Winners of the Second
Mental Health Innovation Prize

Winners of the Second Mental Health Innovation Prize

In Partnership with Mental Health Research Canada (MHRC) 

IAM-MHRC 2022 Mental Health Innovation Prize: Application of Artificial Intelligence in Mental Health

We are proud to announce the innovation initiatives awarded the Mental Health Innovation Prize in partnership with
Mental Health Research Canada (MHRC).

Each initiative will be awarded $50,000 to support the development of predictive models and interventions utilizing AI, Machine Learning, and/or Big Data to better understand and align supports and treatments in the mental illness space. We jointly and proudly announce this Mental Health Innovation national prize in support of mental health innovation in Canada.

Winners of the second Mental Health Innovation Prize:

noun-ai-5907046.png       Predicting opioid overdose risk using machine learning:
       Evaluation of an AI approach.

           Special thanks to ATB Financial  for supporting this project based in Alberta.

         Lead Investigators: Giri Puligandla, Canadian Mental Health Association – Edmonton Region
                                          and Dr. Bo Cao, University of Alberta

Opioid misuse is a major health crisis in need of solutions. In Canada, opioid overdose is related to over 16,000 deaths since 2016.

We are building on an existing model for accurate prediction of opioid overdose and risk factor identification using large-scale health data and machine learning. The potential use-cases for the application of the model are easy to imagine: outreach and emergency services could apply targeted monitoring approaches, service providers could intervene proactively with high-risk individuals, and governments could strategically allocate prevention resources. Nonetheless, there are significant risks such as over-surveillance, stigmatization, and loss of personal autonomy on the part of people with living experience.

Service providers, institutional knowledge users (like government and police), and people with lived/living experience will be engaged to guide the improvement of the predictive model and contribute real-life insights into the potential risks and benefits of its application. This project is ground-breaking in that it engages communities and stakeholders early in developing new technologies rather than imposing tools on them after the fact.


noun-digital-phone-2123653.png Using digital phenotyping measures to predict the symptoms and functional      
outcomes in first episode of psychosis.     

Lead Investigators: Dr. JianLi Wang, Department of Community Health and Epidemiology,      
Dalhousie University and Dr. Phil Tibbo, Professor in the Department of Psychiatry, Dalhousie University       
and Clinical Director of the Early Psychosis Intervention Nova Scotia (EPINS) (Nova Scotia Health)      

Psychotic disorders, including schizophrenia, are severe mental disorders affecting 2-3% of the population and rank among the leading causes of disability worldwide. Although early intervention is effective in improving illness outcome, a significant proportion of first-episode psychosis (FEP) patients experience persistent functional impairment even after clinical remission.

Accurate prediction of FEP patient trajectories will allow clinicians to select better interventions at the beginning, leading to better patient outcomes and quality of life. However, predicting FEP patient outcomes is challenging because assessments of the clinical and behavioral factors are often based on patient self-report, which is vulnerable to recall and reporting biases. The biases can reduce the accuracy of outcome prediction.

Digital phenotyping, which refers to the use of mobile devices (e.g., smartphone, wearable) to initiate data collection in everyday life, has great potential to address these issues.

The goal of this project is to develop and implement prediction models for symptom and functional outcomes in FEP, using both self-reported and digital phenotyping data. If successful, the prediction models will greatly enhance clinicians’ capability of outcome prediction and decision-making, leading to better patient outcomes and quality of life.