Chi-square distance kernel of the gaits for the diagnosis of Parkinson's disease

ATST-Net: A method to identify early symptoms in the upper and lower extremities of PD

Bradykinesia, a core symptom of motor disorders in Parkinson's disease (PD), is a major criterion for screening early PD patients in clinical practice. Currently, many studies have proposed automatic assessment schemes for bradykinesia in PD. However, existing schemes suffer from problems such as dependence on professional equipment, single evaluation tasks, difficulty in obtaining samples and low accuracy. This paper proposes a manual feature extraction- and neural network-based method to evaluate bradykinesia, effectively solving the problem of a small sample size. This method can automatically assess finger tapping (FT), hand movement (HM), toe tapping (TT) and bilateral foot sensitivity tasks (LA) through a unified model. Data were obtained from 120 individuals, including 93 patients with Parkinson's disease and 27 age- and sex-matched normal controls (NCs). Manual feature extraction and Attention Time Series Two-stream Networks (ATST-Net) were used for classification. Accuracy rates of 0.844, 0.819, 0.728, and 0.768 were achieved for FT, HM, TT, and LA, respectively. To our knowledge, this study is the first to simultaneously evaluate the upper and lower limbs using a unified model that has significant advantages in both model training and transfer learning.

Intra-subject enveloped multilayer fuzzy sample compression for speech diagnosis of Parkinson's disease

Machine learning-based Parkinson's disease (PD) speech diagnosis is a current research hotspot. However, existing methods use each corpus sample as the base unit for modeling. Since different corpus samples within the same subject have different sensitive speech features, it is difficult to obtain unified and stable sensitive speech features (diagnostic markers) that reflect the pathology of the whole subject. Therefore, this study aims at compressing the corpus samples within the subject to facilitate the search for diagnostic markers with high diagnostic accuracy. A two-step sample compression module (TSCM) can solve the problem above. It includes two major parts: sample pruning module (SPM) and sample fuzzy clustering mechanism (SFCMD). Based on stacking multiple TSCMs, a multilayer sample compression module (MSCM) is formed to obtain multilayer compression samples. After that, simultaneous sample/feature selection mechanism (SS/FSM) is designed for feature selection. Based on the multilayer compression samples processed by MSCM and SS/FSM, a novel ensemble learning algorithm (EMSFE) is designed with sparse fusion ensemble learning mechanism (SFELM). The proposed EMSFE is validated by visualization of extracted features and performance comparison with related algorithms. The experimental results show that the proposed algorithm can effectively extract the stable diagnostic markers by compressing the corpus samples within the subject. Furthermore, based on LOSO cross validation, the proposed algorithm with extreme learning machine (ELM) classifier can achieve the accuracy of 92.5%, 93.75% and 91.67% on three datasets, respectively. The proposed EMSFE can extract unified and stable sensitive features that accurately reflect the overall pathology of the subject, which can better meet the requirements of clinical applications.The code and datasets can be found in: https://github.com/wywwwww/EMSFE-supplementary-material.git.

Detecting early-stage Parkinson's disease from gait data

Parkinson's disease is a chronic and progressive neurodegenerative disorder with an estimated 10 million people worldwide living with PD. Since early signs are benign, many patients go undiagnosed until the symptoms get severe and the treatment becomes more difficult. The symptoms start intermittently and gradually become continuous as the disease progresses. In order to detect and classify these minute differences between gaits in early PD patients, we propose to use dynamic time warping (DTW). For a given set of gait data from a patient, the DTW algorithm computes the difference between any two gait cycles in the form of a warping path, which reveals small time differences between gait cycles. Once the time-warping information between all possible pairs of gait cycles is used as the main source of gait features, K-means clustering is used to extract the final features. These final features are fed to a simple logistic regression to easily and successfully detect early PD symptoms, which was reported as challenging using conventional statistical features. In addition, the use of DTW ensures that the obtained results are not affected by the differences in the style and speed of walking of a subject. Our approach is validated for the gait data from 83 subjects at early stages of PD, 10 subjects at moderate stages of PD, and 73 controls using the Leave-One-Out and N-fold cross-validation techniques, with a detection accuracy of over 98%. The high classification accuracy validated from a large data set suggests that these new features from DTW can be effectively used to help clinicians diagnose the disease at the earliest. Even though PD is not completely curable, early diagnosis would help clinicians to start the treatment from the beginning thereby reducing the intensity of symptoms at later stages.

Neurodegenerative diseases detection and grading using gait dynamics

Detection of neurodegenerative diseases such as Parkinson's disease, Huntington's disease, Amyotrophic Lateral Sclerosis, and grading of these diseases' severity have high clinical significance. These tasks based on walking analysis stand out compared to other methods due to their simplicity and non-invasiveness. This study has emerged to realize an artificial intelligence-based disease detection and severity prediction system for neurodegenerative diseases using gait features obtained from gait signals. For the detection of the disease, the problem is divided into parts which are subgroups of 4 classes consisting of Parkinson's, Huntington's, Amyotrophic Lateral Sclerosis diseases, and the control group. In addition, the disease vs. control subgroup where all diseases are collected under a single label, the subgroups where each disease is separately against the control group. For disease severity grading, each disease was divided into subgroups and a solution was sought for the prediction problem mentioned by various machine and deep learning methods separately for each group. In this context, the resulting detection performance was measured by the metrics of Accuracy, F1 Score, Precision, and Recall while the resulting prediction performance was measured by the metrics such as R, R2, MAE, MedAE, MSE, and RMSE.

Techniques for the detection and management of freezing of gait in Parkinson's disease - A systematic review and future perspectives

Freezing of Gait (FoG) is one of the most critical debilitating motor symptoms of advanced Parkinson's disease (PD) with a higher rate of occurrence in aged people. PD affects the cardinal motor functioning and leads to non-motor symptoms, including cognitive and neurobehavioral abnormalities, autonomic dysfunctions and sleep disorders. Since its pathogenesis is complex and unclear yet, this paper targets the studies done on the pathophysiology and epidemiology of FoG in PD. Gait disorder and cardinal features vary from festination (involuntary hurrying in walking) to freezing of gait (breakdown of repetitive movement of steps despite the intention to walk) in patients. Hence, it is difficult to assess the FoG in clinical trials. Therefore, the current research emphasizes wearable sensor-based systems over pharmacology and surgical methods.•This paper presents a technological review of various techniques used for the assessment of FoG with a comprehensive comparison.•Researchers are aiming at the development of wireless sensor-based assistive devices to (a) predict the FoG episode in a different environment, (b) acquire the long-term data for real-time analysis, and (c) cue the FoG patients.•We summarize the work done till now and future research directions needed for a suitable cueing mechanism to overcome FoG.

Use of common spatial patterns for early detection of Parkinson's disease

One of the most common diseases that affects human brain is Parkinson's disease. Detection of Parkinson's disease (PD) poses a serious challenge. Robust methods for feature extraction allowing separation between the electroencephalograms (EEG) of healthy subjects and PD patients are required. We used the EEG records of healthy subjects and PD patients which were subject to auditory tasks. We used the common spatial patterns (CSP) and Laplacian mask as methods to allow robust selection and extraction of features. We used the derived CSP whitening matrix to determine those channels that are the most promising in the terms of differentiating between EEGs of healthy controls and of PD patients. Using the selection of features calculated using the CSP we managed to obtain the classification accuracy of 85% when classifying EEG records belonging to groups of controls or PD patients. Using the features calculated using the Laplacian operator we obtained the classification accuracy of 90%. Diagnosing the PD in early stages using EEG is possible. The CSP proved to be a promising technique to detect informative channels and to separate between the groups. Use of the combination of features calculated using the Laplacian offers good separability between the two groups.

A deep learning approach for parkinson’s disease severity assessment

Purpose

Parkinson’s Disease comes on top among neurodegenerative diseases affecting 10 million worldwide. To detect Parkinson’s Disease in a prior state, gait analysis is an effective choice. However, monitoring of Parkinson’s Disease using gait analysis is time consuming and exhaustive for patients and physicians. To assess severity of symptoms, a rating scale called Unified Parkinson's Disease Rating Scale is used. It determines mild and severe cases. Today, Parkinson’s Disease severity assessment is made in gait laboratories and by manual examination. These are time consuming and it is costly for health institutions to build and maintain laboratories. By using low-cost wearables and an effective model, aforementioned problems can be solved.

Methods

We provide a computerized solution for quantifiable assessment of Parkinson’s Disease symptoms severity. By using wearable sensors, our framework can predict exact symptom values to assess Parkinson’s Disease severity. We propose a deep learning approach that utilizes Ground Reaction Force sensors. From sensor signals, features are extracted and fed to a hybrid deep learning model. This model is the combination of Convolutional Neural Networks and Locally Weighted Random Forest.

Results

Proposed framework achieved 0.897, 3.009, 4.556 in terms of Correlation Coefficient, Mean Absolute Error and Root Mean Square Error, respectively. Proposed framework outperformed other machine and deep learning models. We also evaluated classification performance for disease detection. We outperformed most of the previous studies, achieving 99.5% accuracy, 98.7% sensitivity and 99.1% specificity.

Conclusion

This is the first study to use a deep learning regression approach to predict exact symptom value of Parkinson’s Disease patients. Results show that this approach can be effectively employed as a disease severity assessment tool using wearable sensors.

Superior gait performance and balance ability in Latin dancers

Background

Latin dance consists of various fast and stability-challenging movements that require constant body adjustments to maintain proper posture and balance. Although human gaits are assumed to be symmetrical, several factors can contribute to asymmetrical behavior of the lower extremities in healthy adults. These include lower limb dominance, ground reaction forces, lower limb muscle power, foot placement angle, and range of joint motion. Gait impairment can lead to a high risk of falling, diminished mobility, and even cognition impairment. We hypothesized that Latin dancers might have a more symmetric gait pattern and better balance ability than healthy non-dancer controls.

Methods

We investigated the impact of Latin dance training on gait behaviors and body balance. We recruited twenty Latin dancers and 22 normal healthy subjects to conduct walking experiments and one-leg stance tests, and we measured their kinematic data by inertial measurement units. We then defined four performance indexes to assess gait performance and body stability to quantify the potential advantages of dance training.

Results

We found that the two gait asymmetric indexes during the walking test and the two performance indexes during the one-leg stance tests were better in Latin dancers compared with the healthy control group. The results confirmed the superiority of Latin dancers over the healthy control group in gait symmetry and balance stability. Our results suggest that Latin dancing training could effectively strengthen lower limb muscles and core muscle groups, thereby improving coordination and enhancing gait performance and balance.

Conclusion

Latin dance training can benefit gait performance and body balance. Further studies are needed to investigate the effect of Latin dance training on gait and balance outcomes in healthy subjects and patients with gait disorders.

A Systematic Survey of Research Trends in Technology Usage for Parkinson's Disease

Parkinson's disease (PD) is a neurological disorder with complicated and disabling motor and non-motor symptoms. The complexity of PD pathology is amplified due to its dependency on patient diaries and the neurologist's subjective assessment of clinical scales. A significant amount of recent research has explored new cost-effective and subjective assessment methods pertaining to PD symptoms to address this challenge. This article analyzes the application areas and use of mobile and wearable technology in PD research using the PRISMA methodology. Based on the published papers, we identify four significant fields of research: diagnosis, prognosis and monitoring, predicting response to treatment, and rehabilitation. Between January 2008 and December 2021, 31,718 articles were published in four databases: PubMed Central, Science Direct, IEEE Xplore, and MDPI. After removing unrelated articles, duplicate entries, non-English publications, and other articles that did not fulfill the selection criteria, we manually investigated 1559 articles in this review. Most of the articles (45%) were published during a recent four-year stretch (2018-2021), and 19% of the articles were published in 2021 alone. This trend reflects the research community's growing interest in assessing PD with wearable devices, particularly in the last four years of the period under study. We conclude that there is a substantial and steady growth in the use of mobile technology in the PD contexts. We share our automated script and the detailed results with the public, making the review reproducible for future publications.

Analysis of the stance phase of the gait cycle in Parkinson's disease and its potency for Parkinson's disease discrimination

In this study, using vertical ground reaction force (VGRF) data and focusing on the stance phase of the gait cycle, the effect of Parkinson's disease (PD) on gait was investigated. The used dataset consisted of 93 PD and 72 healthy individuals. Multiple comparisons correction ANOVA test and student t-test were used for statistical analyses. Results showed that a longer stance duration with a larger VGRF peak value (p < 0.05) was observed for PD patients during the stance phase. In addition, the VGRF peak value was delayed and blunted in PD cases compared with healthy individuals. These results indicated more time and effort for PD patients for posture stabilization during the stance phase. The time delay for different locations of the foot sole to contact the ground during the stance phase indicated that PD patients might use a different strategy for maintaining their body stability compared with healthy individuals. Although the VGRF time-domain pattern during the stance phase in PD was similar to healthy conditions, its local characteristics like duration and peak value differed significantly. The classification analysis based on the VGRF time-domain extracted features during the stance phase obtained PD recognition with accuracy, sensitivity and specificity of 90.82%, 88.63% and 82.56%, respectively.

Local Pattern Transformation Based Feature Extraction for Recognition of Parkinson's Disease Based on Gait Signals

Parkinson’s disease (PD) is a neuro-degenerative disorder primarily triggered due to the deterioration of dopamine-producing neurons in the substantia nigra of the human brain. The early detection of Parkinson’s disease can assist in preventing deteriorating health. This paper analyzes human gait signals using Local Binary Pattern (LBP) techniques during feature extraction before classification. Supplementary to the LBP techniques, Local Gradient Pattern (LGP), Local Neighbour Descriptive Pattern (LNDP), and Local Neighbour Gradient Pattern (LNGP) were utilized to extract features from gait signals. The statistical features were derived and analyzed, and the statistical Kruskal–Wallis test was carried out for the selection of an optimal feature set. The classification was then carried out by an Artificial Neural Network (ANN) for the identified feature set. The proposed Symmetrically Weighted Local Neighbour Gradient Pattern (SWLNGP) method achieves a better performance, with 96.28% accuracy, 96.57% sensitivity, and 95.94% specificity. This study suggests that SWLNGP could be an effective feature extraction technique for the recognition of Parkinsonian gait.

Implementation of a Deep Learning Algorithm Based on Vertical Ground Reaction Force Time-Frequency Features for the Detection and Severity Classification of Parkinson's Disease

Conventional approaches to diagnosing Parkinson’s disease (PD) and rating its severity level are based on medical specialists’ clinical assessment of symptoms, which are subjective and can be inaccurate. These techniques are not very reliable, particularly in the early stages of the disease. A novel detection and severity classification algorithm using deep learning approaches was developed in this research to classify the PD severity level based on vertical ground reaction force (vGRF) signals. Different variations in force patterns generated by the irregularity in vGRF signals due to the gait abnormalities of PD patients can indicate their severity. The main purpose of this research is to aid physicians in detecting early stages of PD, planning efficient treatment, and monitoring disease progression. The detection algorithm comprises preprocessing, feature transformation, and classification processes. In preprocessing, the vGRF signal is divided into 10, 15, and 30 s successive time windows. In the feature transformation process, the time domain vGRF signal in windows with varying time lengths is modified into a time–frequency spectrogram using a continuous wavelet transform (CWT). Then, principal component analysis (PCA) is used for feature enhancement. Finally, different types of convolutional neural networks (CNNs) are employed as deep learning classifiers for classification. The algorithm performance was evaluated using k-fold cross-validation (kfoldCV). The best average accuracy of the proposed detection algorithm in classifying the PD severity stage classification was 96.52% using ResNet-50 with vGRF data from the PhysioNet database. The proposed detection algorithm can effectively differentiate gait patterns based on time–frequency spectrograms of vGRF signals associated with different PD severity levels.

Electromyography-informed modeling for estimating muscle activation and force alterations in Parkinson's disease

Electromyography (EMG)-driven neuromusculoskeletal modeling (NMSM) enables simulating the mechanical function of multiple muscle-tendon units as controlled by nervous system in the generation of complex movements. In the context of clinical assessment this may enable understanding biomechanical factor contributing to gait disorders such as one induced by Parkinson's disease (PD). In spite of the challenges in the development of patient-specific models, this preliminary study aimed at establishing a feasible and noninvasive experimental and modeling pipeline to be adopted in clinics to detect PD-induced gait alterations. Four different NMSM have been implemented for three healthy controls using CEINMS, an OpenSim-compatible toolbox. Models differed in the EMG-normalization methods used for calibration purposes (i.e. walking trial normalization and maximum voluntary contraction normalization) and in the set of experimental EMGs used for the musculotendon-unit mapping (i.e. 4 channels vs. 15 channels). Model accuracy assessment showed no statistically significant differences between the more complete model (non-clinically viable) and the proposed reduced one (clinically viable). The clinically viable reduced model was systematically applied on a dataset including ten PD's and thirteen healthy controls. Results showed significant differences in the neuromuscular control strategy of the PD group in term of muscle forces and joint torques. Indeed, PD patients displayed a significantly lower magnitude on force production and revealed a higher amount of force variability with the respect of the healthy controls. The estimated variables could become a measurable biomechanical outcome to assess and track both disease progression and its impact on gait in PD subjects.

Time-frequency time-space LSTM for robust classification of physiological signals

Automated analysis of physiological time series is utilized for many clinical applications in medicine and life sciences. Long short-term memory (LSTM) is a deep recurrent neural network architecture used for classification of time-series data. Here time-frequency and time-space properties of time series are introduced as a robust tool for LSTM processing of long sequential data in physiology. Based on classification results obtained from two databases of sensor-induced physiological signals, the proposed approach has the potential for (1) achieving very high classification accuracy, (2) saving tremendous time for data learning, and (3) being cost-effective and user-comfortable for clinical trials by reducing multiple wearable sensors for data recording.

Detection and Classification of Stroke Gaits by Deep Neural Networks Employing Inertial Measurement Units

This paper develops Deep Neural Network (DNN) models that can recognize stroke gaits. Stroke patients usually suffer from partial disability and develop abnormal gaits that can vary widely and need targeted treatments. Evaluation of gait patterns is crucial for clinical experts to make decisions about the medication and rehabilitation strategies for the stroke patients. However, the evaluation is often subjective, and different clinicians might have different diagnoses of stroke gait patterns. In addition, some patients may present with mixed neurological gaits. Therefore, we apply artificial intelligence techniques to detect stroke gaits and to classify abnormal gait patterns. First, we collect clinical gait data from eight stroke patients and seven healthy subjects. We then apply these data to develop DNN models that can detect stroke gaits. Finally, we classify four common gait abnormalities seen in stroke patients. The developed models achieve an average accuracy of 99.35% in detecting the stroke gaits and an average accuracy of 97.31% in classifying the gait abnormality. Based on the results, the developed DNN models could help therapists or physicians to diagnose different abnormal gaits and to apply suitable rehabilitation strategies for stroke patients.

Insight into an unsupervised two-step sparse transfer learning algorithm for speech diagnosis of Parkinson's disease

Speech diagnosis of Parkinson’s disease (PD) as a non-invasive and simple diagnosis method is particularly worth exploring. However, the number of samples of speech-based PD is relatively small, and there exist discrepancies in the distribution between subjects. In order to solve the two problems, a novel unsupervised two-step sparse transfer learning is proposed in this paper to tackle with PD speech diagnosis. In the first step, convolution sparse coding with the coordinate selection of samples and features is designed to learn speech structure from the source domain to replenish sample information of the target domain. In the second step, joint local structure distribution alignment is designed to maintain the neighbor relationship between the respective samples of the training set and test set, and reduce the distribution difference between the two domains at the same time. Two representative public PD speech datasets and one real-world PD speech dataset were exploited to verify the proposed method on PD speech diagnosis. Experimental results demonstrate that each step of the proposed method has a positive effect on the PD speech classification results, and it also delivers superior performance over the existing relative methods.

Gait characteristics in individuals with Parkinson's disease during 1-minute treadmill walking

Background

No previous study has examined the age-dependent characteristics of gait in individuals between 50 and 79 years simultaneously in healthy individuals and individuals with Parkinson’s disease (PD) over continuous gait cycles. This study aimed to investigate age-related differences in gait characteristics on individuals age ranged 50–79 years, including individuals with PD, during a 1-minute treadmill walking session. Additionally, we aimed to investigate the differences associated with spatiotemporal gait parameters and PD compared in age-matched individuals.

Methods

This study included 26 individuals with PD and 90 participants age ranged 50–79 years. The treadmill walking test at a self-preferred speed was performed for 1 min. The embedded inertial measurement unit sensor in the left and right outsoles-based system was used to collect gait characteristics based on tri-axial acceleration and tri-axial angular velocities.

Results

Participants aged >60 years had a decreased gait speed and shortened stride and step, which may demonstrate a distinct shift in aging (all p < 0.005). Individuals with PD showed more of a decrease in variables with a loss of consistency, including gait asymmetry (GA), phase coordination index (PCI) and coefficient of variation (CV) of all variables, than age-matched individuals (all p < 0.001). Gait speed, stride and step length, stance phase, variability, GA and PCI were the variables that highly depended on age and PD.

Discussion

Older adults could be considered those older than 60 years of age when gait alterations begin, such as a decreased gait speed as well as shortened stride and step length. On the other hand, a loss of consistency in spatiotemporal parameters and a higher GA and PCI could be used to identify individuals with PD. Thus, the CV of all spatiotemporal parameters, GA and PCI during walking could play an important role and be useful in identifying individuals with PD.

Conclusion

This study provided the notable aging pattern characteristics of gait in individuals >50 years, including individuals with PD. Increasing age after 60 years is associated with deterioration in spatiotemporal parameters of gait during continuous 1-minute treadmill walking. Additionally, GA, PCI and the CV of all variables could be used to identify PD which would be placed after 70 years of age. It may be useful to determine the decline of gait performance in general and among individuals with PD.

Lesser magnitudes of lower extremity variability during terminal swing characterizes walking patterns in children with autism

BACKGROUND

Anecdotally, children with Autism Spectrum Disorder have highly variable lower extremity walking patterns, yet, this has not been sufficiently quantified. As such, the purpose of this study was to examine walking pattern variability by way of lower extremity coordination and spatio-temporal characteristics in children with autism compared with individuals with typical development during over-ground walking.

METHODS

Bilateral continuous relative phase variability was computed for the thigh-leg, leg-foot, and thigh-foot segment couples for 11 children with autism and 9 children with typical development at each gait sub-phase. Furthermore, left and right stride lengths and stride width were computed and compared. The Model Statistic was utilized to test for statistical differences in variability between each child with autism to an aggregate group with typical development. Effect sizes were computed to determine the meaningfulness between responses for children with autism and typical development. Coefficient of variation and effect sizes were computed for stride lengths and stride width.

FINDINGS

Analysis revealed that children with autism exhibited differences in variability in each gait sub-phase. Notably, all but two children with autism exhibited lesser variability in all segment couples during terminal swing. Differences in stride lengths were relatively minimal, however, greater coefficient of variation magnitudes in stride width were observed in children with autism.

INTERPRETATION

This finding reveals that children with autism may have limited or a preferred movement strategy when preparing the foot for ground contact. The findings from this study suggest variability may be an identifiable characteristic during movement in children with autism.

A Dual-Modal Attention-Enhanced Deep Learning Network for Quantification of Parkinson's Disease Characteristics

It is well known that most patients with Parkinson's disease (PD) have different degree of movement disorders, such as shuffling, festination and akinetic episodes, which could degenerate the life quality of PD patients. Therefore, it is very useful to develop a computerized tool to provide an objective evaluation of PD patients' gait. In this study, we implemented a novel gait evaluating approach to provide not only a binary classification of PD gaits and normal walking, but also a quantification of the PD gaits to relate them to the PD severity level. The proposed system is a dual-modal deep-learning-based model, where left and right gait is modeled separately by a convolutional neural network (CNN) followed by an attention-enhanced long short-term memory (LSTM) network. The left and right samples for model training and testing were segmented sequentially from multiple 1D vertical ground reaction force (VGRF) signals according to the detected gait cycle. Experimental results indicate that our model can provide state-of-the-art performance in terms of classification accuracy. It is expected that the proposed model can be a useful gait assistance to provide a quantitative evaluation of PD gaits with high confidence and accuracy if trained suitably.

Artificial intelligence for assisting diagnostics and assessment of Parkinson's disease-A review

Artificial intelligence, specifically machine learning, has found numerous applications in computer-aided diagnostics, monitoring and management of neurodegenerative movement disorders of parkinsonian type. These tasks are not trivial due to high inter-subject variability and similarity of clinical presentations of different neurodegenerative disorders in the early stages. This paper aims to give a comprehensive, high-level overview of applications of artificial intelligence through machine learning algorithms in kinematic analysis of movement disorders, specifically Parkinson's disease (PD). We surveyed papers published between January 2007 and January 2019, within online databases, including PubMed and Science Direct, with a focus on the most recently published studies. The search encompassed papers dealing with the implementation of machine learning algorithms for diagnosis and assessment of PD using data describing motion of upper and lower extremities. This systematic review presents an overview of 48 relevant studies published in the abovementioned period, which investigate the use of artificial intelligence for diagnostics, therapy assessment and progress prediction in PD based on body kinematics. Different machine learning algorithms showed promising results, particularly for early PD diagnostics. The investigated publications demonstrated the potentials of collecting data from affordable and globally available devices. However, to fully exploit artificial intelligence technologies in the future, more widespread collaboration is advised among medical institutions, clinicians and researchers, to facilitate aligning of data collection protocols, sharing and merging of data sets.

Gait Analysis for Post-Stroke Hemiparetic Patient by Multi-Features Fusion Method

Walking is a basic requirement for participating in daily activities. Neurological diseases such as stroke can significantly affect one's gait and thereby restrict one's activities that are a part of daily living. Previous studies have demonstrated that gait temporal parameters are useful for characterizing post-stroke hemiparetic gait. However, no previous studies have investigated the symmetry, regularity and stability of post-stroke hemiparetic gaits. In this study, the dynamic time warping (DTW) algorithm, sample entropy method and empirical mode decomposition-based stability index were utilized to obtain the three aforementioned types of gait features, respectively. Studies were conducted with 15 healthy control subjects and 15 post-stroke survivors. Experimental results revealed that the proposed features could significantly differentiate hemiparetic patients from healthy control subjects by a Mann-Whitney test (with a p-value of less than 0.05). Finally, four representative classifiers were utilized in order to evaluate the possible capabilities of these features to distinguish patients with hemiparetic gaits from the healthy control subjects. The maximum area under the curve values were shown to be 0.94 by the k-nearest-neighbor (kNN) classifier. These promising results have illustrated that the proposed features have considerable potential to promote the future design of automatic gait analysis systems for clinical practice.

A convolutional neural network Cascade for plantar pressure images registration

BACKGROUND

Plantar pressure image (PPI) recorded in high spatial and temporal resolution is very useful in clinical gait analysis. For functional analysis of PPI, image registration is often performed to maximally correlate source image with a template image. Previous methods estimate the registration parameters by iteratively optimizing different objective functions. These methods are often computational expensive to achieve satisfactory registration accuracy.

RESEARCH QUESTION

Can we develop a single PPI registration technique that performs more rapidly than previous methods, and that also maintains adequate PPI correspondence as defined by various (dis)similarity metrics?

METHODS

A cascaded convolutional neural network (CNN) was proposed for the registration of PPIs. Our model was trained to learn a regression from the difference between the template and misaligned images to the registration parameters. The registration performance was evaluated by three different metrics, i.e. the mean squared error (MSE), the exclusive or (XOR), and the mutual information (MI). For comparison, four previous methods were also implemented. These included the principal axes (PA) method, the center of pressure trajectory (COP) method, the MSE method, and the XOR method.

RESULTS

Experimental results on a dataset with 71 PPI template-source pairs showed that the proposed CNN-based method could obtain comparable registration accuracy to the MSE and XOR method. With regards to the registration speed, registration durations (mean ± sd in seconds) per image pair were: MSE (30.584 ± 2.171), XOR (24.245 ± 1.596), PA (0.016 ± 0.003), COP (25.614 ± 0.341), and the proposed model (0.054 ± 0.007).

SIGNIFICANCE

Our findings indicate that the proposed registration approach can achieve high accuracy but less computational time. Thus, it is more practical to utilize our pre-trained CNN-based model to develop near-real time applications for plantar pressure images registration.

Classification of gait patterns between patients with Parkinson's disease and healthy controls using phase space reconstruction (PSR), empirical mode decomposition (EMD) and neural networks

Parkinson's disease (PD) is a common neurodegenerative disorder that affects human's quality of life, especially leading to locomotor deficits such as postural instability and gait disturbances. Gait signal is one of the best features to characterize and detect movement disorders caused by a malfunction in parts of the brain and nervous system of the patients with PD. Various classification approaches using spatiotemporal gait variables have been presented earlier to classify Parkinson's gait. In this study we propose a novel method for gait pattern classification between patients with PD and healthy controls, based upon phase space reconstruction (PSR), empirical mode decomposition (EMD) and neural networks. First, vertical ground reaction forces (GRFs) at specific positions of human feet are captured and then phase space is reconstructed. The properties associated with the gait system dynamics are preserved in the reconstructed phase space. Three-dimensional (3D) PSR together with Euclidean distance (ED) has been used. These measured parameters demonstrate significant difference in gait dynamics between the two groups and have been utilized to form a reference variable set. Second, reference variables are decomposed into Intrinsic Mode Functions (IMFs) using EMD, and the third IMFs are extracted and served as gait features. Third, neural networks are then used as the classifier to distinguish between patients with PD and healthy controls based on the difference of gait dynamics preserved in the gait features between the two groups. Finally, experiments are carried out on 93 PD patients and 73 healthy subjects to assess the effectiveness of the proposed method. By using 2-fold, 10-fold and leave-one-out cross-validation styles, the correct classification rates are reported to be 91.46%, 96.99% and 98.80%, respectively. Compared with other state-of-the-art methods, the results demonstrate superior performance and the proposed method can serve as a potential candidate for the automatic and non-invasive classification between patients with PD and healthy subjects.

Data-Driven Based Approach to Aid Parkinson's Disease Diagnosis

This article presents a machine learning methodology for diagnosing Parkinson's disease (PD) based on the use of vertical Ground Reaction Forces (vGRFs) data collected from the gait cycle. A classification engine assigns subjects to healthy or Parkinsonian classes. The diagnosis process involves four steps: data pre-processing, feature extraction and selection, data classification and performance evaluation. The selected features are used as inputs of each classifier. Feature selection is achieved through a wrapper approach established using the random forest algorithm. The proposed methodology uses both supervised classification methods including K-nearest neighbour (K-NN), decision tree (DT), random forest (RF), Naïve Bayes (NB), support vector machine (SVM) and unsupervised classification methods such as K-means and the Gaussian mixture model (GMM). To evaluate the effectiveness of the proposed methodology, an online dataset collected within three different studies is used. This data set includes vGRF measurements collected from eight force sensors placed under each foot of the subjects. Ninety-three patients suffering from Parkinson's disease and 72 healthy subjects participated in the experiments. The obtained performances are compared with respect to various metrics including accuracy, precision, recall and F-measure. The classification performance evaluation is performed using the leave-one-out cross validation. The results demonstrate the ability of the proposed methodology to accurately differentiate between PD subjects and healthy subjects. For the purpose of validation, the proposed methodology is also evaluated with an additional dataset including subjects with neurodegenerative diseases (Amyotrophic Lateral Sclerosis (ALS) and Huntington's disease (HD)). The obtained results show the effectiveness of the proposed methodology to discriminate PD subjects from subjects with other neurodegenerative diseases with a relatively high accuracy.

String Grammar Unsupervised Possibilistic Fuzzy C-Medians for Gait Pattern Classification in Patients with Neurodegenerative Diseases

Neurodegenerative diseases that affect serious gait abnormalities include Parkinson's disease (PD), amyotrophic lateral sclerosis (ALS), and Huntington disease (HD). These diseases lead to gait rhythm distortion that can be determined by stride time interval of footfall contact times. In this paper, we present a new method for gait classification of neurodegenerative diseases. In particular, we utilize a symbolic aggregate approximation algorithm to convert left-foot stride-stride interval into a sequence of symbols using a symbolic aggregate approximation. We then find string prototypes of each class using the newly proposed string grammar unsupervised possibilistic fuzzy C-medians. Then in the testing process the fuzzy k-nearest neighbor is used. We implement the system on three 2-class problems, i.e., the classification of ALS against healthy patients, that of HD against healthy patients , and that of PD against healthy patients. The system is also implemented on one 4-class problem (the classification of ALS, HD, PD, and healthy patients altogether) called NDDs versus healthy. We found that our system yields a very good detection result. The average correct classification for ALS versus healthy is 96.88%, and that for HD versus healthy is 97.22%, whereas that for PD versus healthy is 96.43%. When the system is implemented on 4-class problem, the average accuracy is approximately 98.44%. It can provide prototypes of gait signals that are more understandable to human.

Can a Smartphone Diagnose Parkinson Disease? A Deep Neural Network Method and Telediagnosis System Implementation

Parkinson's disease (PD) is primarily diagnosed by clinical examinations, such as walking test, handwriting test, and MRI diagnostic. In this paper, we propose a machine learning based PD telediagnosis method for smartphone. Classification of PD using speech records is a challenging task owing to the fact that the classification accuracy is still lower than doctor-level. Here we demonstrate automatic classification of PD using time frequency features, stacked autoencoders (SAE), and K nearest neighbor (KNN) classifier. KNN classifier can produce promising classification results from useful representations which were learned by SAE. Empirical results show that the proposed method achieves better performance with all tested cases across classification tasks, demonstrating machine learning capable of classifying PD with a level of competence comparable to doctor. It concludes that a smartphone can therefore potentially provide low-cost PD diagnostic care. This paper also gives an implementation on browser/server system and reports the running time cost. Both advantages and disadvantages of the proposed telediagnosis system are discussed.

An automatic non-invasive method for Parkinson's disease classification

BACKGROUND AND OBJECTIVE

The automatic noninvasive identification of Parkinson's disease (PD) is attractive to clinicians and neuroscientist. Various analysis and classification approaches using spatiotemporal gait variables have been presented earlier in classifying Parkinson's gait. In this paper, we present a wavelet transform based representation of spatiotemporal gait variables to explore the potential of such representation in the identification of Parkinson's gait.

METHODS

Here, we present wavelet analysis as an alternate method and show that wavelet analysis combined with support vector machine (SVM) can produce efficient classification accuracy. Computationally simplified features are extracted from the wavelet transformation and are fed to support vector machine for Parkinson's gait identification. We have assessed various gait parameters namely stride interval, swing interval, and stance interval (from both legs) to observe the best single parameter for such classification.

RESULTS

By employing wavelet decomposition of the gait variables as an alternate method for the identification of Parkinson's subjects, the classification accuracy of 90.32% (Confidence Interval; 74.2%-97.9%) has been achieved, at par to recently reported accuracy, using only one gait parameter. Left stance interval performed equally good to Right swing interval showing classification accuracy of 90.32%. The classification accuracy improved to 100% when all the gait parameters from left leg were put together to form a larger feature vector. We have shown that Haar wavelet performed significantly better than db2 wavelet (p = 0.05) for certain gait variables e.g., right stride time series. The results show that wavelet analysis is a promising approach in reducing down the required number of gait variables, however at the cost of increased computations in wavelet analysis.

CONCLUSIONS

In this work a wavelet transform approach is explored to classify Parkinson's subjects and healthy subjects using their gait cycle variables. The results show that the proposed method can efficiently extract relevant features from the different levels of the wavelet towards the classification of Parkinson's and healthy subjects and thus, the present work is a potential candidate for the automatic noninvasive neurodegenerative disease classification.

Metric learning for Parkinsonian identification from IMU gait measurements

Diagnosis of people with mild Parkinson's symptoms is difficult. Nevertheless, variations in gait pattern can be utilised to this purpose, when measured via Inertial Measurement Units (IMUs). Human gait, however, possesses a high degree of variability across individuals, and is subject to numerous nuisance factors. Therefore, off-the-shelf Machine Learning techniques may fail to classify it with the accuracy required in clinical trials. In this paper we propose a novel framework in which IMU gait measurement sequences sampled during a 10m walk are first encoded as hidden Markov models (HMMs) to extract their dynamics and provide a fixed-length representation. Given sufficient training samples, the distance between HMMs which optimises classification performance is learned and employed in a classical Nearest Neighbour classifier. Our tests demonstrate how this technique achieves accuracy of 85.51% over a 156 people with Parkinson's with a representative range of severity and 424 typically developed adults, which is the top performance achieved so far over a cohort of such size, based on single measurement outcomes. The method displays the potential for further improvement and a wider application to distinguish other conditions.

Most suitable mother wavelet for the analysis of fractal properties of stride interval time series via the average wavelet coefficient method

Human gait is a complex interaction of many nonlinear systems and stride intervals exhibiting self-similarity over long time scales that can be modeled as a fractal process. The scaling exponent represents the fractal degree and can be interpreted as a "biomarker" of relative diseases. The previous study showed that the average wavelet method provides the most accurate results to estimate this scaling exponent when applied to stride interval time series. The purpose of this paper is to determine the most suitable mother wavelet for the average wavelet method. This paper presents a comparative numerical analysis of 16 mother wavelets using simulated and real fractal signals. Simulated fractal signals were generated under varying signal lengths and scaling exponents that indicate a range of physiologically conceivable fractal signals. The five candidates were chosen due to their good performance on the mean square error test for both short and long signals. Next, we comparatively analyzed these five mother wavelets for physiologically relevant stride time series lengths. Our analysis showed that the symlet 2 mother wavelet provides a low mean square error and low variance for long time intervals and relatively low errors for short signal lengths. It can be considered as the most suitable mother function without the burden of considering the signal length.

Classification of Parkinson's Disease Gait Using Spatial-Temporal Gait Features

Quantitative gait assessment is important in diagnosis and management of Parkinson's disease (PD); however, gait characteristics of a cohort are dispersed by patient physical properties including age, height, body mass, and gender, as well as walking speed, which may limit capacity to discern some pathological features. The aim of this study was twofold. First, to use a multiple regression normalization strategy that accounts for subject age, height, body mass, gender, and self-selected walking speed to identify differences in spatial-temporal gait features between PD patients and controls; and second, to evaluate the effectiveness of machine learning strategies in classifying PD gait after gait normalization. Spatial-temporal gait data during self-selected walking were obtained from 23 PD patients and 26 aged-matched controls. Data were normalized using standard dimensionless equations and multiple regression normalization. Machine learning strategies were then employed to classify PD gait using the raw gait data, data normalized using dimensionless equations, and data normalized using the multiple regression approach. After normalizing data using the dimensionless equations, only stride length, step length, and double support time were significantly different between PD patients and controls (p < 0.05); however, normalizing data using the multiple regression method revealed significant differences in stride length, cadence, stance time, and double support time. Random Forest resulted in a PD classification accuracy of 92.6% after normalizing gait data using the multiple regression approach, compared to 80.4% (support vector machine) and 86.2% (kernel Fisher discriminant) using raw data and data normalized using dimensionless equations, respectively. Our multiple regression normalization approach will assist in diagnosis and treatment of PD using spatial-temporal gait data.

Parkinson's disease classification using gait analysis via deterministic learning

Gait analysis plays an important role in maintaining the well-being of human mobility and health care, and is a valuable tool for obtaining quantitative information on motor deficits in Parkinson's disease (PD). In this paper, we propose a method to classify (diagnose) patients with PD and healthy control subjects using gait analysis via deterministic learning theory. The classification approach consists of two phases: a training phase and a classification phase. In the training phase, gait characteristics represented by the gait dynamics are derived from the vertical ground reaction forces under the usual and self-selected paces of the subjects. The gait dynamics underlying gait patterns of healthy controls and PD patients are locally accurately approximated by radial basis function (RBF) neural networks. The obtained knowledge of approximated gait dynamics is stored in constant RBF networks. The gait patterns of healthy controls and PD patients constitute a training set. In the classification phase, a bank of dynamical estimators is constructed for all the training gait patterns. Prior knowledge of gait dynamics represented by the constant RBF networks is embedded in the estimators. By comparing the set of estimators with a test gait pattern of a certain PD patient to be classified (diagnosed), a set of classification errors are generated. The average L₁ norms of the errors are taken as the classification measure between the dynamics of the training gait patterns and the dynamics of the test PD gait pattern according to the smallest error principle. When the gait patterns of 93 PD patients and 73 healthy controls are classified with five-fold cross-validation method, the accuracy, sensitivity and specificity of the results are 96.39%, 96.77% and 95.89%, respectively. Based on the results, it may be claimed that the features and the classifiers used in the present study could effectively separate the gait patterns between the groups of PD patients and healthy controls.

Intra- and Inter-Fractional Variation Prediction of Lung Tumors Using Fuzzy Deep Learning

Tumor movements should be accurately predicted to improve delivery accuracy and reduce unnecessary radiation exposure to healthy tissue during radiotherapy. The tumor movements pertaining to respiration are divided into intra-fractional variation occurring in a single treatment session and inter-fractional variation arising between different sessions. Most studies of patients’ respiration movements deal with intra-fractional variation. Previous studies on inter-fractional variation are hardly mathematized and cannot predict movements well due to inconstant variation. Moreover, the computation time of the prediction should be reduced. To overcome these limitations, we propose a new predictor for intra- and inter-fractional data variation, called intra- and inter-fraction fuzzy deep learning (IIFDL), where FDL, equipped with breathing clustering, predicts the movement accurately and decreases the computation time. Through the experimental results, we validated that the IIFDL improved root-mean-square error (RMSE) by 29.98% and prediction overshoot by 70.93%, compared with existing methods. The results also showed that the IIFDL enhanced the average RMSE and overshoot by 59.73% and 83.27%, respectively. In addition, the average computation time of IIFDL was 1.54 ms for both intra- and inter-fractional variation, which was much smaller than the existing methods. Therefore, the proposed IIFDL might achieve real-time estimation as well as better tracking techniques in radiotherapy.

An efficient diagnosis system for Parkinson's disease using kernel-based extreme learning machine with subtractive clustering features weighting approach

A novel hybrid method named SCFW-KELM, which integrates effective subtractive clustering features weighting and a fast classifier kernel-based extreme learning machine (KELM), has been introduced for the diagnosis of PD. In the proposed method, SCFW is used as a data preprocessing tool, which aims at decreasing the variance in features of the PD dataset, in order to further improve the diagnostic accuracy of the KELM classifier. The impact of the type of kernel functions on the performance of KELM has been investigated in detail. The efficiency and effectiveness of the proposed method have been rigorously evaluated against the PD dataset in terms of classification accuracy, sensitivity, specificity, area under the receiver operating characteristic (ROC) curve (AUC), f-measure, and kappa statistics value. Experimental results have demonstrated that the proposed SCFW-KELM significantly outperforms SVM-based, KNN-based, and ELM-based approaches and other methods in the literature and achieved highest classification results reported so far via 10-fold cross validation scheme, with the classification accuracy of 99.49%, the sensitivity of 100%, the specificity of 99.39%, AUC of 99.69%, the f-measure value of 0.9964, and kappa value of 0.9867. Promisingly, the proposed method might serve as a new candidate of powerful methods for the diagnosis of PD with excellent performance.