Quantifying gait and postural control adds valuable information that aids in understanding neurological conditions where motor symptoms predominate and cause considerable functional impairment. Disease-specific clinical scales exist; however, they are often susceptible to subjectivity, and can lack sensitivity when identifying subtle gait and postural impairments in prodromal cohorts and longitudinally to document disease progression. Numerous devices are available to objectively quantify a range of measurement outcomes pertaining to gait and postural control; however, efforts are required to standardise and harmonise approaches that are specific to the neurological condition and clinical assessment. Tools are urgently needed that address a number of unmet needs in neurological practice. Namely, these include timely and accurate diagnosis; disease stratification; risk prediction; tracking disease progression; and decision making for intervention optimisation and maximising therapeutic response (such as medication selection, disease staging, and targeted support). Using some recent examples of research across a range of relevant neurological conditions—including Parkinson’s disease, ataxia, and dementia—we will illustrate evidence that supports progress against these unmet clinical needs. We summarise the novel ‘big data’ approaches that utilise data mining and machine learning techniques to improve disease classification and risk prediction, and conclude with recommendations for future direction.

Parkinson’s disease (PD) is the second most common neurodegenerative disease; gait impairments are typical and are associated with increased fall risk and poor quality of life. Gait is potentially a useful biomarker to help discriminate PD at an early stage, however the optimal characteristics and combination are unclear. In this study, we used machine learning (ML) techniques to determine the optimal combination of gait characteristics to discriminate people with PD and healthy controls (HC). 303 participants (119 PD, 184 HC) walked continuously around a circuit for 2-minutes at a self-paced walk. Gait was quantified using an instrumented mat (GAITRite) from which 16 gait characteristics were derived and assessed. Gait characteristics were selected using different ML approaches to determine the optimal method (random forest with information gain and recursive features elimination (RFE) technique with support vector machine (SVM) and logistic regression). Five clinical gait characteristics were identified with RFE-SVM (mean step velocity, mean step length, step length variability, mean step width, and step width variability) that accurately classified PD. Model accuracy for classification of early PD ranged between 73–97% with 63–100% sensitivity and 79–94% specificity. In conclusion, we identified a subset of gait characteristics for accurate early classification of PD. These findings pave the way for a better understanding of the utility of ML techniques to support informed clinical decision-making.

Considering the challenge of population ageing and the substantial health problem among the elderly population from falls, the purpose of this study was to verify whether it is possible to distinguish accurately between older fallers and non-fallers, based on data from wearable inertial sensors collected during a specially designed test battery. A comprehensive but practical test battery using 5 wearable inertial sensors for multifactorial fall risk assessment was designed. This was followed by an experimental study on 196 community-dwelling Korean older women, categorized as fallers (N₁ = 82) and non-fallers (N₂ = 114) based on prior history of falls. Six machine learning models (logistic regression, naïve bayes, decision tree, random forest, boosted tree and support vector machine) were proposed for faller classification. Results indicated that compared with non-fallers, fallers performed significantly worse on the test battery. In addition, the application of sensor data and support vector machine for faller classification achieved an overall accuracy of 89.4% with 92.7% sensitivity and 84.9% specificity. These findings suggest that wearable inertial sensor based systems show promise for elderly fall risk assessment, which could be implemented in clinical practice to identify “at-risk” individuals reliably to promote proactive fall prevention.

Objective: Gait may be a useful biomarker that can be objectively measured with wearable technology to classify Parkinson's disease (PD). This study aims to: (i) comprehensively quantify a battery of commonly utilized gait digital characteristics (spatiotemporal and signal-based), and (ii) identify the best discriminative characteristics for the optimal classification of PD. Methods: Six partial least square discriminant analysis (PLS-DA) models were trained on subsets of 210 characteristics measured in 142 subjects (81 people with PD, 61 controls (CL)). Results: Models accuracy ranged between 70.42-88.73% (AUC: 78.4-94.5%) with a sensitivity of 72.84-90.12% and a specificity of 60.3-86.89%. Signal-based digital gait characteristics independently gave 87.32% accuracy. The most influential characteristics in the classification models were related to root mean square values, power spectral density, step velocity and length, gait regularity and age. Conclusions: This study highlights the importance of signal-based gait characteristics in the development of tools to help classify PD in the early stages of the disease.

Parkinson’s disease (PD) is a common neurodegenerative disorder resulting in a range of mobility deficits affecting gait, balance and turning. In this paper, we present: (i) the development and validation of an algorithm to detect turns during gait; (ii) a method to extract turn characteristics; and (iii) the classification of PD using turn characteristics. Thirty-seven people with PD and 56 controls performed 180-degree turns during an intermittent walking task. Inertial measurement units were attached to the head, neck, lower back and ankles. A turning detection algorithm was developed and validated by two raters using video data. Spatiotemporal and signal-based characteristics were extracted and used for PD classification. There was excellent absolute agreement between the rater and the algorithm for identifying turn start and end (ICC ≥ 0.99). Classification modeling (partial least square discriminant analysis (PLS-DA)) gave the best accuracy of 97.85% when trained on upper body and ankle data. Balanced sensitivity (97%) and specificity (96.43%) were achieved using turning characteristics from the neck, lower back and ankles. Turning characteristics, in particular angular velocity, duration, number of steps, jerk and root mean square distinguished mild-moderate PD from controls accurately and warrant future examination as a marker of mobility impairment and fall risk in PD.

Falls are the leading cause of mortality, morbidity and poor quality of life in older adults with or without neurological conditions. Applying machine learning (ML) models to gait analysis outcomes offers the opportunity to identify individuals at risk of future falls. The aim of this study was to determine the effect of different data pre-processing methods on the performance of ML models to classify neurological patients who have fallen from those who have not for future fall risk assessment. Gait was assessed using wearables in clinic while walking 20 m at a self-selected comfortable pace in 349 (159 fallers, 190 non-fallers) neurological patients. Six different ML models were trained on data pre-processed with three techniques such as standardisation, principal component analysis (PCA) and path signature method. Fallers walked more slowly, with shorter strides and longer stride duration compared to non-fallers. Overall, model accuracy ranged between 48% and 98% with 43–99% sensitivity and 48–98% specificity. A random forest (RF) classifier trained on data pre-processed with the path signature method gave optimal classification accuracy of 98% with 99% sensitivity and 98% specificity. Data pre-processing directly influences the accuracy of ML models for the accurate classification of fallers. Using gait analysis with trained ML models can act as a tool for the proactive assessment of fall risk and support clinical decision-making.