The goal of this study is to increase magnetic resonance image quality in patients suffering from Parkinson's disease. The main question it aims to answer is: can super-resolution improve clinical magnetic resonance image quality to benefit deep brain stimulation for Parkinson's disease? Participants will receive an additional high-quality MRI scan.

Rationale: Better targeting of the subthalamic nucleus (STN) improves the outcome of deep brain stimulation (DBS) for Parkinson's disease. Yet, the accuracy of delineating the STN, and therefore the targeting, is limited by the spatial resolution of the magnetic resonance (MR) imaging. The current study aims to acquire a high resolution (HR) MR dataset, tailored to visualise the STN, to train a super-resolution model to predict HR MR images based on lower resolution MR input. This model will aid delineating the STN and improve segmentation and targeting.

Objective: To develop a deep learned super-resolution model that predicts high-resolution MR images with a peak signal-to-noise ratio of 37 decibel or higher.

Study design: Prospective observational study.

Study population: Twenty Parkinson's patients considered eligible for DBS surgery at Radboud University Medical Centre will be included.

Main study parameters/endpoints: Peak signal-to-noise ratio measured in decibels.

Inclusion Criteria:

• Patients eligible for bilateral STN-DBS-surgery for Parkinson's disease.
• Signed informed consent

Exclusion Criteria:

• Any intracranial abnormality that is not in line with Parkinson's disease progression
• Previous intracranial surgery
• Any significant medical condition that is likely to interfere with study procedures.
• Pregnancy at the time of enrollment.
• Participation in any other clinical trial (e.g. drug, device, or biologics) that interferes with this study.

In the Radboudumc 50-70 bilateral STN-DBS procedures per year are performed for PD. Before being eligible for the DBS procedure, an extensive screening is performed. Any patient that is eligible for the DBS procedure according to this screening is also eligible for this study. From the patients undergoing this procedure, 20 are needed within the current study.

The programs that have been developed, and the models that have been trained, will be published integrally and open source. Such that it is available freely, for the scientific community to use.

When the results of this study have been published, the data will be shared open source for use in other research. In this way the data can be used to the fullest. Before sharing the data, the faces will be masked. No identifiable data from subjects will be disclosed. The identity of these subjects can also not be discovered using recombination of this dataset with other datasets.