A practical approach to Sasang constitutional diagnosis using vocal features

BMC Complementary and Alternative Medicine

Jong Yeol Kim¹  Keun Ho Kim¹  Jiho Nam¹  Young-Su Kim¹  Boncho Ku¹  Jun-Su Jang¹ 

[1]Medical Engineering R&D Group, Medical Research Division, Korea Institute of Oriental Medicine, 1672 Yuseongdae-ro, Yuseong-gu, Daejeon 305-811, Republic of Korea

关键词: Vocal feature;    Voice;    Diagnosis;    Sasang constitution;   

Others  :  1220808
DOI  :  10.1186/1472-6882-13-307

 received in 2013-03-11, accepted in 2013-10-30,  发布年份 2013

【 摘 要 】

Background

Sasang constitutional medicine (SCM) is a type of tailored medicine that divides human beings into four Sasang constitutional (SC) types. Diagnosis of SC types is crucial to proper treatment in SCM. Voice characteristics have been used as an essential clue for diagnosing SC types. In the past, many studies tried to extract quantitative vocal features to make diagnosis models; however, these studies were flawed by limited data collected from one or a few sites, long recording time, and low accuracy. We propose a practical diagnosis model having only a few variables, which decreases model complexity. This in turn, makes our model appropriate for clinical applications.

Methods

A total of 2,341 participants’ voice recordings were used in making a SC classification model and to test the generalization ability of the model. Although the voice data consisted of five vowels and two repeated sentences per participant, we used only the sentence part for our study. A total of 21 features were extracted, and an advanced feature selection method—the least absolute shrinkage and selection operator (LASSO)—was applied to reduce the number of variables for classifier learning. A SC classification model was developed using multinomial logistic regression via LASSO.

Results

We compared the proposed classification model to the previous study, which used both sentences and five vowels from the same patient’s group. The classification accuracies for the test set were 47.9% and 40.4% for male and female, respectively. Our result showed that the proposed method was superior to the previous study in that it required shorter voice recordings, is more applicable to practical use, and had better generalization performance.

Conclusions

We proposed a practical SC classification method and showed that our model having fewer variables outperformed the model having many variables in the generalization test. We attempted to reduce the number of variables in two ways: 1) the initial number of candidate features was decreased by considering shorter voice recording, and 2) LASSO was introduced for reducing model complexity. The proposed method is suitable for an actual clinical environment. Moreover, we expect it to yield more stable results because of the model’s simplicity.

【 授权许可】

   
2013 Jang et al.; licensee BioMed Central Ltd.

【 图 表 】

Figure 1.

Figure 2.

【 参考文献 】

• [1]Kim JY, Pham DD: Sasang constitutional medicine as a holistic tailored medicine. Evid Based Complement Alternat Med 2009, 6(1):11-19.
• [2]Song IB: An Introduction to Sasang Constitutional Medicine. Jimoondang: Seoul; 2005.
• [3]Lee J: Dongeuisusebowon-chogo. Cheongdam: Seoul; 1999.
• [4]Lee EJ, Sohn EH, Yoo JH, Kim JW, Kim KK, Kho BH, Song IB: The study of Sasangin’s face. J Sasang Constitutional Medicine 2005, 17(3):55-68.
• [5]Shin M, Kim D: A study on the correlation between sound characteristics and Sasang constitution by CSL. J Sasang Constitutional Medicine 1999, 11(1):137-157.
• [6]Yang S, Kim D: A study on the correlation between sound spectrogram and Sasang constitution. J Sasang Constitutional Medicine 1996, 8(2):191-202.
• [7]Kim D, Park S, Gun G: An objective study of Sasang constitution diagnosis by sound analysis. J Sasang Constitutional Medicine 1998, 10(1):65-80.
• [8]Park SJ, Kim DR: A study on the correlation between Sasang constitution and sound characteristics used harmonics and formant bandwidth. J Sasang Constitutional Medicine 2004, 16(1):61-73.
• [9]Kim DJ, Jung WK, Choi JW, Kim DR, Jeon JW: A study on the characteristics of the adult men sound as by Sasang constitution analyzed with PSSC-2004. J Sasang Constitutional Medicine 2005, 17(1):67-83.
• [10]Kim SH, Han DY, Youn JY, Kim DR, Jeon JW: A study on the characteristics of the Korea adult women sound as by Sasang constitution analysed with PSSC-2004. J Sasang Constitutional Medicine 2005, 17(1):84-102.
• [11]Choi JW, Song HS, Han DY, Cho SE: A study on the characteristics of the Korean adult male sound according to Sasang constitution using PSCC with a sentence. J Sasang Constitutional Medicine 2006, 18(3):64-74.
• [12]Kang JH, Yoo JH, Lee HJ, Kim JY: Automated speech analysis applied to Sasang constitution classification. Phonetics and Speech Sciences 2009, 1(3):155-161.
• [13]Kang JH, Do JH, Kim JY: Voice classification algorithm for Sasang constitution using support vector machine. J Sasang Constitutional Medicine 2010, 22(1):17-25.
• [14]Boser B, Guyon I, Vapnik V: A training algorithm for optimal margin classifiers. In Proceedings of the Fifth Annual Workshop on Computational Learning Theory: 27–29 July 1992; Pittsburgh. New York: ACM; 1992:144-152.
• [15]Do JH, Jang E, Ku B, Jang JS, Kim H, Kim JY: Development of an integrated Sasang constitution diagnosis method using face, body shape, voice, and questionnaire information. BMC Complement Altern Med 2012, 12:85. BioMed Central Full Text
• [16]Song KH, Yu SG, Cha S, Kim JY: Association of the apolipoprotein A5 gene – 1131 T > C polymorphism with serum lipids in Korean subjects: impact of Sasang constitution. Evid Based Complement Alternat Med 2012, 2012:598394.
• [17]Godino-Llorente JI, Sáenz-Lechón N, Osma-Ruiz V, Aguilera-Navarro S, Gómez-Vilda P: An integrated tool for the diagnosis of voice disorders. Med Eng Phys 2006, 28(3):276-289.
• [18]Zelcer S, Henri C, Tewfik TL, Mazer B: Multidimensional voice program analysis (MDVP) and the diagnosis of paediatric vocal cord dysfunction. Ann Allergy Asthma Immunol 2002, 88(6):601-608.
• [19]Tibshirani RJ: Regression shrinkage and selection via LASSO. J the Royal Statistical Society B 1996, 58(1):267-288.
• [20]HTK speech recognition toolkit. http://htk.eng.cam.ac.uk webcite
• [21]Proakis JG, Manolakis DG: Digital Signal Processing. New Jersey: Prentice Hall; 1996.
• [22]Mel-frequency cepstrum. http://en.wikipedia.org/wiki/Mel-frequency_cepstrum webcite
• [23]Hasan R, Jamil M, Rabbani G, Rahman S: Speaker Identification Using Mel Frequency Cepstral Coefficients. Dhaka, Bangladesh; 2004:565-568. [Proceedings of the 3rd international conference on electrical & computer engineering: 28–30 december 2004]
• [24]Brys G, Hubert M, Rousseeuw PJ: A robustification of independent component analysis. J Chemometrics 2005, 19(5–7):364-375.
• [25]Hubert M, Vandervieren E: An adjusted boxplot for skewed distribution. Computational Statistics & Data Analysis 2008, 52(12):5186-5201.
• [26]Chatterjee S, Hadi AS: Regression Analysis. New York: John Wiley and Sons; 2002.
• [27]Friedman J, Hastie T, Tibshirani RJ: Regularization paths for generalized linear models via coordinate descent. J Statistical Software 2010, 33(1):1-22.