Predicting Improvement in Parkinson's Disease Patients using Location of Stimulation
Poster No:
3
Submission Type:
Abstract Submission
Authors:
Kezia Susanto1,2, Brian Premchand1, Kai Rui Wan3, Fatimah Misbaah3, Rosa So1,2
Institutions:
1Institute for Infocomm Research (I2R), Agency for Science, Technology and Research (A*STAR), Singapore, Singapore, 2Department of Biomedical Engineering, National University of Singapore, Singapore, Singapore, 3Department of Neurosurgery, National Neuroscience Institute, Singapore, Singapore
First Author:
Kezia Susanto
Institute for Infocomm Research (I2R), Agency for Science, Technology and Research (A*STAR)|Department of Biomedical Engineering, National University of Singapore
Singapore, Singapore|Singapore, Singapore
Institute for Infocomm Research (I2R), Agency for Science, Technology and Research (A*STAR)|Department of Biomedical Engineering, National University of Singapore
Singapore, Singapore|Singapore, Singapore
Co-Author(s):
Brian Premchand
Institute for Infocomm Research (I2R), Agency for Science, Technology and Research (A*STAR)
Singapore, Singapore
Institute for Infocomm Research (I2R), Agency for Science, Technology and Research (A*STAR)
Singapore, Singapore
Rosa So
Institute for Infocomm Research (I2R), Agency for Science, Technology and Research (A*STAR)|Department of Biomedical Engineering, National University of Singapore
Singapore, Singapore|Singapore, Singapore
Institute for Infocomm Research (I2R), Agency for Science, Technology and Research (A*STAR)|Department of Biomedical Engineering, National University of Singapore
Singapore, Singapore|Singapore, Singapore
Introduction:
Parkinson's Disease (PD), which affects at least 1% of people over 60 (Zafar & Yaddanapudi, 2023), is a debilitating neurological disorder characterized by motor symptoms such as rigidity, tremor, slowness of movement (bradykinesia) and freezing of gait. While medications are commonly used to manage symptoms, they often lose effectiveness over time (Marsden, 1994; Zappia et al., 1999). Deep Brain Stimulation (DBS), a recent and proven therapy, provides significant symptom relief by implanting electrodes deep into the brain (Benabid et al., 2009). DBS typically targets the subthalamic nucleus (STN) in the basal ganglia, a region responsible for motor functions (Groiss et al., 2009; Herzog et al., 2004; van den Munckhof et al., 2021).
Despite its superior efficacy, the relationship between the precise DBS stimulation location within the STN and symptom relief remains unclear, with the selection of stimulation parameters often involving trial and error (Santaniello et al., 2018). To address this gap, this study aims to identify how stimulation location affects individual symptom improvement in PD by developing predictive models to optimize DBS protocols and enhance patient outcomes.
Despite its superior efficacy, the relationship between the precise DBS stimulation location within the STN and symptom relief remains unclear, with the selection of stimulation parameters often involving trial and error (Santaniello et al., 2018). To address this gap, this study aims to identify how stimulation location affects individual symptom improvement in PD by developing predictive models to optimize DBS protocols and enhance patient outcomes.
Methods:
This retrospective study included 16 patients from a cohort implanted at the National Neuroscience Institute, Tan Tock Seng Hospital, Singapore. All patients underwent bilateral STN DBS surgery with quadripolar electrodes (model 3389, Medtronic). Comprehensive data were collected, including pre- and postoperative MRI or CT, optimized stimulation settings, and UPDRS-III scores. One month post-surgery, UPDRS-III motor scores were assessed, and optimized stimulation parameters (activated contact(s), amplitude, and frequency) were recorded. Symptom improvement was analyzed by comparing baseline (MedOFF/STIMOFF) and stimulation conditions.
MRI/CT scans were pre-processed and normalized to determine electrode locations using Lead-DBS 3 software (Neudorfer et al., 2023). The software extracted features such as x-, y-, and z-coordinates of active contacts, distances to the STN and its subregions (motor, associative, limbic), and estimations of the VTA and VEF intersecting these regions. Each patient had 32 unique features, totaling 64 when considering both hemispheres.
Pearson correlation identified the seven most significantly associated features for each primary symptom, categorized into rigidity, tremor, bradykinesia, and freezing of gait. With left and right hemisphere features combined, 14 features per symptom were used for model building. Features were normalized and refined for machine learning, with ridge regression applied using leave-one-out and grid search cross-validation for hyperparameter tuning. Model performance was finally evaluated using mean squared error (MSE) and R-squared (R2) scores.
MRI/CT scans were pre-processed and normalized to determine electrode locations using Lead-DBS 3 software (Neudorfer et al., 2023). The software extracted features such as x-, y-, and z-coordinates of active contacts, distances to the STN and its subregions (motor, associative, limbic), and estimations of the VTA and VEF intersecting these regions. Each patient had 32 unique features, totaling 64 when considering both hemispheres.
Pearson correlation identified the seven most significantly associated features for each primary symptom, categorized into rigidity, tremor, bradykinesia, and freezing of gait. With left and right hemisphere features combined, 14 features per symptom were used for model building. Features were normalized and refined for machine learning, with ridge regression applied using leave-one-out and grid search cross-validation for hyperparameter tuning. Model performance was finally evaluated using mean squared error (MSE) and R-squared (R2) scores.
Results:
Our regression models demonstrated high predictive accuracy across UPDRS metrics. For Total-UPDRS, the model achieved an R2 of 0.780 and an MSE of 13.920, while for Rigidity-UPDRS, it achieved exceptional accuracy with an R2 of 0.976 and an MSE of 0.107. Strong performances were also observed for Tremor-UPDRS (R2 = 0.851, MSE = 1.561), Bradykinesia-UPDRS (R2 = 0.852, MSE = 4.669), and Freezing of Gait-UPDRS (R2 = 0.840, MSE = 0.277). These results highlight the strong correlation between DBS stimulation sites and symptom improvement, as visualized in Figure 2.
Conclusions:
This study highlights the potential of personalized DBS to optimize symptom relief in PD by linking stimulation locations to improvements in rigidity, tremor, and bradykinesia. With predictive models achieving high accuracy (R2 = 0.840–0.976), this work lays the foundation for tailored DBS protocols that reduce trial-and-error. Future validation on diverse cohorts could enhance generalizability and further refine personalized treatment strategies to improve patient outcomes and quality of life.
Brain Stimulation:
Deep Brain Stimulation 1
Disorders of the Nervous System:
Neurodegenerative/ Late Life (eg. Parkinson’s, Alzheimer’s) 2
Modeling and Analysis Methods:
Classification and Predictive Modeling
Image Registration and Computational Anatomy
Novel Imaging Acquisition Methods:
Anatomical MRI
Keywords:
Basal Ganglia
Machine Learning
MRI
Other - neuroimaging; deep brain stimulation; parkinson's disease
1|2Indicates the priority used for review
Abstract Information
By submitting your proposal, you grant permission for the Organization for Human Brain Mapping (OHBM) to distribute your work in any format, including video, audio print and electronic text through OHBM OnDemand, social media channels, the OHBM website, or other electronic publications and media.
The Open Science Special Interest Group (OSSIG) is introducing a reproducibility challenge for OHBM 2025. This new initiative aims to enhance the reproducibility of scientific results and foster collaborations between labs. Teams will consist of a “source” party and a “reproducing” party, and will be evaluated on the success of their replication, the openness of the source work, and additional deliverables. Click here for more information. Propose your OHBM abstract(s) as source work for future OHBM meetings by selecting one of the following options:
Please indicate below if your study was a "resting state" or "task-activation” study.
Healthy subjects only or patients (note that patient studies may also involve healthy subjects):
Was this research conducted in the United States?
Were any human subjects research approved by the relevant Institutional Review Board or ethics panel? NOTE: Any human subjects studies without IRB approval will be automatically rejected.
Were any animal research approved by the relevant IACUC or other animal research panel? NOTE: Any animal studies without IACUC approval will be automatically rejected.
Please indicate which methods were used in your research:
For human MRI, what field strength scanner do you use?
Which processing packages did you use for your study?
Provide references using APA citation style.
Groiss, S. J. (2009). Deep brain stimulation in Parkinson’s disease. Therapeutic Advances in Neurological Disorders, 2(6), 20–28. https://doi.org/10.1177/1756285609339382
Herzog, J. (2004). Most effective stimulation site in subthalamic deep brain stimulation for Parkinson’s disease. Movement Disorders, 19(9), 1050–1054. https://doi.org/10.1002/mds.20056
Marsden, C. D. (1994). Problems with long-term levodopa therapy for Parkinson’s disease. Clinical Neuropharmacology, 17 Suppl 2, S32-44.
Neudorfer, C. (2023). Lead-DBS v3.0: Mapping deep brain stimulation effects to local anatomy and global networks. NeuroImage, 268. https://doi.org/10.1016/j.neuroimage.2023.119862
Santaniello, S. (2018). Systems approaches to optimizing deep brain stimulation therapies in Parkinson’s disease. In Wiley Interdisciplinary Reviews: Systems Biology and Medicine (Vol. 10, Issue 5). Wiley-Blackwell. https://doi.org/10.1002/wsbm.1421
van den Munckhof, P. (2021). Targeting of the Subthalamic Nucleus in Patients with Parkinson’s Disease Undergoing Deep Brain Stimulation Surgery. Neurology and Therapy, 10(1), 61–73. https://doi.org/10.1007/s40120-021-00233-8
Zafar, S. (2023, August 7). Parkinson’s Disease. StatPearls. https://www.ncbi.nlm.nih.gov/books/NBK470193/
Zappia, M. (1999). Loss of long-duration response to levodopa over time in PD: implications for wearing-off. Neurology, 52(4), 763–767. https://doi.org/10.1212/wnl.52.4.763