Heartbeat: Bringing AI to the SWB mix
February 2, 2021
Exciting news! Our Trust has partnered with a University of Oxford spin-out company called Brainomix to support our acute stroke service using artificial intelligence (AI).
Artificial Intelligence technology is a set of algorithms and is used to make sense of a variety and large volume of clinical information more quickly than a human. The presentation of acute ischaemic stroke in the emergency department can be difficult for non-specialist clinicians.
A stroke is when the blood supply to the brain is affected. This is a medical emergency, and, urgent treatment is essential. A stroke is usually due to either a blockage or if the blood vessel supplying the brain is weakened and burst. The sooner a patient receives appropriate treatment – damage is less likely to occur.
Strokes are usually treated with medicines to prevent and dissolve blood clots, reduce blood pressure and reduce cholesterol levels. The implementation of the AI to support our clinicians was led by Dr Bahadar Bhatia, Dr John Morlese and Dr Sissi Ispoglou, with hosting support from the Informatics Infrastructure team.
When a possible stroke patient attends the emergency department, they are assessed and referred for a CT medical scan of their head. After the patient has had their CT medical scan, their images are sent to AI, which processes these images within a few minutes, sending its findings to the stroke doctors to review.
A notification may be sent to their mobile phone to inform them that the AI findings are ready for their review alongside their images. The system will provide augmented intelligence for detecting the extent of large vessel occlusions and perfusion within collaterals. This approach provides prompt findings to aid the stroke doctors with the clinical management of the patient.
Dr Ispoglou noted the different angle in viewing these AI findings. She said: “This is a fantastic opportunity to take the care we provide for our stroke patients to a different level. The program will not replace knowledge, clinical skill and common sense. It will, however, support us to make better decisions and ultimately help our patients, with a particular focus on those that have significant clot type of stroke and can benefit from clot busting or clot removal.”
In addition to this, Dr Bhatia has been leading several AI projects with IBM Watson, with another live clinical AI system for a number of chest and abdomen conditions expected early next year.
Dr Sarah Yusuf, Group Director for Imaging, added: “Brainomix eStroke is the first clinical AI solution implemented in real-time within the Trust.
“Brainomix is part of a wider Trust level AI strategy to improve efficiency and patient care. There will be further clinical AI solutions implemented in the future ready for MMUH. Our strategic aim is to embed technology solutions to improve patient pathways and information flow within the hospital and with our strategic partners.
“Imaging is central to the emergency management of stroke patients. Accurate and timely CT head reports are essential to reduce door to needle time for thrombolysis. Brainomix will improve the speed of diagnosis of acute stroke and the clinical management of patients that can have thrombolysis.
Dr John Morlese, a Consultant Radiologist, said: “We are delighted to have this opportunity to give our patients such excellent care. We will soon implement a CT brain perfusion model that will aid delivery of life saving thrombolysis to patients who wake up with an ischaemic stroke. Thank you to Dr Bhatia for his diligence in bringing this to fruition.”