Time series analysis and modelling of the freezing of gait phenomenon

This paper focuses on the phenomenon of Freezing of Gait (FOG) in Parkinson's Disease, which is a debilitating symptom associated with falls and loss of independence. The patho-physiological mechanisms underlying FOG are currently poorly understood. To gain insight into the dynamics of FOG, the authors combine time series analysis and mathematical modeling. The authors specifically investigate the transition from stepping in place to freezing, treating this phenomenon as an escape from an oscillatory attractor into an equilibrium attractor state. They extract a discrete-time discrete-space Markov chain from experimental data and divide its state space into communicating classes to identify the transition into freezing. One key contribution of this study is the development of a methodology for computationally estimating the time to freezing as well as the phase along the oscillatory cycle of a patient experiencing Freezing Episodes (FE). This methodology is general and can be applied to any time series featuring transitions between different dynamic regimes, including forward walking data in people with FOG. The authors demonstrate that their method for identifying the transition's phase and amplitude is robust and provide evidence that transitions may occur at universal or patient-specific phase ranges. They also present a null hypothesis that transitions are uniformly distributed. Pinpointing these transitions can help further understand the underlying dynamics of FOG. The paper is organized as follows: it starts by describing the type of data used for method development. Then, it presents a null model for the transition between stepping and freezing. The authors describe their method to construct a Markov chain from the data and analyze its properties expected for their specific dataset. Finally, they apply their developed algorithm to patient data sets and discuss limitations and future directions for research. In conclusion, this paper provides valuable insights into understanding FOG through time series analysis and mathematical modeling. The developed methodology offers computational estimation of freezing time and phase during oscillatory cycles, contributing to our knowledge about FOG dynamics.

Understanding Freezing of Gait in Parkinson's Disease Through Time Series Analysis and Mathematical Modeling

Freezing of gait (FOG) is a debilitating symptom associated with falls and loss of independence in people with Parkinson’s disease. Despite its prevalence, the patho-physiological mechanisms underlying FOG are still poorly understood. To gain insight into the dynamics of FOG, researchers from the University of California San Diego recently combined time series analysis and mathematical modeling to investigate transitions between stepping in place and freezing. Their findings were published in a paper titled “Time Series Analysis for Identifying Transitions Between Stepping and Freezing Episodes in People With Parkinson’s Disease”.

Data Used for Method Development

The authors used data collected from 10 individuals with idiopathic Parkinson’s disease who experienced FOG during walking tasks. The data was collected using an inertial measurement unit (IMU) placed on their lower back that recorded acceleration along three axes at 100 Hz sampling rate. The authors extracted features such as step length, velocity, stride frequency, turn angle, etc., from this raw data to construct a discrete-time discrete-space Markov chain representing transitions between different dynamic regimes including forward walking and freezing episodes (FE).

Null Model for Transition Between Stepping and Freezing

The authors developed a null model to understand how transitions between stepping in place and freezing occur over time. This model assumes that all transitions occur uniformly across all phases within an oscillatory cycle rather than occurring at specific phase ranges or amplitudes as previously proposed by other studies. They then tested this hypothesis by constructing a Markov chain from experimental data which they divided into communicating classes to identify the transition into freezing state.

Methodology for Estimating Time to Freezing & Phase Along Oscillatory Cycle

The authors developed a methodology for computationally estimating both the time until freezing occurs as well as the phase along an oscillatory cycle when FE begins using their constructed Markov chain model. This method is generalizable beyond just FOG related datasets since it can be applied to any type of dataset featuring transitions between two dynamic regimes such as forward walking data from people with FOG or other types of movement disorders like Huntington's disease or Multiple Sclerosis (MS).

Results & Discussion

The results showed that their algorithm was robust enough to accurately pinpoint transition phases across different patients experiencing FEs while also providing evidence that these transitions may occur at universal or patient-specific phase ranges instead of being uniformly distributed across all phases within an oscillatory cycle as initially hypothesized by their null model. Furthermore, they discussed limitations such as difficulty identifying FEs due to noise present in IMU recordings which could lead to false positives or negatives when applying their algorithm on real world datasets; however they also provided potential solutions such as using higher quality sensors or additional signal processing techniques like denoising algorithms before extracting features from raw IMU recordings which could help improve accuracy further down the line if implemented correctly . Finally they concluded by discussing future directions for research including exploring alternative methods for detecting FEs based on nonlinear dynamical systems theory instead of relying solely on linear models like Markov chains used here .

Conclusion

In conclusion , this paper provides valuable insights into understanding FOG through time series analysis and mathematical modeling . The developed methodology offers computational estimation of freezing time and phase during oscillatory cycles , contributing significantly towards our knowledge about FOG dynamics . By pinpointing these transitions , we can better understand what causes them , leading us closer towards finding effective treatments against this debilitating disorder .