NeuroDiag: Software for Automated Diagnosis of Parkinson's Disease Using Handwriting.

OBJECTIVE: A change in handwriting is an early sign of Parkinson's disease (PD). However, significant inter-person differences in handwriting make it difficult to identify pathological handwriting, especially in the early stages. This paper reports the testing of NeuroDiag, a software-based medical device, for the automated detection of PD using handwriting patterns. NeuroDiag is designed to direct the user to perform six drawing and writing tasks, and the recordings are then uploaded onto a server for analysis. Kinematic information and pen pressure of handwriting are extracted and used as baseline parameters. NeuroDiag was trained based on 26 PD patients in the early stage of the disease and 26 matching controls. METHODS: Twenty-three people with PD (PPD) in their early stage of the disease, 25 age-matched healthy controls (AMC), and 7 young healthy controls were recruited for this study. Under the supervision of a consultant neurologist or their nurse, the participants used NeuroDiag. The reports were generated in real-time and tabulated by an independent observer. RESULTS: The participants were able to use NeuroDiag without assistance. The handwriting data was successfully uploaded to the server where the report was automatically generated in real-time. There were significant differences in the writing speed between PPD and AMC (P<0.001). NeuroDiag showed 86.96% sensitivity and 76.92% specificity in differentiating PPD from those without PD. CONCLUSION: In this work, we tested the reliability of NeuroDiag in differentiating between PPD and AMC for real-time applications. The results show that NeuroDiag has the potential to be used to assist neurologists and for telehealth applications. Clinical and Translational Impact Statement - This pre-clinical study shows the feasibility of developing a community-wide screening program for Parkinson's disease using automated handwriting analysis software, NeuroDiag.

Wearable Accelerometer and Gyroscope Sensors for Estimating the Severity of Essential Tremor.

BACKGROUND: Several validated clinical scales measure the severity of essential tremor (ET). Their assessments are subjective and can depend on familiarity and training with scoring systems. METHOD: We propose a multi-modal sensing using a wearable inertial measurement unit for estimating scores on the Fahn-Tolosa-Marin tremor rating scale (FTM) and determine the classification accuracy within the tremor type. 17 ET participants and 18 healthy controls were recruited for the study. Two movement disorder neurologists who were blinded to prior clinical information viewed video recordings and scored the FTM. Participants drew a guided Archimedes spiral while wearing an inertial measurement unit placed at the mid-point between the lateral epicondyle of the humerus and the anatomical snuff box. Acceleration and gyroscope recordings were analyzed. The ratio of the power spectral density between frequency bands 0.5-4 Hz and 4-12 Hz, and the sum of power spectrum density over the entire spectrum of 2-74 Hz, for both accelerometer and gyroscope data, were computed. FTM was estimated using regression model and classification using SVM was validated using the leave-one-out method. RESULTS: Regression analysis showed a moderate to good correlation when individual features were used, while correlation was high ([Formula: see text] = 0.818) when suitable features of the gyro and accelerometer were combined. The accuracy for two-class classification of the combined features using SVM was 91.42% while for four-class it was 68.57%. CONCLUSION: Potential applications of this novel wearable sensing method using a wearable Inertial Measurement Unit (IMU) include monitoring of ET and clinical trials of new treatments for the disorder.

On the inter-dataset generalization of machine learning approaches to Parkinson's disease detection from voice.

BACKGROUND AND OBJECTIVE: Parkinson's disease is the second-most-common neurodegenerative disorder that affects motor skills, cognitive processes, mood, and everyday tasks such as speaking and walking. The voices of people with Parkinson's disease may become weak, breathy, or hoarse and may sound emotionless, with slurred words and mumbling. Algorithms for computerized voice analysis have been proposed and have shown highly accurate results. However, these algorithms were developed on single, limited datasets, with participants possessing similar demographics. Such models are prone to overfitting and are unsuitable for generalization, which is essential in real-world applications. METHODS: We evaluated the computerized Parkinson's disease diagnosis performance of various machine learning models and showed that these models degraded rapidly when used on different datasets. We evaluated two mainstream state-of-the-art approaches, one based on deep convolutional neural networks and another based on voice feature extraction followed by a shallow classifier (i.e., extreme gradient boosting (XGBoost)). RESULTS: An investigation with four datasets (CzechPD, PC-GITA, ITA, and RMIT-PD) proved that even if the algorithms yielded excellent performance on a single dataset, the results obtained on new data or even a mix of datasets were very unsatisfactory. CONCLUSIONS: More work needs to be done to make computerized voice analysis methods for Parkinson's disease diagnosis suitable for real-world applications.

Computerised prediction of healing for venous leg ulcers.

Early prediction of delayed healing for venous leg ulcers could improve management outcomes by enabling earlier initiation of adjuvant therapies. In this paper, we propose a framework for computerised prediction of healing for venous leg ulcers assessed in home settings using thermal images of the 0 week. Wound data of 56 older participants over 12 weeks were used for the study. Thermal images of the wounds were collected in their homes and labelled as healed or unhealed at the 12th week follow up. Textural information of the thermal images at week 0 was extracted. Thermal images of unhealed wounds had a higher variation of grey tones distribution. We demonstrated that the first three principal components of the textural features from one timepoint can be used as an input to a Bayesian neural network to discriminate between healed and unhealed wounds. Using the optimal Bayesian neural network, the classification results showed 78.57% sensitivity and 60.00% specificity. This non-contact method, incorporating machine learning, can provide a computerised prediction of this delay in the first assessment (week 0) in participants' homes compared to the current method that is able to do this in 3rd week and requires contact digital planimetry.

Wearable sensors during drawing tasks to measure the severity of essential tremor.

Commonly used methods to assess the severity of essential tremor (ET) are based on clinical observation and lack objectivity. This study proposes the use of wearable accelerometer sensors for the quantitative assessment of ET. Acceleration data was recorded by inertial measurement unit (IMU) sensors during sketching of Archimedes spirals in 17 ET participants and 18 healthy controls. IMUs were placed at three points (dorsum of hand, posterior forearm, posterior upper arm) of each participant's dominant arm. Movement disorder neurologists who were blinded to clinical information scored ET patients on the Fahn-Tolosa-Marin rating scale (FTM) and conducted phenotyping according to the recent Consensus Statement on the Classification of Tremors. The ratio of power spectral density of acceleration data in 4-12 Hz to 0.5-4 Hz bands and the total duration of the action were inputs to a support vector machine that was trained to classify the ET subtype. Regression analysis was performed to determine the relationship of acceleration and temporal data with the FTM scores. The results show that the sensor located on the forearm had the best classification and regression results, with accuracy of 85.71% for binary classification of ET versus control. There was a moderate to good correlation (r2 = 0.561) between FTM and a combination of power spectral density ratio and task time. However, the system could not accurately differentiate ET phenotypes according to the Consensus classification scheme. Potential applications of machine-based assessment of ET using wearable sensors include clinical trials and remote monitoring of patients.

Convolutional neural network ensemble for Parkinson's disease detection from voice recordings.

The computerized detection of Parkinson's disease (PD) will facilitate population screening and frequent monitoring and provide a more objective measure of symptoms, benefiting both patients and healthcare providers. Dysarthria is an early symptom of the disease and examining it for computerized diagnosis and monitoring has been proposed. Deep learning-based approaches have advantages for such applications because they do not require manual feature extraction, and while this approach has achieved excellent results in speech recognition, its utilization in the detection of pathological voices is limited. In this work, we present an ensemble of convolutional neural networks (CNNs) for the detection of PD from the voice recordings of 50 healthy people and 50 people with PD obtained from PC-GITA, a publicly available database. We propose a multiple-fine-tuning method to train the base CNN. This approach reduces the semantical gap between the source task that has been used for network pretraining and the target task by expanding the training process by including training on another dataset. Training and testing were performed for each vowel separately, and a 10-fold validation was performed to test the models. The performance was measured by using accuracy, sensitivity, specificity and area under the ROC curve (AUC). The results show that this approach was able to distinguish between the voices of people with PD and those of healthy people for all vowels. While there were small differences between the different vowels, the best performance was when/a/was considered; we achieved 99% accuracy, 86.2% sensitivity, 93.3% specificity and 89.6% AUC. This shows that the method has potential for use in clinical practice for the screening, diagnosis and monitoring of PD, with the advantage that vowel-based voice recordings can be performed online without requiring additional hardware.

Muscle activation strategies of people with early-stage Parkinson's during walking.

INTRODUCTION: Some people with Parkinson's disease (PD) frequently have an unsteady gait with shuffling, reduced strength, and increased rigidity. This study has investigated the difference in the neuromuscular strategies of people with early-stage PD, healthy older adults (HOA) and healthy young adult (HYA) during short-distance walking. METHOD: Surface electromyogram (sEMG) was recorded from tibialis anterior (TA) and medial gastrocnemius (MG) muscles along with the acceleration data from the lower leg from 72 subjects-24 people with early-stage PD, 24 HOA and 24 HYA during short-distance walking on a level surface using wearable sensors. RESULTS: There was a significant increase in the co-activation, a reduction in the TA modulation and an increase in the TA-MG lateral asymmetry among the people with PD during a level, straight-line walking. For people with PD, the gait impairment scale was low with an average postural instability and gait disturbance (PIGD) score = 5.29 out of a maximum score of 20. Investigating the single and double support phases of the gait revealed that while the muscle activity and co-activation index (CI) of controls modulated over the gait cycle, this was highly diminished for people with PD. The biggest difference between CI of controls and people with PD was during the double support phase of gait. DISCUSSION: The study has shown that people with early-stage PD have high asymmetry, reduced modulation, and higher co-activation. They have reduced muscle activity, ability to inhibit antagonist, and modulate their muscle activities. This has the potential for diagnosis and regular assessment of people with PD to detect gait impairments using wearable sensors.

Computerised analysis of writing and drawing by essential tremor phenotype.

We investigated whether computerised analysis of writing and drawing could discriminate essential tremor (ET) phenotypes according to the 2018 Consensus Statement on the Classification of Tremors. The Consensus scheme emphasises soft additional findings, mainly motor, that do not suffice to diagnose another tremor syndrome. Ten men and nine women were classified by blinded assessors according to Consensus Axis 1 definitions of ET and ET plus. Blinded scoring of tremor severity and alternating limb movement was also conducted. Twenty healthy participants acted as controls. Four writing and three drawing tasks were performed on a Wacom Intuos Pro Large digital tablet with a pressure-sensor mounted ink pen. Sixty-seven computerised measurements were obtained, comprising static (dimensional and temporal), kinematic and pen pressure features. The mean age of ET participants was 67.2±13.0 years and mean tremor duration was 21.7±19.0 years. Six were classified as ET, five had one plus feature and eight had two plus features. The computerised analysis could predict the presence and number of ET plus features. Measures of acceleration and variation of pen pressure performed strongly to separate ET phenotypes (p<0.05). Plus features were associated with higher scores on the Fahn-Tolosa-Marin Tremor Rating Scale (p=0.001) and it appeared that ET groups were mainly being separated according to severity of tremor and by compensatory manoeuvres used by participants with more severe tremor. There were, in addition, a small number of negative kinematic correlations suggesting some slowness with ET plus. Abnormal repetitive limb movement was also correlated with tremor severity (R=0.57) by clinical grading. Critics of the Consensus Statement have drawn attention to weaknesses of the ET plus concept in relation to duration and severity of ET. This classification of ET may be too biased towards tremor severity to assist in distinguishing underlying biological differences by clinical measurement.

PPG Derived Respiratory Rate Estimation in Daily living Conditions.

Respiratory rate (RR) derived from photoplethysmogram (PPG) during daily activities can be corrupted due to movement and other artefacts. We have investigated the use of ensemble empirical mode decomposition (EEMD) based smart fusion approach for improving the RR extraction from PPG. PPG was recorded while subjects performed five different activities: sitting, standing, climbing and descending stairs, walking, and running. RR was obtained using EEMD and smart fusion. The median absolute error (AE) of the proposed method is superior, median AE = 3.05 (range 3.01 to 3.18) breath/min in estimating RR during five different activities. Therefore, the proposed method can be implemented for overcoming the artefact problems when recording continuous RR monitoring during activities of daily living.

Is Thermal Imaging a Useful Predictor of the Healing Status of Diabetes-Related Foot Ulcers? A Pilot Study.

INTRODUCTION: In clinical practice, both area and temperature of the ulcer have been shown to be effective in tracking the healing status of diabetes-related foot ulcer (DRFU). However, traditionally, the area of the DRFU is measured regardless of the temperature distribution. The current prospective, observational study used thermal imaging, as a more accurate tool, to measure both the area and the temperature of DRFU. We aimed to predict healing of DRFU using thermal imaging within the first 4 weeks of ulceration. METHOD: A pilot study was conducted where thermal and color images of 26 neuropathic DRFUs (11 healing vs 15 nonhealing) from individuals with type 1 or 2 diabetes were taken at the initial clinic visit (baseline), at week 2, and at week 4. The thermal images were segmented into isothermal patches to identify the wound boundary and area corresponding to temperature distribution. Five parameters were obtained: temperature of the wound bed, area of the isothermal patch of the wound bed, area of isothermal patch of periwound, number of isolated isothermal patches of the wound region, and physical wound bed area from color image. The ulcers were also measured by experienced podiatrists over 4 consecutive weeks and used as the healing reference. RESULTS: For healing cases, the ratio of the area of the wound bed to its baseline measured using thermal images was found to be significantly lower at 2 weeks compared to nonhealing cases and this corresponded with a 50% reduction in area of DRFU at 4 weeks (group rank-based nonparametric analysis of variance P = .036). In comparison, neither the planimetric area measured using color images nor the temperature of the wound bed was associated with the healing. CONCLUSION: This study of 26 patients demonstrates that change in the isothermal area of DRFU can predict the healing status at week 4. Thermal imaging of DRFUs has the advantage of incorporating both area and temperature allowing for early prediction of the healing of these ulcers. Further studies with greater sample sizes are required to test the significance of these results.

Exudate detection in fundus images using deeply-learnable features.

Presence of exudates on a retina is an early sign of diabetic retinopathy, and automatic detection of these can improve the diagnosis of the disease. Convolutional Neural Networks (CNNs) have been used for automatic exudate detection, but with poor performance. This study has investigated different deep learning techniques to maximize the sensitivity and specificity. We have compared multiple deep learning methods, and both supervised and unsupervised classifiers for improving the performance of automatic exudate detection, i.e., CNNs, pre-trained Residual Networks (ResNet-50) and Discriminative Restricted Boltzmann Machines. The experiments were conducted on two publicly available databases: (i) DIARETDB1 and (ii) e-Ophtha. The results show that ResNet-50 with Support Vector Machines outperformed other networks with an accuracy and sensitivity of 98% and 0.99, respectively. This shows that ResNet-50 can be used for the analysis of the fundus images to detect exudates.

High-Resolution Spectral Analysis Accurately Identifies the Bacterial Signature in Infected Chronic Foot Ulcers in People With Diabetes.

Diabetic foot infections are a major cause of hospitalization, and delayed treatment can lead to numerous complications. The aim of this research was to investigate high-resolution spectroscopy of the wound center and periwound area for real-time estimation of multispectral signature of bacteria at the base of diabetic foot ulcers. We investigated the spectrum of the reflected visual light from diabetic foot ulcers and developed a method that identifies the presence of bacteria in the wound infections. We undertook a prospective pilot study on 18 patients with type 1 and type 2 diabetes and chronic diabetic foot ulcers. The spectral coefficients were directly compared with the results from the wound swab. The results of the multispectral analysis demonstrated 100% sensitivity, with 100% negative predictive values of identifying the presence of the bacteria, which was the cause of the infection in the wound. The results of our study suggest that the changes in the multispectral properties of the wound can be used to identify the presence of bacteria in the infected area using a noninvasive device without any contact with the wound. This technique holds great promise for real-time objective evaluation of the wound infection status beyond the standard visual assessment of diabetic foot ulcers.

Computational model to investigate the relative contributions of different neuromuscular properties of tibialis anterior on force generated during ankle dorsiflexion.

This study describes a new model of the force generated by tibialis anterior muscle with three new features: single-fiber action potential, twitch force, and pennation angle. This model was used to investigate the relative effects and interaction of ten age-associated neuromuscular parameters. Regression analysis (significance level of 0.05) between the neuromuscular properties and corresponding simulated force produced at the footplate was performed. Standardized slope coefficients were computed to rank the effect of the parameters. The results show that reduction in the average firing rate is the reason for the sharp decline in the force and other factors, such as number of muscle fibers, specific force, pennation angle, and innervation ratio. The fast fiber ratio affects the simulated force through two significant interactions. This study has ranked the individual contributions of the neuromuscular factors to muscle strength decline of the TA and identified firing rate decline as the biggest cause followed by decrease in muscle fiber number and specific force. The strategy for strength preservation for the elderly should focus on improving firing rate. Graphical abstract Neuromuscular properties of Tibialis Anterior on force generated during ankle dorsiflexion.

Age-Associated Changes in the Spectral and Statistical Parameters of Surface Electromyogram of Tibialis Anterior.

Age-related neuromuscular change of Tibialis Anterior (TA) is a leading cause of muscle strength decline among the elderly. This study has established the baseline for age-associated changes in sEMG of TA at different levels of voluntary contraction. We have investigated the use of Gaussianity and maximal power of the power spectral density (PSD) as suitable features to identify age-associated changes in the surface electromyogram (sEMG). Eighteen younger (20-30 years) and 18 older (60-85 years) cohorts completed two trials of isometric dorsiflexion at four different force levels between 10% and 50% of the maximal voluntary contraction. Gaussianity and maximal power of the PSD of sEMG were determined. Results show a significant increase in sEMG's maximal power of the PSD and Gaussianity with increase in force for both cohorts. It was also observed that older cohorts had higher maximal power of the PSD and lower Gaussianity. These age-related differences observed in the PSD and Gaussianity could be due to motor unit remodelling. This can be useful for noninvasive tracking of age-associated neuromuscular changes.

Automatic Analysis of Retinal Vascular Parameters for Detection of Diabetes in Indian Patients with No Retinopathy Sign.

This study has investigated the association between retinal vascular parameters with type II diabetes in Indian population with no observable diabetic retinopathy. It has introduced two new retinal vascular parameters: total number of branching angles (TBA) and average acute branching angles (ABA) as potential biomarkers of diabetes in an explanatory model. A total number of 180 retinal images (two (left and right) × two (ODC and MC) × 45 subjects (13 diabetics and 32 nondiabetics)) were analysed. Stepwise linear regression analysis was performed to model the association between type II diabetes with the best subset of explanatory variables (predictors), consisting of retinal vascular parameters and patients' demographic information. P value of the estimated coefficients (P < 0.001) indicated that, at α level of 0.05, the newly introduced retinal vascular parameters, that is, TBA and ABA together with CRAE, mean tortuosity, SD of branching angle, and VB, are related to type II diabetes when there is no observable sign of retinopathy.

Application-Dedicated Selection of Filters (ADSF) using covariance maximization and orthogonal projection.

Visible and near-infrared (Vis-NIR) spectra are generated by the combination of numerous low resolution features. Spectral variables are thus highly correlated, which can cause problems for selecting the most appropriate ones for a given application. Some decomposition bases such as Fourier or wavelet generally help highlighting spectral features that are important, but are by nature constraint to have both positive and negative components. Thus, in addition to complicating the selected features interpretability, it impedes their use for application-dedicated sensors. In this paper we have proposed a new method for feature selection: Application-Dedicated Selection of Filters (ADSF). This method relaxes the shape constraint by enabling the selection of any type of user defined custom features. By considering only relevant features, based on the underlying nature of the data, high regularization of the final model can be obtained, even in the small sample size context often encountered in spectroscopic applications. For larger scale deployment of application-dedicated sensors, these predefined feature constraints can lead to application specific optical filters, e.g., lowpass, highpass, bandpass or bandstop filters with positive only coefficients. In a similar fashion to Partial Least Squares, ADSF successively selects features using covariance maximization and deflates their influences using orthogonal projection in order to optimally tune the selection to the data with limited redundancy. ADSF is well suited for spectroscopic data as it can deal with large numbers of highly correlated variables in supervised learning, even with many correlated responses.

Fractal and twin SVM-based handgrip recognition for healthy subjects and trans-radial amputees using myoelectric signal.

Identifying functional handgrip patterns using surface electromygram (sEMG) signal recorded from amputee residual muscle is required for controlling the myoelectric prosthetic hand. In this study, we have computed the signal fractal dimension (FD) and maximum fractal length (MFL) during different grip patterns performed by healthy and transradial amputee subjects. The FD and MFL of the sEMG, referred to as the fractal features, were classified using twin support vector machines (TSVM) to recognize the handgrips. TSVM requires fewer support vectors, is suitable for data sets with unbalanced distributions, and can simultaneously be trained for improving both sensitivity and specificity. When compared with other methods, this technique resulted in improved grip recognition accuracy, sensitivity, and specificity, and this improvement was significant (κ=0.91).

Effect of number of motor units and muscle fibre type on surface electromyogram.

Reduction in number of motor units (nMU) and fast fibre ratio (FFR) is associated with disease or atrophy when this is rapid. There is a need to study the effect of nMU and FFR to analyse the association with ageing and disease. This study has developed a mathematical model to investigate the relationship between nMU and FFR on surface electromyogram (sEMG) of the biceps muscles. The model has been validated by comparing the simulation outcomes with experiments comparing the sEMG of physically active younger and older cohort. The results show that there is statistically significant difference between the two groups, and the simulation studies closely model the experimental results. This model can be applied to identify the cause of muscle weakness among the elderly due to factors such as muscle dystrophy or preferential loss of type F muscle fibres.

Relationship between diabetes and grayscale fractal dimensions of retinal vasculature in the Indian population.

BACKGROUND: Diabetes mellitus is rapidly increasing in the Indian population. The purpose of this study was to identify changes in the retinal vasculature of diabetic people, ahead of visual impairments. Grayscale Fractal Dimension (FD) analysis of retinal images was performed on people with type 2 diabetes from an Indian population. METHODS: A cross-sectional study comprising 189 Optic Disc (OD) centred retinal images of healthy and diabetic individuals aged 14 to 73 years was conducted. Grayscale Box Counting FD of these retinal photographs was measured without manual supervision. Statistical analysis was conducted to determine the difference in the FD between diabetic and healthy (non-diabetic) people. RESULTS: The results show that grayscale FD values for diabetic cases are higher compared to controls, irrespective of the gender. It was also observed that FD was higher for male compared with females. CONCLUSIONS: There is difference in the grayscale fractal dimension of retinal vasculature of diabetic patients and healthy subjects, even when there is no reported retinopathy.

Zone specific fractal dimension of retinal images as predictor of stroke incidence.

Fractal dimensions (FDs) are frequently used for summarizing the complexity of retinal vascular. However, previous techniques on this topic were not zone specific. A new methodology to measure FD of a specific zone in retinal images has been developed and tested as a marker for stroke prediction. Higuchi's fractal dimension was measured in circumferential direction (FDC) with respect to optic disk (OD), in three concentric regions between OD boundary and 1.5 OD diameter from its margin. The significance of its association with future episode of stroke event was tested using the Blue Mountain Eye Study (BMES) database and compared against spectrum fractal dimension (SFD) and box-counting (BC) dimension. Kruskal-Wallis analysis revealed FDC as a better predictor of stroke (H = 5.80, P = 0.016, α = 0.05) compared with SFD (H = 0.51, P = 0.475, α = 0.05) and BC (H = 0.41, P = 0.520, α = 0.05) with overall lower median value for the cases compared to the control group. This work has shown that there is a significant association between zone specific FDC of eye fundus images with future episode of stroke while this difference is not significant when other FD methods are employed.