New AI Technology Can Help Assess Brain Lesions and Atrophy

Reading Time: 3 minutes

A new type of artificial intelligence (AI) software is enabling physicians at the University of Miami Miller School of Medicine to make more precise brain measurements of people with neurologic disorders, including neurodegenerative disorders such as Alzheimer’s, dementia, multiple sclerosis, brain trauma, and epilepsy.

LesionQuant on computer screen.

FDA-approved NeuroQuant and LesionQuant, AI products of CorTechs Labs, can aid in diagnosis and management of patients with various neurological disorders. NeuroQuant segments and measures the volumes of different areas of the brain and compares the results to a normative database of healthy patients. LesionQuant is very useful in patients with multiple sclerosis, providing the same brain structure data while additionally quantifying the number, size, and location of lesions, and comparing them to prior studies of the same patient.

“This AI tool is actually very, very helpful in the assessment of patients with dementia,” said Gaurav Saigal, M.D., chief of the Section of Neuroradiology at the Miller School. “This is a tool that was missing in the overall diagnosis of people with dementia. It completes our armamentarium,” he said. “Combined with other clinical information, the AI analysis tool can improve the accuracy of dementia diagnosis.”

Alzheimer’s disease, for example, results in atrophy of certain parts of the brain. In the early stages of the disease, it is hard to recognize these changes with the naked eye. Without this new technology, “it’s hard for us to know if one part of the brain is atrophied in relation to another part, especially in the early stages of the disease,” Dr. Saigal said.

Furthermore, UM’s clinicians see a wide range of dementia types. “Distinguishing between Alzheimer’s disease and another form of dementia is often tricky,” Dr. Saigal said. “And it’s another instance where the software can help, by determining the area of the brain involved in early stages of the disease process.”

NeuroQuant on computer screen.

The AI software can also play a role in evaluation of people with epilepsy. Different types of epilepsy affect certain small structures in the brain and the changes can be very minimal, according to Dr. Saigal. Using this AI tool, subtle changes in volumes can be easily and accurately determined, thus increasing the radiologist’s confidence in making such a diagnosis, he added.

When it comes to multiple sclerosis, neuroradiologists play an essential role in assessing characteristic brain lesions seen on MRI. The NeuroQuant software can ease the measurement and quantification of these lesions and save radiologists time.

“It helps to assess if the lesions are changing compared to a prior MRI. It also tells us the total volume of the lesions in the brain,” Dr. Saigal said. “Imagine if you have someone with 50 lesions. Before we had to look at those 50 lesions one by one to find out if any had changed or not. With this new AI tool, we can rapidly determine if there is a change in the size and/or number of lesions compared to a prior study.”

Dr. Saigal and his colleagues are extending the use of this tool ever further in the form of a study. Traditionally all patients with multiple sclerosis get gadolinium contrast injections as part of their MRI study to determine if a lesion is active or not.

“Gadolinium has been shown to get deposited in the brain tissues,” Dr. Saigal said. “There is a big movement to avoid gadolinium contrast if possible.”

Although, to date, no side effects have been reported with this contrast deposition in the brain, Dr. Saigal and his colleagues want to see if results obtained using this AI software in real time can aid in determining the need to give contrast or not. They have analyzed two patients to date with the NeuroQuant software. The goal is to enroll 50 patients in this study to determine if obtaining data in real time by using this AI tool can help determine if a contrast injection is needed or not.