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.

Predicting Early Stage Drug Induced Parkinsonism using Unsupervised and Supervised Machine Learning.

Drug Induced Parkinsonism (DIP) is the most common, debilitating movement disorder induced by antipsychotics. There is no tool available in clinical practice to effectively diagnose the symptoms at the onset of the disease. In this study, the variations in gait accelerometer data due to the intermittency of tremor at the initial stages is examined. These variations are used to train a logistic regression model to predict subjects with early-stage DIP. The logistic classifier predicts if a subject is a DIP or control with approximately 89% sensitivity and 96% specificity. This paper discusses the algorithm used to extract the features in gait data for training the classifier to predict DIP at the earliest.Clinical Relevance- Diagnosing the disease and the causative drug is vital as the physical health of a patient who is mentally unstable can deteriorate with prolonged usage of the drug. The proposed model helps clinicians to diagnose the disease at the onset of tremors with an accuracy of 93.58%.

Using accelerometer as a diagnostic tool to detect drug-induced parkinsonism (DIP) secondary to first-generation anti-psychotic medications.

OBJECTIVE: The objective of this study is to examine the effectiveness of an accelerometer-based compact system in detecting and quantifying drug-induced parkinsonism (DIP) in patients with schizophrenia. METHOD: A pilot study controlled clinical trial comprising 6 people with schizophrenia and 11 control subjects was conducted at Alfred Health, Melbourne. Participants had their movements assessed using Barnes Akathisia Rating Scale (BARS), Simpson Angus Scale (SAS) and Movement Disorder Society Unified Parkinson's Disease Rating Scale Part III (MDS-UPDRS III) followed by an assessment of gait using three triaxial accelerometers. RESULTS: Median BARS, SAS, MDS-UPDRS III and accelerometer scores were significantly higher for patients with schizophrenia than controls. Accelerometers detected three times more rest tremor than clinical rating scales. Patients with schizophrenia had 70% of their dynamic acceleration at frequencies between 4 and 10 Hz, which is almost twice that observed in the control population (38%). Accelerometer scores were significantly correlated with BARS scores. CONCLUSION: Accelerometers were able to accurately detect patients with DIP better than some clinical rating scale including the SAS. Further larger-scale studies must be conducted to further demonstrate the accuracy of accelerometers in detecting DIP.

Wavelet PCA for automatic identification of walking with and without an exoskeleton on a treadmill using pressure and accelerometer sensors.

Nowadays portable devices with more number of sensors are used for gait assessment and monitoring for elderly and disabled. However, the problem with using multiple sensors is that if they are placed on the same platform or base, there could be cross talk between them, which could change the signal amplitude or add noise to the signal. Hence, this study uses wavelet PCA as a signal processing technique to separate the original sensor signal from the signal obtained from the sensors through the integrated unit to compare the two types of walking (with and without an exoskeleton). This comparison using wavelet PCA will enable the researchers to obtain accurate sensor data and compare and analyze the data in order to further improve the design of compact portable devices used to monitor and assess the gait in stroke or paralyzed subjects. The advantage of designing such systems is that they can also be used to assess and monitor the gait of the stroke subjects at home, which will save them time and efforts to visit the laboratory or clinic.

Multiscale PCA to distinguish regular and irregular surfaces using tri axial head and trunk acceleration signals.

This study uses multiscale principal component analysis (MSPCA) signal processing technique in order to distinguish the two different surfaces, tiled (regular) and cobbled (irregular) using accelerometry data (recorded from MTx sensors). Two MTx sensors were placed on the head and trunk of the subject while the subject walked freely over the regular and irregular surfaces during a free walk. 3D acceleration signals, vertical, medio lateral (ML) and anterior-posterior (AP) were recorded for the head and trunk segments and compared for the free walk on a defined route. The magnitude of the ML and AP acceleration obtained from the MTx sensors (for both head & trunk) was higher when walking over the irregular (cobbled) surface as compared to the regular (tiled) surface. The accelerometry data was initially analysed using MSPCA and was later classified using naïve Bayesian classifier with >86% accuracy. This research study demonstrates that MSPCA can be used to distinguish the regular and irregular surfaces. The proposed method could be very useful as an automated method for classification of the two surfaces.

Automated method to distinguish toe walking strides from normal strides in the gait of idiopathic toe walking children from heel accelerometry data.

Toe walking mainly occurs in children due to medical condition or physical injury. When there are no obvious signs of any medical condition or physical injury, a diagnosis of Idiopathic Toe Walking (ITW) is made. ITW children habitually walk on their toes, however can modify their gait and walk with a heel-toe gait if they want to. Correct gait assessment in ITW children therefore becomes difficult. To solve this problem, we have developed an automated way to assess the gait in ITW children using a dual axis accelerometer. Heel acceleration data was recorded from the gait of ITW children using boots embedded with the sensor in the heel and interfaced to a handheld oscilloscope. An innovative signal processing algorithm was developed in IgorPro to distinguish toe walking stride from normal stride using the acceleration data. The algorithm had an accuracy of 98.5%. Based on the statistical analysis of the heel accelerometer data, it can be concluded that the foot angle during mid stance in ITW children tested, varied from 36° to 11.5° while as in normal children the foot stance angle is approximately zero. This algorithm was later implemented in a system (embedded in the heel) which was used remotely to differentiate toe walking stride from normal stride. Although the algorithm classifies toe walking stride from normal stride in ITW children, it can be generalized for other applications such as toe walking in Cerebral Palsy or Acquired Brain Injury subjects. The system can also be used to assess the gait for other applications such as Parkinson's disease by modifying the algorithm.

Detecting idiopathic toe-walking gait pattern from normal gait pattern using heel accelerometry data and Support Vector Machines.

Toe walking is commonly seen in children with neurological symptoms such as cerebral palsy. However idiopathic toe walking (ITW) in children is considered to be habitual. ITW children are categorized as toe walkers without any neurological problems, however they walk with their foot plantar-flexed. These children often suffer poor sport performance leading to low exercise levels and the associated consequences. If the condition is not treated, the ITW children eventually develop abnormal gait pattern as adults and could suffer from postural problems. However, ITW gait is difficult to observe since children can modify their gait when made aware of it. Gait analysis using heel accelerometry data in ITW children could provide an objective and quantitative description of their toe walking and may thus be beneficial for observing ITW. In this paper, we propose a technique based on Support Vector Machines (SVM) to recognize ITW gait patterns using heel accelerometry data. Test results indicated that the SVM is able to identify ITW gait patterns with a maximum accuracy of 87.5% when a feature selection algorithm was applied.