Identification and Characterization of Short-Term Motor Patterns in Rest Tremor of Individuals with Parkinson's Disease.

(1) Background: The dynamics of hand tremors involve nonrandom and short-term motor patterns (STMPs). This study aimed to (i) identify STMPs in Parkinson's disease (PD) and physiological resting tremor and (ii) characterize STMPs by amplitude, persistence, and regularity. (2) Methods: This study included healthy (N = 12, 60.1 ± 5.9 years old) and PD (N = 14, 65 ± 11.54 years old) participants. The signals were collected using a triaxial gyroscope on the dorsal side of the hand during a resting condition. Data were preprocessed and seven features were extracted from each 1 s window with 50% overlap. The STMPs were identified using the clustering technique k-means applied to the data in the two-dimensional space given by t-Distributed Stochastic Neighbor Embedding (t-SNE). The frequency, transition probability, and duration of the STMPs for each group were assessed. All STMP features were averaged across groups. (3) Results: Three STMPs were identified in tremor signals (p < 0.05). STMP 1 was prevalent in the healthy control (HC) subjects, STMP 2 in both groups, and STMP3 in PD. Only the coefficient of variation and complexity differed significantly between groups. (4) Conclusion: These results can help professionals characterize and evaluate tremor severity and treatment efficacy.

Open-source data management system for Parkinson's disease follow-up.

BACKGROUND: Parkinson's disease (PD) is a neurodegenerative condition of the central nervous system that causes motor and non-motor dysfunctions. The disease affects 1% of the world population over 60 years and remains cureless. Knowledge and monitoring of PD are essential to provide better living conditions for patients. Thus, diagnostic exams and monitoring of the disease can generate a large amount of data from a given patient. This study proposes the development and usability evaluation of an integrated system, which can be used in clinical and research settings to manage biomedical data collected from PD patients. METHODS: A system, so-called Sistema Integrado de Dados Biomédicos (SIDABI) (Integrated Biomedical Data System), was designed following the model-view-controller (MVC) standard. A modularized architecture was created in which all the other modules are connected to a central security module. Thirty-six examiners evaluated the system usability through the System Usability Scale (SUS). The agreement between examiners was measured by Kendall's coefficient with a significance level of 1%. RESULTS: The free and open-source web-based system was implemented using modularized and responsive methods to adapt the system features on multiple platforms. The mean SUS score was 82.99 ± 13.97 points. The overall agreement was 70.2%, as measured by Kendall's coefficient (p < 0.001). CONCLUSION: According to the SUS scores, the developed system has good usability. The system proposed here can help researchers to organize and share information, avoiding data loss and fragmentation. Furthermore, it can help in the follow-up of PD patients, in the training of professionals involved in the treatment of the disorder, and in studies that aim to find hidden correlations in data.

On the use of histograms of oriented gradients for tremor detection from sinusoidal and spiral handwritten drawings of people with Parkinson's disease.

Parkinson's disease (PD), whose cardinal signs are tremor, rigidity, bradykinesia, and postural instability, gradually reduces the quality of life of the patient, making early diagnosis and follow-up of the disorder essential. This study aims to contribute to the objective evaluation of tremor in PD by introducing and assessing histograms of oriented gradients (HOG) to the analysis of handwriting sinusoidal and spiral patterns. These patterns were digitized and collected from handwritten drawings of people with PD (n = 20) and control healthy individuals (n = 20). The HOG descriptor was employed to represent relevant information from the data classified by three distinct machine-learning methods (random forest, k-nearest neighbor, support vector machine) and a deep learning method (convolutional neural network) to identify tremor in participants with PD automatically. The HOG descriptor allowed for the highest discriminating rates (accuracy 83.1%, sensitivity 85.4%, specificity 80.8%, area under the curve 91%) on the test set of sinusoidal patterns by using the one-dimensional convolutional neural network. In addition, ANOVA and Tukey analysis showed that the sinusoidal drawing is more appropriate than the spiral pattern, which is the most common drawing used for tremor detection. This research introduces a novel and alternative way of quantifying and evaluating tremor by means of handwritten drawings.