1. Title: Laryngeal3 database Three–class problem – normal /organic pathology / functional (neurological) pathology 2. Sources: (a) Creator: Dr. St. Hadjitodorov, Dr. P. Mitev, Dr. B. Boyanov Central Lab on Biomedical Engineering Bulgarian Academy of Sciences Acad.G.Bonchev Str., block 105 1113 Sofia, Bulgaria 3. Past Usage: 1.Mitev P., S.Hadjitodorov. Fundamental frequency estimation of voice of patients with laryngeal disorders. Information Sciences Journal, Vol.156, Issues 1-2, 1 November 2003, pp. 3-19. 2.Hadjitodorov, S. P.Mitev. A computer system for acoustic analysis of pathological voices and laryngeal diseases screening. Medical Engineering and Physiscs, 2002, July, 24:6, 419-429. 3.Mitev, P. , Hadjitodorov, S. A Method for Turbulent Noise Estimation in Voiced Signals, Medical & Biological Engineering & Computing, vol.38, 2000, pp. 625-631. 4.Hadjitodorov,S., B.Boyanov, B.Teston. Laryngeal pathology detection by means of class-specific neural maps. IEEE Trans. on Information Technology in Biomedicine, vol.4, No 1, 2000, pp. 68-73. 5.Boyanov B., Doskov D., Mitev P., Hadjitodorov S., Teston B. New cepstral parameters for description of pathologic voice. Comptes Rendus de L'Academie Bulgare des Sciences (Ann. of Bulgarian Academy of Sciences), vol.53, No 3, 2000, pp.41-44. 6.Boyanov B., S. Hadjitodorov. Acoustic analysis of pathological voices and screening of laryngeal diseases. IEEE Engineering in Medicine and Biology Magazine, vol.16, No 4, 1997, pp.74-82 . 7.Boyanov B., S. Hadjitodorov, G. Baudoin, D. Doskov. Method for evaluation the energy in the singer formant. Comptes Rendus de L'Academie Bulgare des Sciences (Ann. of Bulgarian Academy of Sciences),vol. 48, No 8, 1995, pp.25-28. 4. Relevant Information: Kayelemetrics Inc. sells on CD voice signal records of patients distributed in different classes. We have processed these signals to derive parameters in time, spectral and cepstral domains. A computer decision support system was developed at CLBME (Bulgarian Academy of Sciences) to aid diagnosis of laryngeal pathology and especially in detecting its early stages. The system is currently in use at the Phoniatric department of the University Hospital "Queen Jovanna" in Sofia, Bulgaria. 5. Number of Instances: 353 6. Number of Attributes: 16 attributes (all continuous-valued) plus the class attribute 7. Attribute Information: 1. Class label - 0 – normal, 1 – pathological 2. Duration of the voice signal. 3. DUV - Degree of "unvoiceness": ratio of unvoiced part of the signal to the whole signal. 4. Fo - Fundamental frequency. 5. Standard deviation of fundamental frequency. 6. PPQ - Pitch period perturbation quotient (Jitter). 7. APQ - Amplitude perturbation quotient (Shimmer). 8. HNRf - Harmonic energy to noise energy ratio in frequency domain [dB]. 9. HNRq - Harmonic energy to noise energy ratio – Qi method [dB]. 10. NNE - Normalized noise energy [dB]. 11. DH - Degree of hoarseness: ratio of noise energy to all energy. 12. TNI - Turbulence noise index. 13. NFHE - Normalized first harmonic energy: first harmonic energy to all harmonic energy ratio. 14. HLER - High to low energy ratio: ratio of harmonics in range 0 – 500 Hz to harmonics in range 500 - 3500 Hz. 15. NVB - Number of voiceness breaks. 16. FrTI - Frequency tremors intensity. 17. AmTI - Amplitude tremor intensity. 8. Missing Attribute Values: None 9. Class Distribution: (out of 353 total instances) -- 53 Normal voices -- 218 Organic pathological voices -- 82 Functional pathological voices