Parkinson’s disease is a relatively common illness, constantly progressing, evoking anxiety and depression resulting from significantly restricted independence. Its dominating symptoms include muscle rigidity, rest tremor, slowness and postural disorders. They can lead to progressing gait disorders, hypomimia, extrapyramidal dysarthria and micrography. Acquisition of data imaging the condition of a patient and their objective evaluation can constitute a valuable tool improving the diagnostics and treatment monitoring. The developed data acquisition set enables recording data while the patient is performing selected tests, based on the UDPRS scale. The set consists of an infrared camera, a camera operating within the visible range, a microphone with a preamplifier, a graphic tablet and a laptop. A set of recording devices controlled by a graphic user interface guarantees the acquisition of data for the purposes of studying primarily motor symptoms. The set of recorded data includes a face image in the visible light range and in infrared for studying hypomimia, video images of the limbs for studying finger tapping movement regularity, hand movement regularity, lower limb agility and gait, and spontaneous and forced speech samples to evaluate the strength of the voice, its timbre and quality. In addition, the used graphic tablet enables collecting handwriting samples for testing writing speed and the force used. The suggested solution enables non-invasive quantitative measurements and archiving multimodal data describing the condition of a patient, which after processing can be used in diagnostics, evaluating treatment effectiveness and studying the progression of the disease.
KEYWORDS: Acoustics, Signal processing, Analytical research, Parkinson's disease, Time-frequency analysis, Diagnostics, Signal analyzers, Neurology, Lung, Medical research
Parkinson’s Disease (PD) is a neurodegenerative disease, which is becoming an increasingly greater social issue due to the growing incidence rate caused by population ageing. Over the recent years, the doctors have been focusing on searching for new methods supporting the diagnosis of such disorders. Acoustic voice analysis in Parkinson’s patients can be a valuable and objective tool supporting the diagnosing diseases of neurodegenerative nature. The article discusses a concept of utilizing voice processing techniques in evaluating patients with Parkinson’s disease. Using time analysis, frequency analysis and time-frequency analysis, the authors attempted performing acoustic voice analysis in that group of patients. The research utilized recordings conducted at the Department of Neurology at the Medical University of Warsaw. The study involved both women and men. The recording scenario was divided into several parts. The first part contained two various texts read out by the tested person. Yet another segment involved recording the vowel “a” with prolonged phonation, uttered by a patient for at least 5 seconds. The last part of the recordings involved the patient uttering individual words and sentences according to an assumed scenario. A total of 7 acoustic signals per patient, with an average length of ca. 75 seconds were recorded. The conclusions from the conducted studies will enable determining, which of the applied techniques can be a promising tool for supporting the diagnostics of neurodegenerative disorders, including parkinsonism. Further studies involving larger groups are required in order to confirm the obtained results and the structure of a target diagnostic system.
Engineering support in the field of distinguishing Parkinson's disease from other diseases, diagnosing its progression and monitoring the effectiveness of drug treatment is nowadays implemented by way of recording and analyzing equipment fitted with motion sensors. The time series they provide enable quantitative evaluation of a set of symptoms describing daily activities and motor abilities of patients. The paper presents the preliminary results of fundamental research, which based on known medical observations indicating the diminution of facial expressions and micrographic apart from general motor deterioration, suggest that the clinical studies could utilize the techniques of processing image data acquired during the medical history taking. The image data includes video recording of the face and limbs conducted in the course of the coercions suggested in the study and manual drawings by patients. The image data are redundant and require processing for presentations facilitating their interpretation by a physician and enabling efficient utilization of machine learning algorithms in the next study stage. Within the framework of preliminary processing of acquired images, attempts were made to determine the quantitative measures, such as, e.g. blinking frequency and the indicators generated as a result of analyzing the position of characteristic points within the facial image. In the case of limbs, it is suggested to reproduce the motion on the image using a time series acquired thanks to the fixed markers. Preliminary processing of data coming from a graphic tablet also guarantees the generation of time series for images created by patients.
KEYWORDS: Throat, Speaker recognition, Signal processing, Databases, Signal detection, Diagnostics, Data modeling, System identification, Biometrics, Systems modeling
The paper presents the results of experiments that allow to determine the validity of use of the Automatic Speaker Recognition (ASR) in devices that register voice signal via throat microphones. Throat microphones are used in extremely difficult registration conditions, e.g. in tank headsets, therefore high efficiency of identification of speakers obtained for such a device allows to significantly extend the range of applications of ASR. The presented research results are complemented with results of speakers identification tests carried out using a traditional electret microphone on the same population of voices, which allows for unequivocal comparison of the impact of an acoustic channel on operation of ASR. The paper also includes a brief description of operation of the automatic speaker recognition system based on a cepstral analysis of a voice signal and Gaussian Mixture Models (GMM).
The World Health Organization (WHO) figures clearly indicate that cardiovascular disease is the most common cause of death and disability in the world. Early detection of cardiovascular pathologies may contribute to reducing such a high mortality rate. Auscultatory examination is one of the first and most important step in cardiologic diagnostics. Unfortunately, proper diagnosis is closely related to long-term practice and medical experience. The article presents the author's system of recording phonocardiograms and the way of saving data, as well as the outline of the analysis algorithm, which will allow to assign a case to a patient with heart failure or healthy voluntaries’ with a certain high probability. The results of a pilot study of phonocardiographic signals were also presented as an introduction to further research aimed at the development of an efficient diagnostic algorithm based on spectral analysis of the heart tone.
Diagnosis of part of the visual system, that is responsible for conducting compound action potential, is generally based on visual evoked potentials generated as a result of stimulation of the eye by external light source. The condition of patient’s visual path is assessed by set of parameters that describe the time domain characteristic extremes called waves. The decision process is compound therefore diagnosis significantly depends on experience of a doctor. The authors developed a procedure – based on wavelet decomposition and linear discriminant analysis – that ensures automatic classification of visual evoked potentials. The algorithm enables to assign individual case to normal or pathological class. The proposed classifier has a 96,4% sensitivity at 10,4% probability of false alarm in a group of 220 cases and area under curve ROC equals to 0,96 which, from the medical point of view, is a very good result.
KEYWORDS: Interference (communication), Acoustics, Genetic algorithms, Databases, Speaker recognition, Data modeling, Signal to noise ratio, Optimization (mathematics), Global system for mobile communications, Classification systems
The paper presents the architecture and the results of optimization of selected elements of the Automatic Speaker Recognition (ASR) system that uses Gaussian Mixture Models (GMM) in the classification process. Optimization was performed on the process of selection of individual characteristics using the genetic algorithm and the parameters of Gaussian distributions used to describe individual voices. The system that was developed was tested in order to evaluate the impact of different compression methods used, among others, in landline, mobile, and VoIP telephony systems, on effectiveness of the speaker identification. Also, the results were presented of effectiveness of speaker identification at specific levels of noise with the speech signal and occurrence of other disturbances that could appear during phone calls, which made it possible to specify the spectrum of applications of the presented ASR system.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.