Featured Projects
Featured Projects
Embedded Files
Upper Airway Assessment
Upper Airway Assessment
Upper Airway assessment is of utmost importance before administering anesthesia as sleep apneic patients can develop life-threatening complications. However, current techniques are subjective questionnaires that have very low accuracy. Objective techniques require MRI or CT of the upper airway which are often inaccessible. To address this gap, we introduced an acoustic assessment of the upper airway. We showed the following results:
Upper Airway assessment is of utmost importance before administering anesthesia as sleep apneic patients can develop life-threatening complications. However, current techniques are subjective questionnaires that have very low accuracy. Objective techniques require MRI or CT of the upper airway which are often inaccessible. To address this gap, we introduced an acoustic assessment of the upper airway. We showed the following results:
a) Snoring Sounds can be used to assess upper airway size, length, and its surrounding tissue component,
a) Snoring Sounds can be used to assess upper airway size, length, and its surrounding tissue component,
b) Vowel Sounds can be used to assess the upper airway diameter.
b) Vowel Sounds can be used to assess the upper airway diameter.
Sleep Apnea Diagnosis During Sleep
Sleep Apnea Diagnosis During Sleep
Sleep apnea is a chronic respiratory disorder that happens when the upper airway collapses while sleeping. The biggest hurdles in dealing with sleep apnea are figuring out who has it and knowing why it happens. To help solve these issues, during my grad school studies, we developed a small, easy-to-use device that you can wear on the front of your neck. Furthermore, we have developed several algorithms utilizing conventional machine learning and deep learning techniques to: a) distinguish sleep/awake states, b) detect breath phases c) estimate the tidal volume and respiratory airflow, and d) estimate the severity of sleep apnea detection.
Sleep apnea is a chronic respiratory disorder that happens when the upper airway collapses while sleeping. The biggest hurdles in dealing with sleep apnea are figuring out who has it and knowing why it happens. To help solve these issues, during my grad school studies, we developed a small, easy-to-use device that you can wear on the front of your neck. Furthermore, we have developed several algorithms utilizing conventional machine learning and deep learning techniques to: a) distinguish sleep/awake states, b) detect breath phases c) estimate the tidal volume and respiratory airflow, and d) estimate the severity of sleep apnea detection.
Heart Failure Management
Heart Failure Management
Medly, a Health Canada-cleared Class II software as a medical device is a smartphone-based digital therapeutics program for patients with heart failure (HF) developed by experts from the University Health Network (UHN) and the University of Toronto. Medly uses an in-app rules-based algorithm that delivers personalized self-care messages based on daily input of weight, blood pressure, heart rate, and HF-specific symptoms.
Medly, a Health Canada-cleared Class II software as a medical device is a smartphone-based digital therapeutics program for patients with heart failure (HF) developed by experts from the University Health Network (UHN) and the University of Toronto. Medly uses an in-app rules-based algorithm that delivers personalized self-care messages based on daily input of weight, blood pressure, heart rate, and HF-specific symptoms.
I led three main projects in the next level enhancement of Medly. They are:
I led three main projects in the next level enhancement of Medly. They are:
a) Implement a machine learning model to improve the performance of Medly
a) Implement a machine learning model to improve the performance of Medly
b) Develop a novel algorithm that automatically manages the medication of heart failure patients
b) Develop a novel algorithm that automatically manages the medication of heart failure patients
c) Speech-based prediction of HF decompensation
c) Speech-based prediction of HF decompensation
Impact:
Impact:
Received USD $307,500 worth of funding to lead the projects
Received USD $307,500 worth of funding to lead the projects
Page updated
Google Sites
Report abuse