A new procedure is expected to boost the quality of brain scans so that they can be used to see how strokes impact different people in large-scale studies.
Researchers from the Massachusetts Institute of Technology (MIT), as well as doctors from several institutions, including Massachusetts General Hospital, have created the new method that will enable researchers to better see how genetic factors influence stroke survival and how patients respond to treatment.
“These images are quite unique because they are acquired in routine clinical practice when a patient comes in with a stroke,” Polina Golland, an MIT professor of electrical engineering and computer science, said in a statement. “You couldn’t stage a study like that.”
Images from current scans are often too low in resolution for researchers to analyze the effects of strokes. Scans take images rapidly due to limited scanning time, which results in image slices taken about five-to-seven millimeters apart.
In scientific studies, researchers obtain higher-resolution images with slices only one millimeter apart, but that requires the subjects to stay in the scanner for a longer period of time. Scientists also use specialized computer algorithms to analyze the images but they often don’t work well on the more plentiful but lower-quality patient scans taken in hospitals.
“These research studies are very small because you need volunteers, but hospitals have hundreds of thousands of images,” said lead author Adrian Dalca, a post-doctoral student in MIT’s Computer Science and Artificial Intelligence Laboratory.
The researchers essentially filled in the data that was missing from each patient scan by taking information from the entire set of scans and using it to recreate anatomical features that are.
“The key idea is to generate an image that is anatomically plausible, and to an algorithm looks like one of those research scans, and is completely consistent with clinical images that were acquired,” Golland said. “Once you have that, you can apply every state-of-the-art algorithm that was developed for the beautiful research images and run the same analysis, and get the results as if these were the research images.”
When the images are generated, researchers then can run a set of algorithms to help with analyzing the anatomical features, including the alignment of slices and a process called skull-stripping that eliminates everything but the brain from the images.
The algorithm keeps track of which pixels were from the original scans and which were added afterward, meaning that measuring the extent of brain damage can only be performed on information from the original scans.
“In a sense, this is a scaffold that allows us to bring the image into the collection as if it were a high-resolution image and then make measurements only on the pixels where we have the information,” Golland said.
The new scans could also be used to study other disorders including Alzheimer’s disease. For now, the researchers plan to apply the new scan to a large set of stroke images which includes 4,000 scans from 12 different hospitals.
“Understanding spatial patterns of the damage that is done to the white matter promises to help us understand in more detail how the disease interacts with cognitive abilities of the person, with their ability to recover from stroke, and so on,” Golland said.