University of Toronto hosted a lunch in honour of the 2020 RBC Post-Doctoral & Graduate Fellowship winners as part of U of T Entrepreneurship Week.
The awards are independent of the Borealis AI Graduate Fellowships, which will be announced in the coming weeks.
Entrants of the RBC Post-Doctoral & Graduate Fellowships must meet two key critieria: potential interest in commercialization, knowledge translation and company creation in addition to an outstanding record of scholarship; and they must attend U of T.
The two graduate fellows and two post-doctoral fellows represent a number of disciplines across the university. All winning entries focused on addressing healthcare issues including senior care, dementia, patient-centered healthcare and regenerative therapy.
Learn more about their winning research proposals below and follow the Borealis AI Twitter feed for highlights and pictures from the awards ceremony.
2020 RBC Graduate Fellows
PhD candidate in Biomedical Engineering at the UofT and Toronto Rehab Institute.
Research topic: Using IoT sensors and AI to assess and predict frailty syndrome for older adults in home settings.
Research interests: Chao co-founded Nightingale.ai, a digital health startup that uses AI to help clinicians automate functional assessment for older adults. He has won several entrepreneurial awards in the field of technology and ageing during his PhD. He was an IoT & AI developer at IBM Canada.
“Improving and maintaining functional ability is the key for healthy ageing. Physiotherapists are at the front line to assess and maintain physical function of older adults. Current practice of physical function assessment includes standardized functional tests such as Timed Up and Go and Sit to Stand, and paper-based documentation of test results. The assessment and documentation process are subjective, time-consuming and varied between therapists.
“In my research, I am building an AI-powered mobile application that automates and standardizes physical function assessments and documentation process. I am using computer vision to automatically track body movement of patients during physical function assessments. With every motion being captured, I will then train machine learning models using the visual data to evaluate the assessment performance. The assessment accuracy using this tool will be validated against the current paper-based assessment in a clinical validation study. The outcomes of this work will be a validated portable tool for physiotherapists to conduct objective, standardized physical function assessments with automatic documentation.”
PhD candidate at the Department of Computer Science, UofT
Research topic: Building intelligent note-taking interfaces for patient encounters utilizing speech and natural language processing.
Research interests: Machine learning and deep learning and their applications in speech, natural language processing and healthcare.
“Currently, for every hour that physicians spend on direct patient care, they spend two hours on administrative and documentation tasks with Electronic Medical Record (EMR) systems. This causes growing dissatisfaction and burnout among physicians, decreasing the quality of care that they are able to deliver to patients. Moreover, during patient encounters, the interaction between physicians and patients could be hindered by the frequent use of monitors and keyboards, which makes healthcare more computer-centered instead of patient-centered.
“My overall goal for this PhD study is to enable more patient-centered healthcare and ease the work burden of physicians on clinical documentation through speech and natural language processing (NLP) technologies and refined interfaces. I have been working on an intelligent note-taking interface for patient encounters, named PhenoPad. All my research topics are tightly around it to spread. The success of projects like PhenoPad depends heavily on their ability to transcribe, ingest and synthesize physician-patient conversations. I work on the pipeline from audio signals all the way to medical notes or electronic medical records that involves multiple technologies, such as speech recognition, speaker diarization, spoken language understanding, question answering and natural language generation.”
2020 Post-Doctoral Fellows
PhD, Electrical Engineering
Research topic: Agitation prediction in people living with dementia using multi-modal sensors.
Research interests: Signal processing, machine learning and data analysis.
“Dementia is one of the major causes of cognitive disabilities. Approximately 50 million people globally suffer from dementia with nearly 10 million new cases every year. People Living with Dementia (PLwD) often exhibit behavioral and psychological symptoms, such as episodes of agitation. During these episodes, they can hurt themselves, other patients and caregivers.
“My goal is to develop methods for prediction of agitation behaviors based on the multi-modal sensor data using machine and deep learning techniques. By predicting the agitation episodes, clinical staff/caregivers would have enough time to prevent them, ensuring the well-being of PLwD. The outcomes of this research will bring improvements upon the existing techniques currently used in clinical practice for agitation assessment by the inclusion of objective sensor measurements.”
Research topic: Optical micromanipulation, microfluidics, micro-/nano-robotics, cell biology, and biomolecule sensing.
Research interests: The development and use of optical micromanipulation technologies for biomedical applications, such as cell sorting, cell analysis, studying intercellular interaction and micro-assembly of cellular/tissue microstructures.
“Neural stem cells are multipotent, self-renewing cells that are responsible for building the fundamental components of the vertebrate central nervous system. Intensive study of these rare, precious cells could lead to a watershed moment in the field of neurobiology, wherein regenerative therapies for ailments such as Alzheimer’s disease, dementia, and traumatic brain injury become an achievable reality.
“Here, l propose to develop a powerful new robotic micromanipulation tool by exploiting multiple artificial intelligence (AI) paradigms for classification and collection of neural stem cells in complex biological samples. The system will utilize state-of-the art AI programs to target and harvest neural stem cells from brain tissues, creating new possibilities for analysis and study of these therapeutically promising cells that will support the development of new regenerative treatments.”