The Michael J. Fox Foundation for Parkinson's Research

The Michael J. Fox Foundation for Parkinson's Research is using its 2019 distribution to fund:

PROJECT TITLE: Mapping brain response patterns to Deep Brain Stimulation with fMRI

Investigator/Author: Dr. Andres Lozano and Dr. Radhika Madhavan

Objective: We will examine whether a brain imaging technique called functional Magnetic Resonance Imaging (fMRI) - a technique that maps brain activity - can help improve the current programming approach by making it easier, faster, and more precise. Specific aims are to (1) detect the specific brain circuits engaged with DBS using fMRI and (2) assess fMRI as a tool to select optimal DBS settings.

Background: Deep brain stimulation (DBS) targets malfunctioning brain circuits. Commonly used to treat Parkinson’s disease (PD), this surgical therapy can produce striking clinical benefits when the appropriate electrical stimulation settings have been selected (i.e., programmed). However, DBS programming often requires multiple clinic visits to test the large number of possible stimulation parameters. PD DBS patients are thus in a position to benefit from a novel tool that improves on the current programming approach.

Methods/Design: First, patients who are already programmed at their best DBS settings will be scanned with fMRI. These images will help us understand which brain areas are activated by ‘optimal’ settings. We will then build computer algorithms that identify these brain responses. Second, we will recruit patients who have not been programmed yet and split them into 2 groups: one group in which their DBS settings are conventionally chosen by their physician, and another group in which they will have fMRI and the computer algorithms to guide the selection of their settings. We will then compare the clinical benefits achieved with both programming methods.  

Relevance to Diagnosis/Treatment of Parkinson’s Disease: To date, over 150,000 patients have been implanted with DBS for Parkinson’s disease. The process of programming DBS systems takes numerous clinic visits, resulting in a long, expensive, and tiring process for patients. By predicting the best settings with a single MRI session, the process can be simplified and improved for patients and practitioners. This new technology when validated could be shared with other DBS centers worldwide, improving the care of PD patients undergoing DBS surgery.

November 2020 Project Update:

Lozano’s team has made considerable progress toward their goals for the first phase of the project despite facing the impact of COVID-19. They have completed fMRI imaging on 24 people with Parkinson’s and generated individual maps of brain activation patterns evoked in response to either ‘optimal’ DBS settings (correlating with the best clinical improvement) or non-optimal DBS settings. Researchers then used these brain activation imaging data as inputs for a machine-learning algorithm, which has resulted in a model that is able to distinguish optimal from non-optimal DBS settings with 76% accuracy. With the first phase of the project complete, the investigators will now begin patient recruitment and clinical enrollment for the second phase: a prospective trial to assess the clinical utility of the machine-learning algorithm to automate the programming of DBS settings compared to standard clinical assessment-based programming methods.