Thousands of stroke victims are disabled every year because the right treatment comes too slowly. Some emerging technologies may change that.
The technologies use computer algorithms to cross-reference a stroke patient’s brain scan with a vast database of scans from other victims, allowing a precise diagnosis in minutes. It can be done by less-experienced doctors in outlying hospitals or by paramedics with portable scanning devices in an ambulance. The scans and results can be quickly sent to a specialist at a regional stroke treatment center who can confirm the diagnosis.
The result: Victims of strokes—an interruption of the brain’s blood supply that deprives brain tissue of oxygen—can be rushed to a stroke center with the most-skilled doctors who can remove blood clots. That contrasts with a too-common scenario: languishing for hours at a less-equipped hospital waiting for a diagnosis that often comes too late to help.
“There is no more time-sensitive treatment in all of medicine than treating the stroke victim,” said
Thomas G. Devlin,
chairman of neurology at the University of Tennessee’s Erlanger Medical Center in Chattanooga, a top stroke-treatment center. “The new technology has the potential for shaving off critical minutes, sometimes even hours, in the diagnosis, triage and treatment of stroke.”
Companies developing different forms of this new technology include Viz.ai Inc. of San Francisco and Neural Analytics Inc. of Los Angeles. Both have been testing their new imaging at Erlanger and elsewhere. Another important new technology, called RAPID from IschemaView Inc. of Redwood City, Cal., produces automated images showing how much brain tissue is salvageable after a severe stroke. Many stroke neurologists say this technology could break new ground in getting stroke patients the right treatment.
At issue with the technologies are the most severe blockages in major arteries caused by clots.
Under current protocols, patients with severe strokes are typically rushed to the nearest hospital, regardless of the facility’s ability to treat the most-severely ill patients. Once there, such patients undergo tests and scans for a diagnosis that often takes hours. With the clock ticking, many patients find their best chance for survival and recovery is an innovative procedure called a thrombectomy, which allows doctors to remove a clot to restore blood flow, according to stroke doctors, medical records and journal articles.
The thrombectomy has proven highly effective and is transforming stroke treatment. But as The Wall Street Journal recently reported, many stroke patients lack timely access to the complex procedure.
Many hospitals don’t have the expertise or the facilities, leaving them with two choices: Lose precious time transferring a patient to a better-equipped stroke center, or treat the patient with a clot-dissolving drug. This drug works well for some moderately ill stroke patients. But the therapy often fails for thousands of severely stricken patients—those with large clots blocking major arteries.
A study in the journal Circulation last year of 984 such patients showed that treatment delays for them led to worse outcomes if they were transferred between hospitals instead of going straight to a hospital that could pull clots out with a thrombectomy. The new technologies could help fix that.
Erlanger treated 1,721 patients last year with clot-based strokes, 180 of whom got thrombectomies. The hospital employs a far-flung network of helicopters and ambulances that can rush stroke patients from up to 150 miles away. But like most top stroke centers, Erlanger depends on fast referrals from other hospitals for optimal treatment. Minutes make all the difference between recovery, disability or death, doctors say.
In one case during January, a 64-year-old stroke patient arrived at another Chattanooga hospital at 1:07 p.m. After two brain scans, she was transferred to Erlanger at 8:05 p.m., too late for a thrombectomy, Dr. Devlin said. She now resides in a nursing home, unable to speak or use her right leg and arm, her husband said.
The technology from Viz.ai (the ai is for artificial intelligence) recently gained Food and Drug Administration approval. The company and Erlanger are talking to other Chattanooga hospitals about installing its diagnostic software there, which would allow neurologists to see brain scans almost simultaneously.
A Viz.ai study of 300 patients, which prompted the FDA approval, showed that the company’s software was able to notify a stroke neurologist on average 7.3 minutes after the brain imaging took place—compared with the hours that it sometimes takes with conventional diagnostic procedures. The technology accurately identified severe strokes as precisely as expert stroke radiologists do, the study found.
The system holds promise for improved outcomes, medical experts say, but only if it can be widely adopted by hospitals. One concern is cost. Companies developing the technology haven’t set a price but say it won’t be cost-prohibitive.
Some neurologists who specialize in thrombectomy say many doctors at outlying hospitals will resist transferring patients and losing their business. But Dr. Devlin said several hospitals are already planning to install Viz.ai technology in their facilities.
The idea for Viz.ai’s system came in part from
a British neurosurgeon who in 2014 participated in a successful operation on a brain-injured woman. But she later died because it had taken four hours to get her to the right hospital.
Dr. Mansi later enrolled in Stanford Business School where he took an interest in developing technology to speed such patients to the right specialist for better outcomes. He also took a course on venture capital for entrepreneurs, taught by former Google chief
At the course’s end, Dr. Mansi made a presentation that impressed Mr. Schmidt. His venture-capital fund and others invested $7.5 million in what became Viz.ai. Dr. Mansi serves as Viz.ai’s president and chief executive.
“I like the medical imaging area, as I think AI will lead to much better outcomes,” Mr. Schmidt said. “There are so many opportunities to change and improve medical care through AI.”
The other company developing similar stroke technology, Neural Analytics, uses ultrasound devices that would attach to the patient’s head in the ambulance to produce images that measure blood flow in the brain. Neural Analytics’ Lucid system compares patients’ brain blood-flow images with a database of thousands of such images to quickly pinpoint the problem. Paramedics could then direct the image immediately to a stroke hospital.
the company’s chief scientific officer, was a Ph.D. student in bioengineering at UCLA when he and colleagues thought of this technology for other types of brain injuries. They later realized that it could also be used to rapidly detect strokes caused by blood clots. The company’s most recent study, conducted at Erlanger by Dr. Devlin, showed its technology detects 97% of large-artery strokes. It is seeking FDA approval and hopes to be selling its product by this fall.
“By detecting in the ambulance which patients have blood flow disruption,” said Dr. Devlin, this device “has the potential to revolutionize stroke care.”
Write to Thomas M. Burton at firstname.lastname@example.org