The growth rate in the number of people diagnosed with Parkinsonism is substantial. Estimates indicate thatfrom 1990 to 2015 the number of Parkinsonism diagnoses doubled, with more than 6 million people currentlydiagnosed. By 2040, there will be between 12-14 million people diagnosed with Parkinsonism. Parkinson'sdisease (PD), multiple system atrophy Parkinsonian variant (MSAp), and progressive supranuclear palsy (PSP)are neurodegenerative forms of Parkinsonism, which can be difficult to diagnose as they share similar motor andnon-motor features, and they each have an increased chance of developing dementia. In the first five years of aPD diagnosis, about 58% of PD are misdiagnosed, and of these misdiagnoses about half have either MSA orPSP. Since PD, MSAp, and PSP require unique treatment plans and different medications, and clinical trialstesting new medications require the correct diagnosis, there is an urgent need for clinic ready diagnostic levelmarkers for differential diagnosis of PD, MSAp, and PSP. A promising approach to identify different forms ofParkinsonism is diffusion magnetic resonance imaging (dMRI), as there is no contrast drug, the technique is safeand is already used clinically in traumatic brain injury and stroke. The data collection takes 6-12 minutes and iscompatible on current 3 Tesla MRI systems worldwide. Based on academic research at University of Florida,Automated Imaging Diagnostics, LLC is developing a commercial software package using free-water diffusionimaging as an innovative biomarker to help in the diagnosis of PD, MSAp, and PSP. The software currentlydistinguishes PD, MSAp, and PSP with over 90% accuracy, and can achieve this accuracy on different scannermanufactures. Our next goal in this Phase I project is to further improve the innovation and accuracy of oursoftware technology by employing deep learning classification algorithms for the diagnosis of Parkinsonism. Thespecific aim of this current Phase I project is to substitute and compare the use of our existing Support VectorMachine (SVM) method with two different Residual Deep Neural Network (ResDNN) architectures for estimatingdisease type (PD/MSAp/PSP) through the following two milestones. First, in Milestone 1 we will determine if aResDNN method that processes the same feature vector as our SVM solution improves the accuracy fordifferentiating a) PD and atypical Parkinsonism (MSAp/PSP) and b) MSAp and PSP by 5%. Second, in Milestone2, we will determine if a ResDNN method that processes directly the raw input image data (instead of our derivedfeature vector) improves the accuracy for differentiating a) PD and atypical Parkinsonism (MSAp/PSP) and b)MSAp and PSP by 5%. This Phase I project will facilitate our long-term objective of developing a high-precisiondiagnostic software that can be used by radiologists for diagnosing different types of Parkinsonism.
Public Health Relevance Statement: NARRATIVE
Automated Imaging Diagnostics, LLC is developing a commercial software package using free-water diffusion
imaging as an innovative biomarker to help in the diagnosis of Parkinsonism. The software uses deep learning
methods within an automated image processing pipeline that can function within the workflow of a radiologist.
The optimized solution will be able to diagnose different forms of Parkinsonism and will provide a safe and cost
effective solution because the imaging is using a contrast-free approach that can function on 3 Tesla MRI
systems worldwide.
Project Terms:
