Viz.ai-Applying AI to Reduce Time to Life Saving Treatments
Viz.ai recently raised $40M in growth capital from CIBC Innovation Banking. The additional funding brings Viz’s total fundraising to approximately $292M. Via has developed a software platform based on artificial intelligence (AI) that is designed to improve communication between care teams handling emergency patients (first applied to stroke patients) by helping improve care coordination and dramatically improved response times. The company will use the funds to help increase expansion and power its expansion, including the possibility of acquisitions.
While the total cost of strokes in the U.S. was approximately $220 billion, the cost due to under-employment was $38.1 billion, and $30.4 billion from premature mortality (Via.ai & Journal of the Neurological Sciences)
The risk of having a first stroke is nearly twice as high for blacks as for whites and blacks have the highest rate of death due to stroke (American Stroke Association)
Each one minute [delay in care for stroke victims] translates into 2 million brain cells that die (Viz.ai)
Stroke is the number one cause of adult disability in the U.S. and the fifth leading cause of death (American Stroke Association )
Viz.ai was founded by, Dr. Chris Mansi and David Golan. While working as a neurosurgeon in the U.K. Dr. Mansi observed situations in which a successful surgery was performed yet patients would not survive because of extended time lapses between diagnosis and surgery, which was particularly true with strokes. For example, when doctors believe there has been a stroke, they typically would order x-rays and a series of CT scans, and while the scans themselves typically happened quickly, there often was a sizeable delay before the studies could be read by a competent professional. Once the readings were performed by a radiologist, there was often a further delay in care as clinicians had to inform a local stroke center of the diagnosis and then ensure the patient was transferred to that center for treatment.
Dr. Mansi met Golan in graduate school at Stanford while studying for his M.B.A. Golan was suspected of having suffered a stroke prior to entering Stanford and the two classmates lamented the lack of available data for stroke treatment. As noted by the company, “Mansi learned how undertreated large vessel occlusion (LVO) strokes were and wanted to be an agent of change.” Mansi and Golan, collaborated on a plan to apply A.I. to increase the data for stroke treatment and viz.ai was born.
As noted in a recent Forbes article, the company’s “software cross-references CT images of a patient’s brain with its database of scans to find early signs of LVO strokes. It then alerts doctors, who see [and communicate] about the images on their phones” and allows those clinicians to communicate with specialists at stroke centers and arrange for patients to be transferred there for care. According to the company, this leads to dramatic decreases in the time it takes for patients to go from diagnosis to procedure, commonly referred to as “door to groin puncture times.” Originally developed for LVO’s Viz has now received FDA approval for 7 A.I. imaging solutions and has extended treatment from LVOs to cerebral aneurysms (February 2022), subdural hemorrhage (July 2022) and most recently hypertrophic cardiomyopathy-HCM (pending).
As noted above, Viz.ai’s system enables it to automatically scan all images in a hospital system and scan for the noted conditions, then alerts clinicians if any are detected. In the case of LVOs the system then allows doctors to view images of patient scans on their phones, exchange messages, and cut crucial time off diagnosis and treatment. As the company notes, this is particularly important for smaller facilities which often lack specialists to interpret scans and arrange for transitions in care. For example, according to a study in the American Journal of Neuroradiology looking at stroke treatment at a facility using viz.ai technology, researchers found “robust improvement” in other stroke response metrics, including door-to-device and door-to-recanalization and a 22% overall decline in time to treatment. This is particularly important in the case of strokes which are the number one cause of disability and the fifth leading cause of death, as time is of the essence for stroke victims with each minute of delay adding one week of disability.
Applying technology to help reduce delays in diagnosis and treatment is one of the most promising applications of artificial intelligence because of the vast amounts of data they can process in short periods of time. While over time, many hope and some fear that these types of technologies will be able to be “taught” how to diagnose and treat illness, in the near term their greatest use lies in augmenting the skills of clinicians by allowing them to focus their attention on areas most in need of an experienced, nuanced diagnosis. This is particularly true for brain injuries where literally every second and every minute count. For example, as noted by the company “every one minute [delay in care for stroke victims] translates into 2 million brain cells that die.” Given that the loss of brain cells results in loss of brain function, disability, or worse the costs to society can be quite high. According to the company, strokes cost the U.S. healthcare system about $220 billion annually and each LVO patient that you are able to treat with a timely thrombectomy” costs one-tenth, or almost $1 million less than those that who aren’t. It is practical examples of the clinical applications of A.I. such as this, which attack very concrete and tangible problems, that are likely to pave the way for acceptance of more complex applications in the healthcare delivery system.