Doctors tell boy, 15, he had a migraine after rugby tackle - but he was actually suffering a paralyzing stroke which nearly killed him
The neurologist replacement here:
Scientists at Imperial College London and the University of Edinburgh in Britain have created machine-learning software to identify and measure the severity of small vessel disease more accurately than some current methods. Their findings were published in the journal Radiology.
"This is the first time that machine learning methods have been able to accurately measure a marker of small vessel disease in patients presenting with stroke or memory impairment who undergo CT scanning," lead author Dr. Paul Bentley, a clinical lecturer at Imperial College London, said. "Our technique is consistent and achieves high accuracy relative to an MRI scan -- the current gold standard technique for diagnosis."
Doctors now diagnose small vessel disease by looking for changes to white matter in the brain during MRI or CT scans. But Bentley said it is often difficult to detect the edges of the SVD, making it difficult to estimate the severity of the disease from CT scans. MRIs are more sensitive, but the scanner might not be available and suited for emergency or older patients.
"Current methods to diagnose the disease through CT or MRI scans can be effective, but it can be difficult for doctors to diagnose the severity of the disease by the human eye," Bentley said. "The importance of our new method is that it allows for precise and automated measurement of the disease."
Studied were historical data of 1,082 CT scans of stroke patients across 70 hospitals in Britain between 2000 and 2014, including cases from the Third International Stroke Trial.
The software, identifying and measuring a marker of SVD, gave a score of how severe the disease was, ranging from mild to severe. These results were compared with information from panel of expert doctors who estimated SVD severity from the same scans. The software was as good as the experts.
In addition, 60 MRI and CT scans were checked in the same subjects. The software was 85 percent accurate at predicting the severity of SVD.
"This is a first step in making a scan reading tool that could be useful in mining large routine scan datasets and, after more testing, might aid patient assessment at hospital admission with stroke," Dr. Joanna Wardlaw, head of neuroimaging sciences at the University of Edinburgh, said.
Bentley said the software can estimate the likely risk of hemorrhage in patients, including whether to treat with clot busters. And he suggested the software can statistically indicate the likelihood of patients developing dementia or immobility because of slowly progressive SVD.