Objectives: This paper focuses on developing a regularization-based feature selection approach to select the most effective attributes from the Parkinson’s speech dataset. Parkinson’s disease is a medical condition that progresses as the dopamine-producing nerve cells are affected. Early diagnosis often reduces the effect on the individuals, minimizes the advancement over time. In recent times, intelligent computational models are used in many complex cases to diagnose a clinical condition with high precision. These models are intended to find meaningful representation from the data to diagnose the disease. Machine learning acts as a tool, gears up the model learning process through a mathematical baseline. But, not in all cases, machine learning will be demanded to perform optimally. It comes with a few constraints, mainly the representation of the data. The learning models expect a clean, noise-free input, which in-turns produces better discriminative patterns over different categories of classes. Methods: The proposed model identified five candidate features as predictors. This feature subset is trained with different varieties of supervised classifiers to trace out the best-performing model. Results: The results are validated through accuracy, precision, recall, and receiver’s operational characteristic curves. The proposed regularization- based feature selection model outperformed the benchmark algorithms by attaining 100% accuracy on most of the classifiers, other than linear discriminant analysis (99.90%) and naïve Bayes (99.51%). Conclusions: This paper exhibits the need for intelligent models to analyze complex data patterns to assist medical practitioners in better disease diagnosis. The results exhibit that the regularization methods find the best features based on their importance score, which improved the model performance over other feature selection methods.
2
Dostęp do pełnego tekstu na zewnętrznej witrynie WWW
The lack of dopamine in the human brain is the cause of Parkinson disease (PD) which is a degenerative disorder common globally to older citizens. However, late detection of this disease before the first clinical diagnosis has led to increased mortality rate. Research effort towards the early detection of PD has encountered challenges such as: small dataset size, class imbalance, overfitting, high false detection rate, model complexity, etc. This paper aims to improve early detection of PD using machine learning through data augmentation for very small datasets. We propose using Spline interpolation and Piecewise Cubic Hermite Interpolating Polynomial (Pchip) interpolation methods to generate synthetic data instances. We further investigate on reducing dimensionality of features for effective and real-time classification while considering computational complexity of implementation on real-life mobile phones. For classification we use Bidirectional LSTM (BiLSTM) deep learning network and compare the results with traditional machine learning algorithms like Support Vector Machine (SVM), Decision Tree, Logistic regression, KNN and Ensemble bagged tree. For experimental validation we use the Oxford Parkinson disease dataset with 195 data samples, which we have augmented with 571 synthetic data samples. The results for BiLSTM shows that even with a holdout of 90%, the model was still able to effectively recognize PD with an average accuracy for ten rounds experiment using 22 features as 82.86%, 97.1\%, and 96.37% for original, augmented (Spline) and augmented (Pchip) datasets, respectively. Our results show that proposed data augmentation schemes have significantly (p < 0.001) improved the accuracy of PD recognition on a small dataset using both classical machine learning models and BiLSTM.
Parkinson’s disease is a neurodegenerative and progressive disease of the central nervous system. It affects more than 10 million patients worldwide and the symptoms allow for little to no control for movement. These symptoms appear because the chemical messenger dopamine is being made in small quantities from impaired cells. However, the disease often forms when there is a mutation in the LRRK2 gene, as the functions of the protein become abnormal. The IC50 value is essential information about molecules because it measures their effectiveness. The goal of this research was to design molecules with a lower IC50 value. This was first done by modeling molecules on the molecular modeling program, Gaussian 09. Modifications were made to molecules that were said to bind to the LRRK2 protein. Modifications ranged from adding a single atom or replacing atoms with groups. After running these molecules on the program, the total energy was found. Using the equation found from the correlation between 1/IC50 and the total energy, the IC50 value was predicted for each of the modified molecules. Many of the modified molecules portrayed a positive percent difference between the original IC50 value and the new one. This saves both time and money because the molecules with lower IC50 values can be made, preserving the resources. After creating the molecule with a low IC50 value, further experimental procedures can be taken; this is a large step in assisting researchers to reach a potential treatment because it is more efficient.
4
Dostęp do pełnego tekstu na zewnętrznej witrynie WWW
Przedmiotem pracy jest kliniczny system oceny stopnia dysfunkcji manualnej osób cierpiących na schorzenia drżenia rąk. Podlegająca późniejszej ocenie informacja o drżeniu rąk pobierana jest poprzez panel dotykowy tablet-PC, w sposób symulujący klasyczną metodę spirali Archimedesa. Ruch ręki sparametryzowano przez 15 arbitralnie dobranych innowacyjnych współczynników amplitudowych, czasowych oraz częstotliwościowych. Ich analizę przeprowadza sztuczna sieć neuronowa, wystawiająca badanym oceny w skali 0-10. Trafny dobór parametrów potwierdzany jest przez średni stopień ich korelacji z ocenami. Dalsza optymalizacja modułu analitycznego może w konsekwencji umożliwić stosowanie prototypowanego rozwiązania w praktyce klinicznej.
EN
The main aim of current article is to introduce prototyped system of hand tremor analysis for people suffering from Parkinson's disease. It was designed after authors found out, that there is significant need for system that would allow assessing peoples hand tremor level in some normalized manner [1, 4]. Whole application, developed using LabVIEW 2009 graphical environment [11], was divided into two independent modules - acquisition and analysis. The signal acquisition is bases on classic Archimedean Spiral method, involving tablet-PC touch screen and specialized pen instead of paper and pencil. Patients hand movement is recorded during the whole process of drawing spiral, and then described using 15 innovative amplitude, acceleration and frequency parameters. Those coefficients stand as input values for the second module - analysis. In this case a three layer cascade artificial neural network with backpropagation was utilized for best performance and flexibility [4, 8]. Network was implemented using MATLAB R2008a environment and taught by marks given to patients by neurologists. Studies carried out on group of patients with Parkinson's disease shown, that chosen parameters have correlation level of 0.58, with marks given by specialists. Moreover, database structure allows provided software to be connected with Hospital Information System [2], so the developed application is well suitable for being widely used in clinical practice.
Cel pracy: Ocena wpływu ćwiczeń ruchowych na czynność mięśni działających na stawy kolanowe u pacjentów z chorobą Parkinsona. Grupa badana: 10 pacjentów (kobiet) z chorobą Parkinsona, średni wiek 72,9 lat (grupa badana), oraz 12 osób zdrowych (kobiety), słuchaczki Uniwersytetu Trzeciego Wieku, średni wiek 65,0 lat (grupa kontrolna). Metoda badań: Wykonano badania czynnościowe mięśni zginaczy i prostowników stawu kolanowego na stanowisku do badań izokinetycznych Biodex System 3 Multi Joint. Badania przeprowadzono przed przystąpieniem do usprawniania pacjentów leczonych z powodu choroby Parkinsona oraz po okresie dwóch miesięcy regularnych ćwiczeń. Wyniki: W grupie pierwszej, w przypadku prędkości kątowej 60%, uzyskano istotną poprawę wszystkich analizowanych parametrów mięśni zginaczy i prostowników, zarówno w obrębie kończyny dolnej prawej, jak i lewej. Natomiast w przypadku prędkości kątowej 1807s istotnie wzrosły analizowane parametry dla mięśni prostowników i zginaczy stawu kolanowego prawego oraz dla mięśni prostowników stawu kolanowego lewego. W grupie 2. nie stwierdzono żadnych istotnych różnic. Wnioski: Ćwiczenia ruchowe prowadzone u osób z chorobą Parkinsona poprawiają czynność mięśni kończyn dolnych, w związku z powyższym mogą zmniejszyć ryzyko występowania upadków oraz powikłań z nimi związanych.
EN
The influence of the regular exercises on the function of lower extremities muscles acting on the knee joints in patients with Parkinson's disease. Subjects: 10 women with Parkinson's disease at the average age of 72,9 and 12 healthy females at the average age of 65,0 (control group). Methods: Biodex dynamometer was used to measure strength - velocity parameters in the knee flexors and extensors muscles. The examination were performed before the period of patients rehabilitation and after two months of regular exercises. Results: In a group 1, in case of the angular velocity of 60°/s, a significant improvement of all analyzed parameters of flexor and extensor muscles was obtained, both in the right and the left lower extremity. In the case of the angular velocity of 1807s, a significant increase of analyzed parameters of flexor and extensor muscles was noted. No significant differences were found in a group 2. Conclusions: Exercises in case of patients with Parkin-son's disease increase the muscular function of lower extremities and therefore may decrease the risk of falls and complications resulting from injures.
Ocenia się, że dysartryczne zaburzenia mowy w chorobie Parkinsona występują u 70-90% chorych, a leczenie farmakologiczne przynosi niewielkie efekty. Celem pracy była analiza komputerowa wpływu postępowania logopedycznego na strukturę akustyczną dźwięków mowy w grupie chorych z rozpoznaniem choroby Parkinsona. Materiał stanowiło 10 chorych (6 kobiet, 4 mężczyzn, w wieku 59-75 lat, średnio - 68,2), z rozpoznaniem choroby Parkinsona i z zaburzeniami mowy o charakterze dysar-trycznym. Wszyscy byli leczeni preparatami L-dopy. Przeprowadzono u nich terapię z uwzględnieniem ćwiczeń fonacji, artykulacji; ukierunkowanych na zmniejszenie sztywności mięśni ust i twarzy oraz ćwiczeń oddechowych. W celu oceny skuteczności postępowania logopedycznego wykorzystano komputerową akustyczną analizę mowy, ze szczególnym uwzględnieniem następujących parametrów: selektywności poszczególnych elementów wypowiedzi, tempa oraz stabilności dźwięków mowy. W analizie akustycznej zastosowano programy komputerowe IRIS oraz Wavelab. Stwierdzono, że systematycznie stosowane postępowanie logopedyczne u pacjentów z rozpoznaniem choroby Parkinsona i towarzyszącymi zaburzeniami mowy o typie dysartrii wpływa na stabilizację wybranych parametrów akustycznych mowy oraz poprawę sprawności mowy, co znalazło potwierdzenie w akustycznej analizie komputerowej.
EN
Dysarthria occurs in case of 70-90% of patients with Parkinson's disease (PD). Pharmacological treatment of this symptom is not effective enough. The aim of this work was the computer analysis of the influence of speech therapy on the acoustic structure of speech sounds in patients with Parkinson's disease. Material and methods: 10 patients (6 women and 4 men) aged from 59 to 75 years (mean age - 68.2) with PD and dysarthria were examined. All patients were treated by the L-Dopa and the logopedie examination was carried out. The administrated pronation and articulation exercises aimed to facial and oral muscle rigidity reduction. The breathing training was included in the therapeutic programme, as well. Computer speech analysis was applied for the evaluation of speech rehabilitation efficacy. Selectiveness of the individual parts of speech and stability of frequency estimation were analyzed using IRIS and Wavelab computer programs. Systematic logopedie therapy in PD patients with dysarthria influences the speech acoustic parameters and enables the speech efficiency improvement. Computer acoustic analysis allows to follow the therapy efficacy.
JavaScript jest wyłączony w Twojej przeglądarce internetowej. Włącz go, a następnie odśwież stronę, aby móc w pełni z niej korzystać.