aifred health is all about helping clinicians predict which treatments will get the best results with the fewest side effects, so that clinicians and patients can have more information when deciding on a treatment plan together. Our solution is a clinical decision aid that can be used by clinicians to input patient features and output personalized lists of treatments that include predicted efficacy and side-effect profile.
We are also a proud official IBM Watson AI XPRIZE team, headquartered in Montreal, Canada.
Depression is a serious mental illness that globally affects more than 300 million people. Some patients can spend years finding the right treatment regimen for them, and many patients won’t improve after the first treatment. The inability to predict any given individual’s unique response to medical treatment is a huge bottleneck to recovery.
To address this challenge, we are using a deep learning approach to build a predictive model based on the best available data.
Initially, we will be focusing on treatments for depression, but we plan to scale aifred to encompass all mental health conditions in order to amplify clinical utility. At its core, aifred is leveraging the collective intelligence of the scientific and medical community to bring better healthcare to all.
Something unique to every machine learning company is the precise nature of their hyperparameter optimization and goals of their model. We will optimize aifred with the help of a distributed network of domain experts in psychiatry--a collaboration unique to aifred health. We are implementing attention networks responsible for removing the “black-box” nature of neural networks. As well, we are analysing the quality of model predictions, allowing both for greater interpretability of model decisions and the generation of new basic research questions, which are going to be unique to the data-set and optimization techniques we develop in-house. By training aifred on reliable datasets, we are able to ensure quality input to our model. De-identified patient outcomes will feed back into our neural networks to continuously improve aifred’s predictive power. Feature engineering is an important part of determining which inputs go into a network and varies how it’s done for every team- once again, this will be undertaken with the support of diverse group of experts we are recruiting.
At aifred health, we strongly believe in the potential for artificial intelligence to augment medical practice and support (but not replace) clinical decision-making. We aim to ease the integration of such technology into medical communities. Following this principle, every product feature is developed to meet physicians’ wants and needs. The aifred solution combines innovative and powerful machine learning techniques with a friendly user experience.
We are building several features that clinicians will find helpful: banks of standardized questionnaires, the ability to visualize patient symptoms over time, scheduling software- all of it modular and capable of being tailored to clinicians needs.
Most importantly, our product is being developed with both patient and clinician advisors so that it can best meet their needs.
In the diagram you can see how a clinician-patient interaction leads to clinician-derived data (as well as data from patient self-report) being inputted into the machine learning model, which will then use its training on high-quality data to predict a list of the most effective treatments for that given patient, with personalized side-effect profiles. Clinicians and patients can then decide what treatment to try together, and patient outcomes can then be fed back into the model to generate new predictions - and continuously improve the model itself. In addition, insights from the model can help generate new questions for research.
Our research team is conducting a series of systematic literature reviews to curate predictors of treatment response and side effect burden in depression. We are evaluating the state of precision psychiatry in domains including genetics, endocrinology, immunology, metabolic biochemistry, and neuroimaging, as well as examining the feasibility of including biomarker testing in routine clinical practice. The results of these reviews will serve to validate our model and inform the input feature space by integrating these multimodal biomarkers along with sociodemographic and clinical factors.
Clinical research is focused on validating our model in controlled and real-world conditions. We are designing three kinds of research trials:
Ease-of-use: The model must be user-friendly and provide clinicians with features they want and need, so we must study the aifred solution’s integration into clinical workflow and any effects on clinician efficiency and doctor-patient interaction.
Open label trials: Safety and effectiveness of the model must be assessed in open-label trials where both clinicians and patients know when our model is being used. A group of physicians using our model will be compared to a group practicing usual care, and patient outcomes will be compared between the two.
Randomized control trial: After open label studies, we will conduct one or more randomized control trials, testing our model against a “dummy” model and against a “practice as usual” group. This will help us determine how efficacious the aifred solution is.
Safety is critical, so our clinical team, which includes two physicians, will be making sure to review our model’s predictions and ensure hard-coded safety features so that model treatment recommendations are safe.
We are blazing the trail when it comes to clinical validation of deep-learning based clinical decision aids, and as such are investing heavily in the development of ethical principles to guide development and testing. In fact ethical development is so important to us that we have created our own ethical framework, known as Meticulous Transparency, to guide our work. We also never store identifiable patient information, to protect patient privacy. Contact us to find out more!
Dr. Gustavo Turecki, MD, PhD - Genetics, Dataset Access Content Expert
Dr. Marc Miresco, MD, MSc - External Psychiatry Services Content Expert
Dr. Leon Tourian, MD - Medical Education Content Expert
Dr. Thomas Milroy, MD - Electroconvulsive Therapy Content Expert
Dr. Gail Myhr, MDCM, Dip Psy, MSc, FRCP - Cognitive Behavioral Therapy Content Expert
Dr. Eduardo Chachamovitch, MD, PhD - Psychometrics, Mood disorders and Special Populations Content Expert
Dr. Simon Ducharme, MD, MSc, FRCPC - Neuroimaging Content Expert
Dr. Marcelo Berlim, MD, MSc - Literature Review and Neuromodulation Content Expert
Dr. Howard Margolese, MD - Clinical Trial Expert
Dr. Wendell Wallach - Bioethics and AI Ethics Content Expert
Tristan Sylvain - Machine Learning Content Expert
Dr. Margaux Luck - Machine Learning Content Expert
Dr. Daniel Blumberger, MD, MSc, FRCPC - rTMS and Neuromodulation Content Expert
Dr. Sagar Parikh, MD - Guidelines, Best Practices, Cultural Safety Content Expert
Dr. Simone Vigod, MD, MSc, FRCPC - Special Populations (Women and Perinatal) Content Expert
Dr. Anthony J. Levitt, MD, MBBS, FCPC - Depression Treatment Optimization Content Expert
Dr. Roger S. McIntyre, MD, FRCPC - Psychiatry and Neuropharmacology Content Expert
Dr. Jonathan Roiser, PhD - Computational Psychiatry Content Expert