Dr. Doug Arnold, chief of staff/VP Medicine at Timmins and District Hospital (Photo credit: Paige Weir, Timmins and District Hospital)
  • Unity Health Toronto and Timmins and District Hospital will begin testing an artificial intelligence (AI) tool called ASIST-TBI at the Timmins, Ontario hospital.  
  • The AI tool, originated at Unity Health’s St. Michael’s Hospital, predicts when patients who have a traumatic brain injury (TBI) are going to need neurosurgery to help prioritize urgent cases. 
  • Timmins and District Hospital is the first hospital to enrol in a study that aims to test and validate ASIST-TBI at 18 Ontario hospitals, with the goal of making the AI tool available to emergency physicians in real-time. 
  • Already in use at St. Michael’s, testing and validating the model with Timmins and District Hospital’s imaging data is critical to ensure its predictions are accurate.  
  • The successful adoption of the AI tool could help to manage and conserve two critical resources throughout Ontario: emergency transport and neurosurgical expertise.   

Unity Health Toronto and Timmins and District Hospital will begin testing an artificial intelligence (AI) tool called ASIST-TBI at the Timmins, Ontario hospital. The tool predicts when patients who have a traumatic brain injury (TBI) are going to need neurosurgery. ASIST-TBI helps to prioritize the most urgent cases so — when every second counts — patients are routed to specialist care as quickly as possible. 

Timmins and District Hospital is the first hospital to enrol in a study that aims to test and validate ASIST-TBI at 18 Ontario hospitals, with the goal of making the AI tool available to emergency physicians in real-time.  

The AI tool originated at Unity Health’s St. Michael’s Hospital and was developed by Dr. Christopher Witiw, neurosurgeon; Christopher Smith, PhD and AI scientist, and Dr. Alun Ackery, emergency and trauma physician and medical director of CritiCall Ontario. The tool is now being used in practice at St. Michael’s. 

“Though this tool was developed by a trauma center that has neurosurgical expertise in-house, it’s more relevant for places that don’t have neurosurgical care,” says Ackery. “In a rural hospital, having tools that support specialist decision-making can reduce delays, especially when specialists are often hours away in city centres.”  

Enjoying this story? Sign up for the Unity Health Toronto newsletter, a monthly update on the latest news, stories, patient voices and research emailed directly to subscribers.  

Timmins and District Hospital is medium-sized community hospital that is geographically isolated in northern Ontario. The closest neurosurgical centre is in Sudbury, hours away.  

The St. Michael’s team partnered closely with CritiCall Ontario, OCINet and Dr. Doug Arnold, chief of staff/VP Medicine at Timmins and District Hospital, who sponsored the project to gain the necessary agreements and permissions to begin testing the tool.  

“It seemed like a no-brainer. Like many emergency departments across the province, our ED physicians are often overloaded and overwhelmed with the volume of patients they are seeing,” says Arnold. “Trying to connect with specialists to arrange definitive care and determine whether a transfer is required consumes a lot of their time. By contributing our imaging data, we can help ensure this tool is accurate and effective in supporting care decisions for patients.”   

While the tool is already in use at St. Michael’s, testing and validating the model with Timmins and District Hospital’s imaging data is critical to ensure its predictions are accurate.  

“It’s documented in literature that medical imaging models can falter when used on different CT scanner manufacturers,” explains Smith. “You’ll also see different reasons for head injuries in a rural hospital — I don’t think anyone’s showing up with a skiing injury in downtown Toronto — which could impact the model’s performance.” 

The lightbulb moment  

The idea for ASIST-TBI came to Witiw in the middle of the night when he was the neurosurgeon on call at St. Michael’s. He was staring at the head CT scan of a patient with a TBI while he was on hold to speak to the treating physician — an emergency doctor at another hospital — to provide a consult. The conversation with the emergency doctor would give additional context — but the scan alone was enough for Witiw to determine that the patient needed surgery.  

That’s when the idea struck him – what if there was an AI tool that could look at head CT scans and predict which patients would need neurosurgery and which ones wouldn’t to help prioritize care? 

St. Michael’s receives neurosurgical consult referrals through CritiCall Ontario from 24 hospitals in Ontario that don’t have neurosurgical capabilities – in addition to the high volume of consult referrals that come through its emergency department as a Level 1 trauma centre. 

“If you take every other specialty in the hospital — surgical and medical — and combine all the consults they receive from outside hospitals, it is still less than the volume of consults that the neurosurgeon on call receives,” explains Witiw. “Yet, a large number of these consults don’t require patient transfer to a neurosurgical center.” 

That high volume of consults, particularly lower-urgency consults, draw neurosurgeons away from other important work. Often, they are in the middle of a surgery having to step away from a patient on the operating table to provide a consult.  

“ASIST-TBI supports emergency doctors to know when they need to be in touch with a neurosurgeon immediately versus when they can wait until the neurosurgeon on call is out of surgery,” says Witiw. Witiw and Ackery brought their idea to St. Michaels’ Hospital Foundation’s research competition in 2024, where they received the Odette Award for Health Innovation. That early investment, combined with additional funding from Michael and Janet Norris, gave the team the momentum to take ASIST-TBI to the next level.

Expanding across Ontario 

Broader testing and validation of ASIST-TBI as part of a multi-hospital research study is the first step in making the tool available province-wide. 

“Timmins and District Hospital are the first and it was largely thanks to Dr. Doug Arnold championing it there,” notes Ackery. “He saw that ASIST-TBI could address a problem they are experiencing and wanted to put his support behind it.” 

The successful adoption of the AI tool could help to manage and conserve two critical resources throughout Ontario: emergency transport and neurosurgical expertise.

Ornge provides air ambulance service and emergency medical transport in Ontario. It serves more than 14 million people over more than one million square kilometres of land.  CritiCall Ontario recently integrated with Ornge and plays a central role in connecting hospital-based physicians in Ontario with specialists – including neurosurgeons like Dr. Witiw – who can provide consultations for urgent, emergent and critically ill patients. CritiCall facilitates close to 60,000 cases per year with less than 35 percent resulting in a transfer. 

“People often think, ‘Why don’t you just transfer everybody? That’s the safest thing to do,’” says Witiw. “If we do that, we’re taking patients away from their hometowns and families. We’re also using transport resources and potentially taking those transport resources away from patients that really need it.” 

Further, only 13 of Ontario’s 140 public hospitals are designated neurosurgery centres and neurosurgery has one of the longest training periods of any medical specialty.    

“If we can do something to make sure neurosurgeons are seeing people who have the highest likelihood of needing their services, it will help with the equity of caring for people scattered throughout our huge province,” says Arnold. 

This study is possible through the support of provincial partners including OCINet, Ontario’s Clinical Imaging Network and Ornge, which now includes CritiCall Ontario. Together, they provide Ontario’s air ambulance and critical care transport service and consultations for emergent and critically ill patients in the province. 

By Robyn Cox

Related Tags