Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2021 Jun 8;23(6):e25247.
doi: 10.2196/25247.

Deep Learning Application for Vocal Fold Disease Prediction Through Voice Recognition: Preliminary Development Study

Hao-Chun Hu  1   2   3 , Shyue-Yih Chang  4 , Chuen-Heng Wang  5 , Kai-Jun Li  2 , Hsiao-Yun Cho  2   6 , Yi-Ting Chen  5 , Chang-Jung Lu  4 , Tzu-Pei Tsai  4 , Oscar Kuang-Sheng Lee  1   7   8   9
Affiliations

Deep Learning Application for Vocal Fold Disease Prediction Through Voice Recognition: Preliminary Development Study

Hao-Chun Hu et al. J Med Internet Res. .

Abstract

Background: Dysphonia influences the quality of life by interfering with communication. However, a laryngoscopic examination is expensive and not readily accessible in primary care units. Experienced laryngologists are required to achieve an accurate diagnosis.

Objective: This study sought to detect various vocal fold diseases through pathological voice recognition using artificial intelligence.

Methods: We collected 189 normal voice samples and 552 samples of individuals with voice disorders, including vocal atrophy (n=224), unilateral vocal paralysis (n=50), organic vocal fold lesions (n=248), and adductor spasmodic dysphonia (n=30). The 741 samples were divided into 2 sets: 593 samples as the training set and 148 samples as the testing set. A convolutional neural network approach was applied to train the model, and findings were compared with those of human specialists.

Results: The convolutional neural network model achieved a sensitivity of 0.66, a specificity of 0.91, and an overall accuracy of 66.9% for distinguishing normal voice, vocal atrophy, unilateral vocal paralysis, organic vocal fold lesions, and adductor spasmodic dysphonia. Compared with the accuracy of human specialists, the overall accuracy rates were 60.1% and 56.1% for the 2 laryngologists and 51.4% and 43.2% for the 2 general ear, nose, and throat doctors.

Conclusions: Voice alone could be used for common vocal fold disease recognition through a deep learning approach after training with our Mandarin pathological voice database. This approach involving artificial intelligence could be clinically useful for screening general vocal fold disease using the voice. The approach includes a quick survey and a general health examination. It can be applied during telemedicine in areas with primary care units lacking laryngoscopic abilities. It could support physicians when prescreening cases by allowing for invasive examinations to be performed only for cases involving problems with automatic recognition or listening and for professional analyses of other clinical examination results that reveal doubts about the presence of pathologies.

Keywords: artificial intelligence; convolutional neural network; dysphonia; pathological voice; vocal fold disease; voice pathology identification.

PubMed Disclaimer

Conflict of interest statement

Conflicts of Interest: None declared.

Figures

Figure 1
Figure 1
Illustration of the changes of the loss function value over the training and validation sets.
Figure 2
Figure 2
Confusion matrix of 2, 3, 4, and 5 classifications. AN = pathological voice; NC = normal voice; SD = adductor spasmodic dysphonia; PAATOL = unilateral vocal paralysis/vocal atrophy/organic vocal fold lesions; OL = organic vocal fold lesions; PAAT = unilateral vocal paralysis/vocal atrophy; PA = unilateral vocal paralysis; AT = vocal atrophy.
Figure 3
Figure 3
Receiver operating characteristic curves of 2, 3, 4, and 5 classifications. NC = normal voice; SD = adductor spasmodic dysphonia; PAATOL = unilateral vocal paralysis/vocal atrophy/organic vocal fold lesions; OL = organic vocal fold lesions; PAAT = unilateral vocal paralysis/vocal atrophy; PA = unilateral vocal paralysis; AT = vocal atrophy.
Figure 4
Figure 4
Confusion matrix of 5 classifications in human specialists. NC = normal voice; SD = adductor spasmodic dysphonia; OL = organic vocal fold lesions; PA = unilateral vocal paralysis; AT = vocal atrophy.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Cohen SM, Dupont WD, Courey MS. Quality-of-life impact of non-neoplastic voice disorders: a meta-analysis. Ann Otol Rhinol Laryngol. 2006 Feb;115(2):128–134. doi: 10.1177/000348940611500209. - DOI - PubMed
    1. Cohen SM, Kim J, Roy N, Asche C, Courey M. Prevalence and causes of dysphonia in a large treatment-seeking population. Laryngoscope. 2012 Feb;122(2):343–348. doi: 10.1002/lary.22426. - DOI - PubMed
    1. Davids T, Klein AM, Johns MM. Current dysphonia trends in patients over the age of 65: is vocal atrophy becoming more prevalent? Laryngoscope. 2012 Feb;122(2):332–335. doi: 10.1002/lary.22397. - DOI - PubMed
    1. Stachler RJ, Francis DO, Schwartz SR, Damask CC, Digoy GP, Krouse HJ, McCoy SJ, Ouellette DR, Patel RR, Reavis CCW, Smith LJ, Smith M, Strode SW, Woo P, Nnacheta LC. Clinical Practice Guideline: Hoarseness (Dysphonia) (Update) Otolaryngol Head Neck Surg. 2018 Mar;158(1_suppl):S1–S42. doi: 10.1177/0194599817751030. - DOI - PubMed
    1. Fang SH, Tsao Y, Hsiao MJ, Chen JY, Lai YH, Lin FC, Wang CT. Detection of Pathological Voice Using Cepstrum Vectors: A Deep Learning Approach. J Voice. 2019 Sep;33(5):634–641. doi: 10.1016/j.jvoice.2018.02.003. - DOI - PubMed

Publication types

LinkOut - more resources