration include heart rate variability, voice tone characteristics, and even the content of written messages, combined to obtain a contextual assessment of the user’s psychological state.
One of the distinctive elements of the platform is the creation of a so-called "digital twin," which is a dynamic digital profile that is constantly updated. This model integrates three different categories of information: the patient's clinical history, including any diagnoses and medications, a knowledge base derived from clinical psychology literature, and emotionally relevant data collected in real-time from sensors and user interactions.
The combination of this information allows the language model to generate responses that are not only contextually relevant but also more personalized and clinically based, surpassing the more generic approach typical of common LLM-based chatbots.
According to Karim Alghoul, a professor at the School of Electrical Engineering and Computer Science at the University of Ottawa and the study's lead author, the project's goal is to provide timely, secure, and personalized support for mental well-being, even when the patient is far from the specialists' offices.
The prototype has undergone a preliminary evaluation in its reactive mode involving 24 participants. At the same time, licensed psychotherapists have assessed the therapeutic quality of the responses generated by the system. According to the results presented by the researchers, UbiMyTherapist received particularly positive evaluations regarding empathy and personalization compared to traditional configurations based on large language models like ChatGPT.
However, the researchers emphasize that the project is not intended to replace psychotherapists. The goal is rather to extend mental health support beyond the confines of healthcare facilities, offering an additional tool especially to those facing economic, geographic, or social barriers to accessing care.
The next stages of development include integrating genuinely proactive interventions activated almost in real-time by biometric data from smartwatches and a broader collaboration with qualified psychotherapists to keep the system aligned with clinical practices. For now, UbiMyTherapist remains a research project, described in a study published in IEEE, and is not yet intended for distribution as a commercial application.