JMIR Publications

ABSTRACT

Background:

Background:

With the rapid development of digital biomarkers in Parkinson’s disease (PD) research, it has become increasingly important to explore the current research trends and key areas of focus.

Objective:

Objective:

This study aims to comprehensively evaluate the current status, hotspots, future trends of global Parkinson’s digital biomarker research, and provide a systematic review of deep learning models for gait freezing (FOG) digital biomarkers.

Methods:

Methods:

This study employs bibliometric analysis based on the Web of Science Core Collection to conduct a comprehensive analysis of the multi-dimensional landscape of Parkinson’s digital biomarkers. After identifying research hotspots, the study also follows the PRISMA-ScR guidelines for a scope review of deep learning models for FOG from five databases: Web of Science, PubMed, IEEE Xplore, Embase, and Google Scholar.

Results:

Results:

A total of 750 studies were included in the bibliometric analysis, and 39 studies were included in the scope review. The analysis revealed an increasing number of related publications; 3,700 researchers were involved in these studies, with the highest annual participation rate of 66% from scholars with a background in neurology. The United States contributed the most research, with the largest number of participating institutions (210). In the study of deep learning models for gait freezing, the average accuracy of the models was 0.92, sensitivity was 0.87, specificity was 0.89, and AUC was 0.91. Additionally, 31 studies indicated that the best models were primarily convolutional neural networks (CNNs) or CNN-based architectures

Conclusions:

Conclusion: Research on digital biomarkers for PD is currently at a stable stage of development, with widespread global interest from countries, institutions, and researchers. However, challenges remain, including insufficient interdisciplinary and inter-institutional collaboration, as well as a lack of corporate funding for related projects. Current research trends primarily focus on motor-related studies, particularly gait freezing monitoring. However, deep learning models for gait freezing still lack external validation and standardized performance reporting. Future research will likely progress toward deeper applications of artificial intelligence, enhanced inter-institutional collaboration, comprehensive analysis of different data types, and the exploration of digital biomarkers for a broader range of Parkinson’s symptoms.