Session Time: 1:45pm-3:15pm
Location: Hall 3FG
Objective: To evaluate the importance of dystonia features that can predict concurrent tremor prevalence and tremor irregularity. (2) To cluster dystonia cases based on important data-driven features using a large, multi-institutional cohort of 2362 patients, and adopting state of the art feature selection and clustering methods of machine learning.
Background: Dystonia is often accompanied by regular or irregular tremor. The dystonia features that increase the likelihood of concurrent tremor are unclear. Dystonia traits associated with irregular tremor are also not well defined.
Methods: We used a permutation-based feature selection algorithm to evaluate various dystonia attributes and select the relevant ones to be used in predicting tremor prevalence and irregularity. We performed clustering analyses to group dystonia patients into clusters with similar characteristics using an agglomerative hierarchical clustering algorithm.
Results: The first feature selection analysis indicated that body part affected by dystonia provides the most useful information for predicting tremor prevalence. Duration of dystonia, total Global Dystonia Rating Scale score, and age at dystonia onset also play a significant role in determining whether dystonia and tremor coexist. With these parameters, a random forest classifier (RFC) was able to classify a test data set with 69% accuracy. The clustering analysis yielded 4 distinct clusters with 16%, 31%, 62% and 67% tremor prevalence rates. The second feature selection analysis showed that tremor irregularity is sensitive to the extent to which dystonia and tremor locations overlap. Investigator site is also an important feature that discriminates between regular and irregular tremor. RFC was able to predict irregularity with 79% test accuracy, and clustering analysis formed 4 distinct clusters with 28%, 76%, 79%, and 84% irregular tremor rates. Handedness, gender, and race were found unimportant for predicting tremor prevalence and irregularity in dystonia.
Conclusions: We identified the most relevant dystonia traits for predicting concurrent tremor prevalence and irregularity using modern machine learning methods. Our results also exemplify the benefit of these methods in understanding the relationship between subtypes of heterogeneous movement disorders.
To cite this abstract in AMA style:S. Balta Beylergil, L. Scorr, A. Cotton, H. Jinnah, A. Shaikh. A machine learning approach to determine the important patient characteristics for tremor prevalence and tremor irregularity in dystonia [abstract]. Mov Disord. 2018; 33 (suppl 2). https://www.mdsabstracts.org/abstract/a-machine-learning-approach-to-determine-the-important-patient-characteristics-for-tremor-prevalence-and-tremor-irregularity-in-dystonia/. Accessed December 11, 2023.
« Back to 2018 International Congress
MDS Abstracts - https://www.mdsabstracts.org/abstract/a-machine-learning-approach-to-determine-the-important-patient-characteristics-for-tremor-prevalence-and-tremor-irregularity-in-dystonia/