LPC Spectrum First Peak Analysis for Voice Pathology Detection

José Manuel Matos Ribeiro da Fonseca

doi:10.1016/j.protcy.2013.12.123

LPC Spectrum First Peak Analysis for Voice Pathology Detection

Research output: Contribution to journal › Article › peer-review

Abstract

The detection of voice pathologies using speech processing techniques can be a useful contribution for the diagnose of larynx diseases. The main objective of this work was to inspect the spectral envelope of the voice signal searching for information that allows voice pathology detection. The frequency and bandwidth of the first peak from the spectral envelope obtained from Linear Predictive Coefficients (LPC) of 30th order was found to be a valuable feature being used for voice pathology detection. In the corpus considered in this work we obtained a 100% discrimination between healthy and unhealthy subjects and a 87% discrimination between nodules and Reinke's edema.

Original language	English
Pages (from-to)	1104-1111
Number of pages	8
Journal	Procedia Technology
Volume	9
Issue number	NA
DOIs	https://doi.org/10.1016/j.protcy.2013.12.123
Publication status	Published - 2013

Keywords

voice pathologies
nodule
Reinkes's edema
LPC spectrum

Access to Document

10.1016/j.protcy.2013.12.123Licence: CC BY

https://ac.els-cdn.com/S2212017313002776/1-s2.0-S2212017313002776-main.pdf?_tid=8ea8a888-cc58-47d4-bc48-22c45be67424&acdnat=1539943736_69ef3474dc62360cf7c7acf7b9f93b80Licence: CC BY

Cite this

@article{a78769c2b1b340f0aabcfbc3a4c90cde,

title = "LPC Spectrum First Peak Analysis for Voice Pathology Detection",

abstract = "The detection of voice pathologies using speech processing techniques can be a useful contribution for the diagnose of larynx diseases. The main objective of this work was to inspect the spectral envelope of the voice signal searching for information that allows voice pathology detection. The frequency and bandwidth of the first peak from the spectral envelope obtained from Linear Predictive Coefficients (LPC) of 30th order was found to be a valuable feature being used for voice pathology detection. In the corpus considered in this work we obtained a 100% discrimination between healthy and unhealthy subjects and a 87% discrimination between nodules and Reinke's edema.",

keywords = "voice pathologies, LPC spectrum, nodule, Reinkes's edema, voice pathologies, nodule, Reinkes's edema, LPC spectrum",

author = "Fonseca, {Jos{\'e} Manuel Matos Ribeiro da}",

year = "2013",

doi = "10.1016/j.protcy.2013.12.123",

language = "English",

volume = "9",

pages = "1104--1111",

journal = "Procedia Technology",

issn = "2212-0173",

publisher = "Elsevier",

number = "NA",

}

TY - JOUR

T1 - LPC Spectrum First Peak Analysis for Voice Pathology Detection

AU - Fonseca, José Manuel Matos Ribeiro da

PY - 2013

Y1 - 2013

N2 - The detection of voice pathologies using speech processing techniques can be a useful contribution for the diagnose of larynx diseases. The main objective of this work was to inspect the spectral envelope of the voice signal searching for information that allows voice pathology detection. The frequency and bandwidth of the first peak from the spectral envelope obtained from Linear Predictive Coefficients (LPC) of 30th order was found to be a valuable feature being used for voice pathology detection. In the corpus considered in this work we obtained a 100% discrimination between healthy and unhealthy subjects and a 87% discrimination between nodules and Reinke's edema.

AB - The detection of voice pathologies using speech processing techniques can be a useful contribution for the diagnose of larynx diseases. The main objective of this work was to inspect the spectral envelope of the voice signal searching for information that allows voice pathology detection. The frequency and bandwidth of the first peak from the spectral envelope obtained from Linear Predictive Coefficients (LPC) of 30th order was found to be a valuable feature being used for voice pathology detection. In the corpus considered in this work we obtained a 100% discrimination between healthy and unhealthy subjects and a 87% discrimination between nodules and Reinke's edema.

KW - voice pathologies

KW - LPC spectrum

KW - nodule

KW - Reinkes's edema

KW - voice pathologies

KW - nodule

KW - Reinkes's edema

KW - LPC spectrum

U2 - 10.1016/j.protcy.2013.12.123

DO - 10.1016/j.protcy.2013.12.123

M3 - Article

SN - 2212-0173

VL - 9

SP - 1104

EP - 1111

JO - Procedia Technology

JF - Procedia Technology

IS - NA

ER -

LPC Spectrum First Peak Analysis for Voice Pathology Detection

Abstract

Keywords

Access to Document

Fingerprint

Cite this