A Study of Upper-Limb Motion using Kinematic Measures for Clinical Assessment of Cerebellar Ataxia.

Cerebellar Ataxia (CA) is a group of diseases affecting the cerebellum, which is responsible for movement coordination. It causes uncoordinated movements and can also impact balance, speech, and eye movements. There are no approved disease-modifying medications for CA, so clinical studies to assess potential treatments are crucial. These studies require robust, objective measurements of CA severity to reflect changes in the progression of the disease due to medication. In recent years, studies have used kinematic measures to evaluate CA severity, but the current method relies on subjective clinical observations and is insufficient for telehealth. There is a need for a non-intrusive system that can monitor people with CA regularly to better understand the disease and develop an automated assessment system. In this study, we analyzed kinematic measures of upper-limb movements during a ballistic tracking test, which primarily involves movements at the shoulder joint. We aimed to understand the challenges of identifying CA and evaluating its severity when measuring such movements. Statistical features of the kinematic signals were used to develop machine learning models for classification and regression. The Gradient Boosting Classifier model had a maximum accuracy of 74%, but the models had low specificity and performed poorly in regression, suggesting that kinematic measures from shoulder-dominated movements during ballistic tracking are not as viable for CA assessment as other measures.

An Enhanced Classification Framework for Limited IoHT Time Series Data Using Ensemble Deep Learning and Image Encoding.

Recent studies have illuminated the potential of harnessing the power of Deep Learning (DL) and the Internet of Health Things (IoHT) to detect a variety of disorders, particularly among patients in the middle to later stages of the disease. The utilization of time series data has proven to be a valuable asset in this endeavour. However, the development of effective DL architectures for time series classification with limited data remains a critical gap in the field. Although some studies have explored this area, it is still an understudied and undervalued topic. Thus, there is a crucial need to address this gap and provide insights into designing effective architectures for time series classification with limited data, specifically in the context of healthcare-related time series data for rare diseases. The goal of this study is to investigate the possibility of making accurate predictions with a smaller time series dataset by using an Ensemble DL architecture. This framework is composed of a deep CNN model and transfer learning approaches like ResNet and MobileNet. The ensemble model proposed in this study was supplied with 3D images that were generated from time series data by using Recurrence Plot (RP), Gramian Angular Field (GAF), and Fuzzy Recurrence Plot (FRP) as the transformation techniques. The proposed method has shown promising classification accuracy, even when applied to a small dataset, and surpassed the performance of other state-of-the-art methods when tested on the ECG5000 dataset.Clinical relevance- The proposed deep learning architecture is capable of effectively handling limited clinical time series datasets, enabling the construction of robust models and accurate predictions.

Machine Learning-Based Scoring System to Predict the Risk and Severity of Ataxic Speech Using Different Speech Tasks.

The assessment of speech in Cerebellar Ataxia (CA) is time-consuming and requires clinical interpretation. In this study, we introduce a fully automated objective algorithm that uses significant acoustic features from time, spectral, cepstral, and non-linear dynamics present in microphone data obtained from different repeated Consonant-Vowel (C-V) syllable paradigms. The algorithm builds machine-learning models to support a 3-tier diagnostic categorisation for distinguishing Ataxic Speech from healthy speech, rating the severity of Ataxic Speech, and nomogram-based supporting scoring charts for Ataxic Speech diagnosis and severity prediction. The selection of features was accomplished using a combination of mass univariate analysis and elastic net regularization for the binary outcome, while for the ordinal outcome, Spearman's rank-order correlation criterion was employed. The algorithm was developed and evaluated using recordings from 126 participants: 65 individuals with CA and 61 controls (i.e., individuals without ataxia or neurotypical). For Ataxic Speech diagnosis, the reduced feature set yielded an area under the curve (AUC) of 0.97 (95% CI 0.90-1), the sensitivity of 97.43%, specificity of 85.29%, and balanced accuracy of 91.2% in the test dataset. The mean AUC for severity estimation was 0.74 for the test set. The high C-indexes of the prediction nomograms for identifying the presence of Ataxic Speech (0.96) and estimating its severity (0.81) in the test set indicates the efficacy of this algorithm. Decision curve analysis demonstrated the value of incorporating acoustic features from two repeated C-V syllable paradigms. The strong classification ability of the specified speech features supports the framework's usefulness for identifying and monitoring Ataxic Speech.

A Sensor-Based Comprehensive Objective Assessment of Motor Symptoms in Cerebellar Ataxia.

Human observer-based assessments of Cerebellar Ataxia (CA) are subjective and are often inadequate to track mild motor symptoms. This study examines the potential use of a comprehensive sensor-based approach for objective evaluation of CA in five domains (speech, upper limb, lower limb, gait and balance) through the instrumented versions of nine bedside neurological tests. A total of twenty-three participants diagnosed with CA to varying degrees and eleven healthy controls were recruited. Data was collected using wearable inertial sensors and Kinect camera. In our study, an optimal feature subset based on feature importance in the Random Forest classifier model demonstrated an impressive performance accuracy of 97% (F1 score = 95.2%) for CA-control discrimination. Our experimental findings also indicate that the Romberg test contributed most, followed by the peripheral tests, while the Gait test contributed least to the classification. Sensor-based approaches, therefore, have the potential to complement existing clinical assessment techniques, offering advantages in terms of consistency, objectivity and informed clinical decision-making.

Objective Assessment of Cerebellar Ataxia: A Comprehensive and Refined Approach.

Parametric analysis of Cerebellar Ataxia (CA) could be of immense value compared to its subjective clinical assessments. This study focuses on a comprehensive scheme for objective assessment of CA through the instrumented versions of 9 commonly used neurological tests in 5 domains- speech, upper limb, lower limb, gait and balance. Twenty-three individuals diagnosed with CA to varying degrees and eleven age-matched healthy controls were recruited. Wearable inertial sensors and Kinect camera were utilised for data acquisition. Binary and multilabel discrimination power and intra-domain relationships of the features extracted from the sensor measures and the clinical scores were compared using Graph Theory, Centrality Measures, Random Forest binary and multilabel classification approaches. An optimal subset of 13 most important Principal Component (PC) features were selected for CA-control classification. This classification model resulted in an impressive performance accuracy of 97% (F1 score = 95.2%) with Holmesian dimensions distributed as 47.7% Stability, 6.3% Timing, 38.75% Accuracy and 7.24% Rhythmicity. Another optimal subset of 11 PC features demonstrated an F1 score of 84.2% in mapping the total 27 PC across 5 domains during CA multilabel discrimination. In both cases, the balance (Romberg) test contributed the most (31.1% and 42% respectively), followed by the peripheral tests whereas gait (Walking) test contributed the least. These findings paved the way for a better understanding of the feasibility of an instrumented system to assist informed clinical decision-making.

Quantitative Assessment of Speech in Cerebellar Ataxia Using Magnitude and Phase Based Cepstrum.

The clinical assessment of speech abnormalities in Cerebellar Ataxia (CA) is time-consuming and inconsistent. We have developed an automated objective system to quantify CA severity and thereby facilitate remote monitoring and optimisation of therapeutic interventions. A quantitative acoustic assessment could prove to be a viable biomarker for this purpose. Our study explores the use of phase-based cepstral features extracted from the modified group delay function as a complement to the features obtained from the magnitude cepstrum. We selected a combination of 15 acoustic measurements using RELIEF feature selection algorithm during the feature optimisation process. These features were used to segregate ataxic speakers from normal speakers (controls) and objectively assess them based on their severity. The effectiveness of our study has been experimentally evaluated through a clinical study involving 42 patients diagnosed with CA and 23 age-matched controls. A radial basis function kernel based support vector machine (SVM) classifier achieved a classification accuracy of 84.6% in CA-Control discrimination [area under the ROC curve (AUC) of 0.97] and 74% in the modified 3-level CA severity estimation (AUC of 0.90) deduced from the clinical ratings. The strong classification ability of selected features and the SVM model supports this scheme's suitability for monitoring CA related speech motor abnormalities.

Automated tongue-twister phrase-based screening for Cerebellar Ataxia using Vocal tract Biomarkers.

Cerebellar Ataxia (CA) is a neurological condition that leads to uncoordinated muscle movements, even affecting the production of speech. Effective biomarkers are necessary to produce an objective decision-making support tool for early diagnosis of CA in non-clinical environments. This paper investigates the reliability and effectiveness of vocal tract acoustic biomarkers for assessing CA speech. These features were tested on a database consisting of 52 clinically rated tongue-twister phrase 'British Constitution' and its 4 consonant-vowel (CV) excerpts /ti/, /ti/', /tu/, /tion/ acquired from 30 ataxic patients and 22 healthy controls. Such a marker could be applied to objectively assess the severity of CA from a simple speaking test, contributing to the possibility of being translated into a computer based automatic module to screen the disease from the speech. All the vocal tract features explored in this study were statistically significant using Kolmogorov-Smirnov test at 5% level in distinguishing healthy and CA speech. Several machine learning classifiers with 5-fold cross-validations were implemented on the vocal features. It was observed that the intensity ratios corresponding to the 4 C-V excerpts in CA group showed an increased variability and produced the best classification accuracy of 84.6% using KNN classifier. Results motivate the use of vocal tract features for monitoring CA speech.

Quantitative Assessment of Syllabic Timing Deficits in Ataxic Dysarthria.

Parametric analysis of Cerebellar Dysarthria (CD) may be valuable and more informative compared to its clinical assessment. A quantifiable estimation of the timing deficits in repeated syllabic utterance is described in the current study. Thirty-five individuals were diagnosed with cerebellar ataxia to varying degrees and twenty-six age-matched healthy controls were recruited. To automatically detect the local maxima of each syllable in the recorded speech files, a topographic prominence incorporated concept is designed. Subsequently, four acoustic features and eight corresponding parametric measurements are extracted to identify articulatory deficits in ataxic dysarthria. A comparative study on the behaviour of these measures for dysarthric and non-dysarthric subjects is presented in this paper. The results are further explored using a dimensionreduction tool (Principal Component Analysis) to emphasize variation and bring out the strongest discriminating patterns in our feature dataset.