Research

Research focus

I am currently working on the detection of Parkinson’s disease (PD) with Electroencephalography (EEG) and Local field potentials (LFP) from human and rodent models of PD and development of a neural mass model of cortico-basal ganglia to simulate the EEG and LFP responses in PD with theoretical analysis.


Recent updates

Linear predictive coding distinguishes spectral EEG features of Parkinson's disease

We have developed and validated a novel EEG-based signal processing approach to distinguish PD and control patients: Linear-predictive-coding EEG Algorithm for PD (LEAPD). This method efficiently encodes EEG time series into features that can detect PD in a computationally fast manner amenable to real time applications.

Highlights

•A novel machine-learning approach to diagnose Parkinson's disease with EEG.•Cross-validation and out-of-sample tests yield more than 85% accuracy.•Outperforms other state-of-the-art EEG methods.•Computationally efficient and amenable to real-time implementation.

Original Research ARTICLE

Parkinsonism & Related Disorders. Vol 79 | https://doi.org/10.1016/j.parkreldis.2020.08.001

Dataset and Code : https://narayanan.lab.uiowa.edu/article/datasets

Linear Predictive Approaches Separate Field Potentials in Animal Model of Parkinson's Disease

We test an approach based on linear predictive coding (LPC), which fits autoregressive (AR) models to time-series data. Parameters of these AR models can be calculated by fast algorithms in real time. We compare LFPs from the striatum in an animal model of PD with dopamine depletion in the absence and presence of the dopamine precursor levodopa, which is used to treat motor symptoms of PD. We show that in dopamine-depleted mice a first order AR model characterized by a single LPC parameter obtained by LFP sampling at 1 kHz for just 1 min can distinguish between levodopa-treated and saline-treated mice and outperform current methods.

Original Research ARTICLE

Front. Neurosci., 24 April 2020 | https://doi.org/10.3389/fnins.2020.00394

Dataset and Code : https://narayanan.lab.uiowa.edu/article/datasets

Linear Predictive Approaches Separate Field Potentials in Animal Model of Parkinson's Disease

Parkinson’s disease (PD) can cause significant changes in cortical and subcortical brain activity which are captured by electroencephalography (EEG) or intracranial recordings of local field potentials (LFP). Such signals can guide adaptive deep-brain stimulation (aDBS) as part of PD therapy which requires the identification of triggers of neuronal activity dependent on real-time monitoring and analysis. We test a novel approach based on linear predictive coding (LPC) to detect PD in human and animal model which meets those requirements.

Highlights

•A feature extraction method based on Linear Predictive Coding for detecting absence and presence of levodopa in animal model of PD using a single parameter.•A LPC based method for generating a PD detection score to diagnose human PD subjects with good accuracy.

Poster Presentation

Neuromodulation: Inputs, Outputs and Outcomes 2019 | Iowa Neuroscience Institute Workshop | Spt 2019

Apparatus, systems and methods for diagnosing parkinsons disease from electroencephalography data

United States Patent Application No. 17/020,432

The disclosed apparatus, systems and methods relate to diagnosing Parkinson's disease from electroencephalography (EEG) data. Embodiments herein have practical applications, including diagnosing Parkinson's disease. The methods and systems of the various implementations herein generate a diagnostic index which reflects the probability of the patient having Parkinson's disease. It uses a novel feature extraction method based on Linear Predictive Coding (LPC) which is used to extract Parkinson's disease related features from EEG recordings of the patient and a novel classification method based on Principal Component Analysis (PCA) is used to calculate the diagnostic index from these features.


patents.google.com/patent/US20210076962A1/en