Wieslaw Wszolek

A vibroacoustic model of selected human larynx diseases.

With the present development of digital registration and methods for processing speech it is possible to make effective objective acoustic diagnostics for medical purposes. These methods are useful as all pathologies and diseases of the human vocal tract influence the quality of a patient’s speech signal. Diagnostics of the voice organ can be defined as an unambiguous recognition of the current condition of a specific voice source. Such recognition is based on an evaluation of essential acoustic parameters of the speech signal. This requires creating a vibroacoustic model of selected deformations of Polish speech in relation to specific human larynx diseases. An analysis of speech and parameter mapping in 29-dimensional space is reviewed in this study. Speech parameters were extracted in time, frequency and cepstral (quefrency) domains resulting in diagrams that qualified symptoms and conditions of selected human larynx diseases. The paper presents graphically selected human larynx diseases.

Recognition of defects in high voltage transmission lines using the acoustic signal of corona effect

The paper deals with the analysis of the possible application of neural networks to the recognition of typical damage of UHV transmission lines. The acoustic signal generated as a result of corona effects is used as a damage symptom, as its intensity is usually increased after damage occurrence or after contamination of the surface of a conductor or an insulator string. The primary problem in the diagnostic process is the distinguishing between signals generated as results of damage and contamination. The problem is not solved by methods based on the RF signal interference or by the classical methods of acoustic signal analysis. The construction and verification of the assumed diagnostic model have been carried out by experimental studies in laboratory conditions, where typical damage and contamination of the transmission line elements have been simulated.

Processing and classification of deformed speech using neural networks

In many problems of medical diagnosis as well as therapy and rehabilitation, evaluation of the deformed speech signal quality is required. In problems of distorted speech diagnosis the regular methods of speech signal preprocessing and classification, used in speech or voice recognition, totally fail. Also the standard speech signal parametrization techniques (e.g. LPC or cepstral coefficients) cannot satisfactorily describe pathological speech because of its dissimilar phonetic and acoustic structure compared with correct speech, and also because the aim of the recognition process is totally different. In the paper a new method for processing and classification of pathologically deformed speech, based on neural networks techniques, is presented and discussed.

Artificial Intelligence Methods in Diagnostics of the Pathological Speech Signals

In the work excerpts of research are presented, concerning the application of modified acoustic signal processing methods in the problem of “understanding” of selected pathologies of vocal tract. The concept of the research scheme is based on the technique of advanced acoustic signal analysis and it refers to the analysis of artificial neural networks functioning in the task of recognition of selected types of vocal tract pathologies. The method is based on utilization of an internal model of the considered signal’s generator and it is directed towards such a structure analysis of the examined sound. The described method allows to achieve more subtle differentiation for signal characterized by small diversification of measurable features, observed for the classes being recognized, what is the case in the problem of identification of selected pathologies considered here.

Effectiveness of classical chordectomy in the treatment of cancer of the glottis

From 1980 to 1992, 85 patients diagnosed with tumour of the glottis, localized exclusively in the vocal cord area, had undergone surgery. A 90% 5-year survival rate without recurrences had been obtained. Taking into account the cases of life-saving surgery, the percentage of the 5-year survivals amounted to 94%. Considerably better results were obtained by the excision of the entire vocal cord and not just one of its sections (1/2, 2/3 etc.). The degree of deformation of the voice following chordectomy was assessed by means of acoustic and laryngographic methods (Laryngograph Processor PCLX). A 13% rate of irregularity in the functioning of the neoglottis, following surgery, was observed. The Jitter-Shimmer co-efficients were established. The results of the deformation of the voice following chordectomy were presented on the J-S scale and compared with other partial surgeries performed on patients with tumours of the glottis. All of the acoustic and laryngographic findings were presented in the from of mean values most representative of chordectomy.

The analysis of the environmental influence of drilling installation

Drilling work, due to its character, settlement, and whole‐day service, has potential acoustic danger for the environment. The degree of influence depends on many factors, e.g., the type of drilling equipment, applied acoustic screens, localization of drilling and its surroundings, the type of drilled‐trough rocks, and applied drilling technology. The main sources of acoustic noise are motors, pumps, and current generators. The distinction can be made between some procedures that are more noisy than the normal drilling process, for example, tripping. Usually the drilling rig is poorly equipped with noise reduction installation. Experiments were carried out on eight types of drilling assemblies. The received results show that different drilling equipment has an individual influence on the environment. The evident proof of that is the azimuthal acoustic characteristic and additionally the noise level dependence from the cycle of drilling work. The level of LAeq in the surroundings of the examined drilling r...

The methods of pathological speech visualization [using Kohonen neural networks]

In tasks related to the analysis and recognition of pathological speech it is often more important to provide the respective person (e.g. physician) with guidelines for a qualitative evaluation of this speech than to achieve a very accurate automated recognition. By ear it is easy to judge whether the speech is regular or deformed, but any attempt of a quantitative evaluation is not satisfactory. If the speech is transformed to a graphic form, by a proper visualization method, it is easier for a person to estimate its deformation degree by comparing the respective graphical patterns. The new visualization method proposed is based on the results obtained by application of Kohonen neural networks.

Automatic Understanding of Speech Pathology

The paper presents selected aspects of research concerning a new concept in application of computer technology to the analysis of pathological speech. This new concept assumes, that during the analysis of pathological speech the study is not focused on determining some or other signal parameters, neither it is focused on the signal classification, but it is supposed to lead to an automated understanding of the origins of the deformation, which can be revealed in analyzed signal. The presented concept therefore assumes a replacement of the well-known process of recognition of the pathological speech signal by a more advanced method of its analysis, consisting of comparison of the features revealed during the signal processing with a set of expected features resulting from application of the system’s knowledge database, provided that a specific, selected medical hypothesis is true, and concerning the pathological factors affecting the form of that speech signal. In the paper the basic elements of the proposed method are described. The examples, showing the essence of the method, have been taken from applications to selected problems of larynx pathology analyses.