We know that both Alzheimer’s and Parkinson’s diseases affect certain parts of the brain first, before affecting the rest of the brain, and often before people start to show signs of the disease. These brain structures can be analyzed precisely in a living person using 3D MRI imaging and computer algorithms. Studying changes in the shape of these structures — such as the hippocampus and the caudate, responsible for memory and movement — can help us predict the onset of disease. Because the two diseases share some similarities, studying the similarity between subcortical shape changes between Alzheimer’s and Parkinson’s can also tell us more precisely how the two conditions are similar.
We hypothesize that change in the shape of several specific brain structures over time will be predictive of disease onset, and will share some similarities between Parkinson’s and Alzheimer’s.
There are already several hundred brain images of people with Alzheimer’s and Parkinson’s, taken over time. Taking advantage of this data, we will apply algorithms for separating the specific regions of the MRI scans that correspond to the brain structures of interest. Once the regions are “segmented,” a model of their boundaries will be used as the starting point for analysis. Changes in the shape of these boundaries will then be used to discover a signature of how the disease progresses over time and differentiate it from healthy aging.
Impact on Diagnosis/Treatment of Parkinson’s Disease:
Our proposed method is non-invasive and may be used to score a patient as more or less likely to progress to Parkinson’s disease, as well as alert the clinician to possible other dementias that are present, but harder to detect.
Next Steps for Development:
There are two avenues to pursue: A clinical trial-oriented software distribution to help other researchers use the newly discovered signatures of disease to evaluate disease-modifying drugs. The other is a direct application in the clinical setting, helping doctors better diagnose patients.