Eine Recherche mit dem KI Werkzeug CLAUDE bestätigt, was regelmäßige Beobachtung der Forschungsergebnisse und medizinischen Fortschritte im Bereich der tiefen Hirnstimulation (Deep Brain Stimulation, DBS) in den letzen gut 10 Jahren nachhaltig erkennen ließen: das von A. Husch im Rahmen seiner Promotion hier entwickelte Werkzeug PaCER zur präzisen und automatischen Lokalisierung der Stimulationselektroden wird weltweit als Referenz eingesetzt. CLAUDE fasst diese Ergebnisse wie folgt zusammen:

A Decade of PaCER for DBS Electrode Reconstruction (2014-2024): An AI-Conducted Survey of Development, Validation, and Application

Abstract

Background: Developed between 2014-2016 and published in 2017, the Precise and Convenient Electrode Reconstruction (PaCER) algorithm has emerged as a pivotal tool for automated deep brain stimulation (DBS) electrode localization. This AI-conducted systematic review synthesizes a decade of development and application (2014-2024) to assess PaCER's validation, clinical adoption, and scientific impact in the DBS research community.

Methods: An artificial intelligence-assisted literature search was conducted across major scientific databases using web search capabilities. Papers were identified that either validated PaCER's accuracy, utilized the algorithm for clinical or research applications, or integrated it into broader neuroimaging pipelines. Key metrics analyzed included validation accuracy, multi-center adoption, electrode types supported, and integration into established software platforms.

Results: Precursor work on electrode displacement and deformation assessment was published in 2015 (Husch et al., Bildverarbeitung für die Medizin), establishing the preprocessing pipeline foundations. The original PaCER validation study (2017) demonstrated unprecedented phantom accuracy with trajectory reconstruction errors below 100 μm (0.046 ± 0.025 mm) across 44 electrodes from two centers. Independent validation on 111 participants confirmed clinical utility with median localization errors of 0.191 mm, well below typical imaging voxel resolution (1 mm³). PaCER has been integrated into Lead-DBS, the leading open-source DBS imaging platform, contributing to over 500 peer-reviewed publications. The algorithm has been employed across diverse applications including: symptom-specific network mapping in cohorts of 237+ patients, sweet spot identification for motor and non-motor outcomes, postoperative stimulation management, brain shift analysis, and multi-center clinical trials. PaCER supports multiple electrode manufacturers (Medtronic 3387/3389, Boston Scientific Vercise) and accommodates varying CT scan protocols, with robust performance across different slice thicknesses and scanner types.

Key Advantages: Literature review identified PaCER's distinctive capabilities: (1) fully automated operation eliminating user subjectivity, (2) accurate reconstruction of curved electrode trajectories accounting for brain shift, (3) validated sub-millimeter accuracy exceeding manual expert localization, (4) seamless integration into existing MATLAB-based workflows, and (5) open-source availability fostering widespread adoption.

Impact and Adoption: PaCER has become the standard automated pre-reconstruction method in Lead-DBS and is utilized in major research applications spanning Parkinson's disease, essential tremor, dystonia, and neuropsychiatric indications. The algorithm has enabled large-scale connectivity analyses, directional lead optimization studies, and is incorporated into educational resources (Lead-Tutor) for training the next generation of DBS researchers.

Limitations: The review identified ongoing challenges including: primary limitation to CT imaging modalities, reduced accuracy with low-resolution scans (>1 mm slice thickness requiring fallback detection strategies), need for manual refinement in cases with extreme electrode angles, and limited validation for newer directional lead designs beyond initial supported models.

Conclusions: Over the past decade from initial development (2014-2016) through widespread application (2017-2024), PaCER has transformed DBS electrode reconstruction from a time-consuming, subjective manual process into a rapid, objective, and highly accurate automated procedure. Its phantom-validated sub-millimeter accuracy, successful multi-center validation, and integration into the most widely-used DBS imaging platform have established it as an essential tool in modern DBS research and clinical assessment. The algorithm's open-source nature and robust performance have democratized access to high-quality electrode localization, enabling sophisticated network analyses and personalized stimulation strategies that were previously impractical. Future developments should focus on extending capabilities to MRI-based reconstruction, supporting emerging electrode designs, and further streamlining clinical workflow integration.

Keywords: Deep brain stimulation, electrode localization, automated reconstruction, image processing, PaCER, Lead-DBS, validation study, brain shift, phantom accuracy