Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Advancing Parkinson's Disease Progression Prediction: Comparing Long Short-Term Memory Networks and Kolmogorov-Arnold Networks

Abhinav Roy, Bhavesh Gyanchandani, Aditya Oza, Abhishek Sharma

TL;DR

This study tackles predicting Parkinson's disease progression by comparing Long Short-Term Memory (LSTM) networks and Kolmogorov-Arnold Networks (KAN) using longitudinal $MDS-UPDRS$ scores and CSF-MS proteomics from 248 patients. The methodology combines dataset description, careful preprocessing, feature selection via correlation and Kernel Density Estimation, and parallel training of LSTM and KAN models to forecast future UPDRS trajectories. Results show that KAN achieves superior accuracy across $RMSE$, $MSE$, and $SMAPE$ metrics, though with longer training times, indicating a favorable trade-off when prediction quality is prioritized. Overall, the work demonstrates the potential of KAN for clinical time-series forecasting in PD and highlights AI's role in enhancing disease management and patient outcomes.

Abstract

Parkinson's Disease (PD) is a degenerative neurological disorder that impairs motor and non-motor functions, significantly reducing quality of life and increasing mortality risk. Early and accurate detection of PD progression is vital for effective management and improved patient outcomes. Current diagnostic methods, however, are often costly, time-consuming, and require specialized equipment and expertise. This work proposes an innovative approach to predicting PD progression using regression methods, Long Short-Term Memory (LSTM) networks, and Kolmogorov Arnold Networks (KAN). KAN, utilizing spline-parametrized univariate functions, allows for dynamic learning of activation patterns, unlike traditional linear models. The Movement Disorder Society-Sponsored Revision of the Unified Parkinson's Disease Rating Scale (MDS-UPDRS) is a comprehensive tool for evaluating PD symptoms and is commonly used to measure disease progression. Additionally, protein or peptide abnormalities are linked to PD onset and progression. Identifying these associations can aid in predicting disease progression and understanding molecular changes. Comparing multiple models, including LSTM and KAN, this study aims to identify the method that delivers the highest metrics. The analysis reveals that KAN, with its dynamic learning capabilities, outperforms other approaches in predicting PD progression. This research highlights the potential of AI and machine learning in healthcare, paving the way for advanced computational models to enhance clinical predictions and improve patient care and treatment strategies in PD management.

Advancing Parkinson's Disease Progression Prediction: Comparing Long Short-Term Memory Networks and Kolmogorov-Arnold Networks

TL;DR

This study tackles predicting Parkinson's disease progression by comparing Long Short-Term Memory (LSTM) networks and Kolmogorov-Arnold Networks (KAN) using longitudinal scores and CSF-MS proteomics from 248 patients. The methodology combines dataset description, careful preprocessing, feature selection via correlation and Kernel Density Estimation, and parallel training of LSTM and KAN models to forecast future UPDRS trajectories. Results show that KAN achieves superior accuracy across , , and metrics, though with longer training times, indicating a favorable trade-off when prediction quality is prioritized. Overall, the work demonstrates the potential of KAN for clinical time-series forecasting in PD and highlights AI's role in enhancing disease management and patient outcomes.

Abstract

Parkinson's Disease (PD) is a degenerative neurological disorder that impairs motor and non-motor functions, significantly reducing quality of life and increasing mortality risk. Early and accurate detection of PD progression is vital for effective management and improved patient outcomes. Current diagnostic methods, however, are often costly, time-consuming, and require specialized equipment and expertise. This work proposes an innovative approach to predicting PD progression using regression methods, Long Short-Term Memory (LSTM) networks, and Kolmogorov Arnold Networks (KAN). KAN, utilizing spline-parametrized univariate functions, allows for dynamic learning of activation patterns, unlike traditional linear models. The Movement Disorder Society-Sponsored Revision of the Unified Parkinson's Disease Rating Scale (MDS-UPDRS) is a comprehensive tool for evaluating PD symptoms and is commonly used to measure disease progression. Additionally, protein or peptide abnormalities are linked to PD onset and progression. Identifying these associations can aid in predicting disease progression and understanding molecular changes. Comparing multiple models, including LSTM and KAN, this study aims to identify the method that delivers the highest metrics. The analysis reveals that KAN, with its dynamic learning capabilities, outperforms other approaches in predicting PD progression. This research highlights the potential of AI and machine learning in healthcare, paving the way for advanced computational models to enhance clinical predictions and improve patient care and treatment strategies in PD management.
Paper Structure (14 sections, 3 equations, 7 figures, 7 tables)

This paper contains 14 sections, 3 equations, 7 figures, 7 tables.

Table of Contents

  1. Introduction
  2. Related Works
  3. Methodology
  4. Dataset description
  5. Data preprocessing
  6. Feature selection
  7. Correlation Matrix
  8. Kernel Density Estimate
  9. Model training
  10. Results and discussion
  11. Evaluation metrics
  12. Evaluation of model
  13. Comparative Analysis
  14. Conclusion

Figures (7)

  • Figure 1: Proposed Model with KAN and LSTM
  • Figure 2: Correlation matrix for merged dataset
  • Figure 3: Density plots for five different variables peptides against total UPDRS
  • Figure 4: Structures of our models
  • Figure 5: Plot of train and validation loss of LSTM
  • ...and 2 more figures