Scientists from CSIRO, Australia’s national scientific institution, and Queensland University of Technology are working together to develop a “world’s first benchmark” to measure brain atrophy, or thinning, in neurodegenerative diseases such as Alzheimer’s. Partnered to use artificial intelligence.
Alzheimer’s disease is the most common form of dementia, accounting for 60-80% of cases. The Commonwealth Science and Industrial Research Organization says one way to measure that progress is with MRI images that show cortical thinning. sub-millimeter range.
“Using magnetic resonance imaging (MRI) of the brain to assess the onset and progression of Alzheimer’s disease has traditionally been difficult because changes in cortical thickness are very small, often in the submillimeter range. was difficult,” points out CSIRO.
“Advanced machine learning techniques are routinely used in brain research to assess changes in cortical thickness, but until now, clinically accurate ‘ground truth’ datasets have been lacking. , could not assess its sensitivity to the detection of small atrophy levels.
|
“Prior to this breakthrough, the only way to obtain an accurate measure of cortical thickness was to study the postmortem brain. You will not get it.”
Filip Rusak, a research scientist at CSIRO’s Australian e-Health Research Centre, said: Cortical atrophy—the thinning of the brain’s cortex—may begin 10 years or more before clinical symptoms of Alzheimer’s disease appear.
“We need very precise methods to see these signs on brain imaging when they start to appear so that we can address them earlier rather than later,” said Dr. Rusak. I’m here.
“Using the power of machine learning, we are able to generate a series of artificial MRI images of the brain with predefined signs of neurodegeneration in the cortical regions, the outer layer of the brain most affected by Alzheimer’s disease. I was.
“Until these findings were available, there was no way to conclusively determine the sensitivity of the different methods used to measure cortical thickness in patients with Alzheimer’s disease,” he said.
According to CSIRO, this new technique allows researchers to set the amount and location of brain degeneration they want to compare, giving them a clear picture of which method of cortical thickness quantification works best. They add that the technique can test the sensitivity of the method. down to the slightest level. You can determine if you can detect thickness changes as small as 0.01 mm.
The survey results are medical image analysis CSIRO says the work is already having international impact.
Michael Rebsamen, University of Bern, Switzerland They say there is strong evidence that DL+DiReCT, a deep-learning-based method for measuring cortical thickness, is robust and sensitive to subtle changes in atrophy.
“Until now, we have not been able to quantify the level of atrophy that we can actually measure, because we don’t have a reference MRI,” Dr. Rebsamen said.
“CSIRO’s innovative benchmark fills this gap and marks an important milestone for assessing cortical thickness methods,” he said.
According to CSIRO, the method can be applied to study any brain disease that involves neurodegeneration, is an important step toward a better understanding of dementia and other debilitating brain diseases, and predicts an individual’s level of cortical degeneration. It may also be used for Expect time to pass.
According to Rusak, the technology is all based on commonly used and relatively inexpensive MRI images.
“The findings will help researchers choose the right tools for the job, and the right tools will increase the chances of accurately assessing disease progression,” said Dr. Rusak.
“So we don’t need new medical infrastructure,” he said.
An image of the synthetic dataset was created publicly available As such, clinicians and scientists can use the composite image to make their own assessment of how cortical thickness is quantified.
B-roll footage, including an interview with Dr. Rusak on MRI imaging, can be accessed at: this link.