The Use of fMRI for Detecting Abnormal Neural Synchronization

Application for Early Diagnosis, Follow Up, and Therapeutic Strategies for Parkinson's Disease

Gadi Goelman, Ph.D.

Hadassah Hebrew University Hospital

Funded in March, 2003: $40000 for 2 years


back to top

Using fMRI to Determine Whether Nerve Firing is Unsynchronized in Parkinson’s Disease

Based on the researchers’ Dana-supported demonstration that a new MRI imaging technique can non-invasively detect asynchronous firing of neurons, the investigators now will determine whether this abnormal firing occurs in patients with Parkinson’s disease.  If so, their research could provide a better understanding of the disease process and new approaches to treating it.

Previous studies have shown that Parkinson’s disease patients have pronounced cell loss within the brain’s basal ganglia, and a correspondingly low level of brain dopamine, a neurotransmitter that facilitates cell to cell communication.  Decreased dopamine levels disrupt the connection between the basal ganglia and the brain’s cortex.  This disruption results in the disease’s familiar symptoms of tremor, difficulty initiating movements, and muscle rigidity.  Surgical intervention in selective parts of the basal ganglia can alleviate most of these motor symptoms.

Yet some experimental findings suggest that this model is over-simplified.  The researchers hypothesize that brain cells in the basal ganglia do not fire synchronously in a coherent pattern.  This asynchronous firing decreases the brain cells’ connectivity locally in the basal ganglia and globally to the cortex, resulting in the Parkinson’s symptoms.  The investigators have shown that a new MRI technique, called “RCC” (Radical Correlation Contrast), can non-invasively reveal basal ganglia cell firing.  They will use this technique to compare synchronization patterns at rest and during activation in healthy and Parkinson’s disease model rats.  Thereafter, they will use MRI-RCC to compare synchronization patterns in Parkinson’s patients to those of healthy adults.  If the investigators are correct, this study would provide new insight into the disease process, a new method for diagnosing and defining the disease’s severity, and a way to assess therapies designed to re-synchronize nerve cell firing.

Significance:  If this imaging technique reveals that asynchronous firing of nerve cells in the basal ganglia occurs in Parkinson’s disease, the findings could lead to new insights into the disease process, improved early diagnosis, and assessment of treatments designed to restore synchronization.


back to top

The Use of fMRI for Detecting Abnormal Neural Synchronization

At present the function of the basal ganglia (BG) in the control of motor and cognitive functions is still unresolved. The common models of the circuitry between the BG nuclei and the cortex that are used to understand BG clinical disorders can not explain all clinical symptoms of the disease or recent findings of pre-frontal overactivation. Our recent studies on MPTP, a primate model of Parkinson's disease (PD), suggest that temporal patterning and temporal coherence of neural signals may have a key role in the processing carried out by the basal ganglia - cortex circuit. We therefore hypothesize that the loss of processing in spatially and temporally independent channels in the basal ganglia-cortical networks, is an important pathophysiological principle of hypokinetic (e.g., Parkinson's disease) and hyperkinetic (L-DOPA induced dykinesias) movement disorders. However, today only invasive methods can be used for evaluating the synchronicity of neural network and even these methods give limited information.

In this project we propose to use our recently developed MRI Radial Correlation Contrast (RCC) method that, non-invasively, gives information on neuronal synchronization to study the changes in BG-cortex synchronization in Parkinsonian state. In the new RCC approach (as extension of the function connectivity MRI method), two complimentary types of images are obtained from each set of functional MRI data. One is proportional to the average neuronal synchronization of each volume element with its vicinity, and the other gives the weighted direction of this synchronization. The method was demonstrated on high spatial resolution data from rat brain. It was shown that functional clusters, like cortical layers, are distinguishable based on their different inter-communication. The RCC method will be optimized for human studies including its extension to 3D and will be used to map the changes in the BG-cortex synchronization circuit in human as well as in rodent 6-OHDA model of Parkinson's disease.

The expected outcome of the project is two-fold. One is better understanding the role of the BG in normal and pathological states and the other is a unique non-invasive method to map functional neuronal synchronization.


back to top

1. Changes in neuronal synchronization along and within the basal ganglia cortex circuit are essential to understand Parkinson's disease and might be used for early diagnosis, objective definition, and follow up for future therapeutic methods.

2. Our new non-invasive Radial Correlation Contrast (RCC) MRI Method [1],  which measures changes in local neuronal synchronization by grouping neighboring voxels in relation to their temporal cross-correlation, is sensitive enough to detect such changes. 

1. Establish a non-invasive MRI technique that estimates the level of neuronal synchronization.

2. Use this technique to compare 6-OHDA (a Parkinson's diseases rat model) and control rats at various locations expected to have different neuronal activity/synchronization.

3. Apply this test on PD patients and controls to evaluate if indeed this non-invasive measure can be use for diagnostic and disease follow up.

Animal studies: High spatial resolution fMRI BOLD contrast data (156 x 156 x 1000m3) were collected from 12 rat brains subjected to forepaw stimulation (6 6-OHDA and 6 shams).  Amplitude and phase RCC maps were calculated for the OFF and the ON time-segments separately. Using the rat brain atlas, the forepaw sensory and motor cortex for each hemisphere were defined and comparison between groups was performed.

Human studies: Two different fMRI human studies were performed: One, focusing on the cortex (3 PD patients and 2 controls) and the other focusing on the basal ganglia nuclei (3 PD patient and one control). In both, we measured the temporal fluctuations [4] and the radial correlation contrast (RCC). Whereas the former is assumed to be a measure of activity, the latter is of synchronization.



back to top

Animal studies: We introduced the Radial Correlation Contrast (RCC) method as a non-invasive method to obtain clusters of interconnected volumes. Sensory stimulation causes increase in these cluster size and values in control rats. However, no significant differences were found in RCC values and cluster size in cortical locations between control and 6-OHDA rats.

Human studies:
1. fMRI measurements focusing on the cortex were performed on 3 PD patients and 2 controls. For comparison, region of interest (ROI) analysis based on anatomy, was done. No significant temporal fluctuation levels or RCC values in the cortex were found in agreement with animal results.
2. fMRI measurements in 3 PD patients and in one age matched control was performed on the lower brain slices containing the thalamus and basal ganglia. ROI, anatomy based, analysis was performed testing the distribution of RCC and temporal fluctuation values in each ROI. Since a Gaussian distribution could approximate each, two parameters were compared: the distribution width and center. Significant differences between PD patients and control were found in the right thalamus and in the right head of caudate. For the head of caudate the following significant differences were found:
  • The width of the temporal fluctuation during stimulation-ON, in PD was wider (p<0.0007), while its center did not change significantly.
  • The width of the RCC values during stimulation OFF and during stimulation ON was wider for the PD patients (p<0.005).
  • The RCC center during stimulation-OFF was higher.

Wider distribution of temporal fluctuation values suggests inhomogeneous activity. This for example could result from activation in isolated islands. The finding of wider RCC distribution suggests larger range of temporal coherences in PD patients. This finding agrees with the idea of isolated islands of high synchronization on one hand, and poor synchronization on the other hand. The fact that the RCC center in the head of caudate was shifted to higher values supports the idea that, on average, synchronization is increased. These results, however, are preliminary and more data is needed to support these findings.



back to top

Goelman G., Pelled G., Dodd S., and Koretsky A.  Observation of two distinct spatial-temporal BOLD clusters during sensory stimulation in rats. Neuroimage. 2007 Feb 1;34(3):1220-6.

Kipervaser Z.G., Pelled G., and Goelman G.  Statistical framework and noise sensitivity of the amplitude radial correlation contrast method. Magn Reson Med. 2007 Sep;58(3):554-61.

Goelman G.  Radial correlation contrast--a functional connectivity MRI contrast to map changes in local neuronal communication. Neuroimage. 2004 Dec;23(4):1432-9.

Pelled G. and Goelman G. Different physiological MRI noise between cortical layers. Magn Reson Med. 2004 Oct;52(4):913-6.