Speech analysis systems: an evaluation.

C Read; E H Buder; R D Kent

doi:10.1044/jshr.3502.314

Home/Archive/PubMed

PubMed · 1573872

Speech analysis systems: an evaluation.

C Read·E H Buder·R D Kent

View original source ↗Export metadata ↓

Abstract

Performance characteristics are reviewed for seven systems marketed for acoustic speech analysis: CSpeech, CSRE, ILS-PC, Kay Elemetrics model 5500 Sona-Graph, MacSpeech Lab II, MSL, and Signalyze. The characteristics reviewed include system components, basic capabilities (signal acquisition, waveform operations, analysis, and other functions), documentation, user interface, data formats and journaling, speed and precision of spectral analysis, and speed and precision of fundamental frequency analysis. Basic capabilities are also tabulated for three recently introduced systems: the Sensimetrics SpeechStation, the Kay Elemetrics Computerized Speech Lab (CSL), and the LSI Speech Workstation. In addition to the capability and performance summaries, this article offers suggestions for continued development of speech analysis systems, particularly in data exchange, journaling, display features, spectral analysis, and fundamental frequency analysis.

Explore related subjects

Diagnosis, Computer-Assisted Documentation Efficiency Humans Microcomputers Sound Spectrography Speech Production Measurement User-Computer Interface Journal Article Research Support, U.S. Gov't, P.H.S.Review

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

C Read, E H Buder, R D Kent. 1992. Speech analysis systems: an evaluation.. https://doi.org/10.1044/jshr.3502.314

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related citations

PubMed2001

Is dermoscopy useful for the diagnosis of melanoma?

Explore the source record for details and available documents.

Diagnosis, Computer-Assisted↗

PubMed2001

An artificial neural network approach to diagnosing epilepsy using lateralized bursts of theta EEGs.

Determining the cause of seizures is a significant medical problem, as misdiagnosis can result in increased morbidity and even mortality of patients. The reported research evaluates the efficacy of using an artificial neural network (ANN) for determining epileptic seizure occurrences for patients with lateralized bursts of theta (LBT) EEGs. Training and test cases are acquired from examining records of 1,500 consecutive adult seizure patients. The small resulting pool of 92 patients with LBT EEGs requires using a jack-knife procedure for developing the ANN categorization models. The ANNs are evaluated for accuracy, specificity, and sensitivity on classification of each patient into the correct two-group categorization: epileptic seizure or non-epileptic seizure. The original ANN model using eight variables produces a categorization accuracy of 62%. Following a modified factor analysis, an ANN model utilizing just four of the original variables achieves a categorization accuracy of 68%.

Diagnosis, Computer-Assisted↗

PubMed2001

TERVA: system for long-term monitoring of wellness at home.

Long-term monitoring of physiological and psychosocial variables in out-hospital conditions would be beneficial for investigating changes in wellness status of an individual or to understand interaction between physiological and behavioral processes. We aimed to design a personal wellness monitoring system (TERVA), which would allow monitoring of wellness-related variables at home for several weeks or even months. The designed TERVA system runs on a laptop computer and interfaces with different measurement devices through a serial interface. Measured variables include beat-to-beat heart rate, motor activity, blood pressure, weight, body temperature, respiration, ballistocardiography, movements, and sleep stages. In addition, self-assessments of daily well-being and activities are stored by keeping a behavioral diary. To test the system, one healthy man used the system for 10 weeks. The system was successfully applied in out-hospital conditions. The success rate of the measurements was 70-91%, depending on the variable under consideration. The pilot study indicated that the recorded data accurately reflected the health status of the subject. The TERVA system provides a method to record and investigate wellness-related data over several weeks, or even months, outside the hospital among subjects capable of using a personal computer. Several applications of the system are discussed.

Diagnosis, Computer-Assisted↗