• 제목/요약/키워드: Multi-dimensional voice program

검색결과 34건 처리시간 0.021초

개별화자의 음성파라미터 추출에 관한 연구: 음성파라미터의 상관관계를 중심으로 (A Study of Extracting Acoustic Parameters for Individual Speakers)

  • 고도흥
    • 음성과학
    • /
    • 제10권2호
    • /
    • pp.129-143
    • /
    • 2003
  • Fundamental frequency (Fo), jitter, shimmer, and harmonics-to-noise ratio (NHR) have been measured to see their interactions between the parameters using Multi-Dimensional Voice Program (MDVP). 100 Korean normal adults (50 males and 50 females) ranging from their early 20's to their early 30's produced the eight sustained vowels including /a/, /i/, /u/, /c/, /e/,/$\varepsilon$/, /i/, and /e/. The subjects were asked to read the above vowels five times in isolation with the interval of five seconds, respectively. Male voices, on the average, showed 130.7 Hz in Fo, 0.6696% in jitter, 1.8151% in shimmer, and 0.12 in NHR, while female voices showed 232.8 Hz in Fo, 0.9222% in jitter, 1.9199% in shimmer, and 0.1098 in NHR. As to the correlation coefficient, it was found that for male speakers jitter vs. shimmer, shimmer vs. NHR, Fo vs. shimmer, and Fo vs. NHR are statistically significant. It was found that for female subjects jitter vs. shimmer and Fo vs. shimmer are statistically significant. However, it is concluded that the correlation coefficient in females are not meaningful in a practical way though they are all statistically significant.

  • PDF

후두 미세수술 중 병변 내 스테로이드 주입이 음성에 미치는 효과 분석 (Analysis of the Effect of Intralesional Steroid Injection on the Voice During Laryngeal Microsurgery)

  • 박재선;강현석;이인범;진성민;이상혁
    • 대한후두음성언어의학회지
    • /
    • 제33권3호
    • /
    • pp.166-171
    • /
    • 2022
  • Background and Objectives Vocal fold (VF) scar is known to be the most common cause of dysphonia after laryngeal microsurgery (LMS). Steroids reduce postoperative scar formation by inhibiting inflammation and collagen deposition. However, the clinical evidence of whether steroids are helpful in reducing VF scar formation after LMS is still lacking. The purpose of this study is to determine whether intralesional VF steroid injection after LMS helps to reduce postoperative scar formation and voice quality. Materials and Method This study was conducted on 80 patients who underwent LMS for VF polyp, Reinke's edema, and leukoplakia. Among them, 40 patients who underwent VF steroid injection after LMS were set as the injection group, and patients who had similar sex, age, and lesion size and who underwent LMS alone were set as the control group. In each group, stroboscopy, multi-dimensional voice program, Aerophone II, and voice handicap index (VHI) were performed before and 1 month after surgery, and the results were statistically analyzed. Results There were no statistically significant differences in the distribution of sex, age, symptom duration, occupation and smoking status between each group. Both groups consisted of VF polyp (n=21), Reinke's edema (n=11), and leukoplakia (n=9). On stroboscopy, the lesion disappeared after surgery, and the amplitude and mucosal wave were symmetrical on both sides of the VFs in all patients. Acoustic parameters and VHI significantly improved after surgery in all patients. However, there was no significant difference between the injection and control group in most of the results. Conclusion There was no significant difference in the results of stroboscopy, acoustic, aerodynamic, and subjective evaluation before and after surgery in the injection group and the control group.

라인케씨 부종 환자에서 경윤상 갑상막 접근을 통한 성대 내 스테로이드 주입술의 효용 (The Efficacy of Percutaneous Steroid Injection via Cricothyroid Membrane for Reinke's Edema)

  • 남우주;김선우;진성민;이상혁
    • 대한후두음성언어의학회지
    • /
    • 제30권2호
    • /
    • pp.101-106
    • /
    • 2019
  • Background and Objectives Reinke's edema is a benign vocal fold disease caused by an edematous laryngeal superficial layer of lamina propria. The first line treatment is cessation of smoking and laryngeal microsurgery. The aim of the study is to evaluate the feasibility and efficacy of percutaneous steroid injection via cricothyroid membrane in patients with Reinke's edema. Materials and Method From Jan 2010 to July 2018, 33 Patients with Reinke's edema managed by vocal fold steroid injection via the cricothyroid membrane were included in this study. We compared medical records of laryngoscopy, stroboscopy and Multi-Dimensional Voice Program analysis at pre-treatment and post-treatment. Subjective voice improvement was evaluated using Voice Handicap Index-30 (VHI-30). Results 75.7% of the patients showed partial response and 6.06% showed complete response. 93.94% were present smokers and only 4 patients ceased smoking after the treatment. In acoustic analysis, the pre-treatment mean value of jitter, shimmer, and noise to harmonic ratio was 2.30±3.21, 9.34±10.37, 1.11±2.90 each. The post-treatment value was 2.20±1.89, 6.96±5.30, 0.20±0.09 respectively and none of the parameters were statistically significant. For subjective symptom improvement, 25 (75.8%) patients showed a better score on post-treatment VHI-30 compared to pre-treatment. Conclusion According to our study, steroid injection is a relatively safe and effective procedure for patients with Reinke's edema. A vocal fold steroid injection via the cricothyroid membrane can be an alternative treatment option for those who are not able to undergo conventional laryngeal microscopic surgery, however cessation of smoking is necessary for effective treatment.

이진 분류문제에서의 딥러닝 알고리즘의 활용 가능성 평가 (Feasibility of Deep Learning Algorithms for Binary Classification Problems)

  • 김기태;이보미;김종우
    • 지능정보연구
    • /
    • 제23권1호
    • /
    • pp.95-108
    • /
    • 2017
  • 최근 알파고의 등장으로 딥러닝 기술에 대한 관심이 고조되고 있다. 딥러닝은 향후 미래의 핵심 기술이 되어 일상생활의 많은 부분을 개선할 것이라는 기대를 받고 있지만, 주요한 성과들이 이미지 인식과 자연어처리 등에 국한되어 있고 전통적인 비즈니스 애널리틱스 문제에의 활용은 미비한 실정이다. 실제로 딥러닝 기술은 Convolutional Neural Network(CNN), Recurrent Neural Network(RNN), Deep Boltzmann Machine (DBM) 등 알고리즘들의 선택, Dropout 기법의 활용여부, 활성 함수의 선정 등 다양한 네트워크 설계 이슈들을 가지고 있다. 따라서 비즈니스 문제에서의 딥러닝 알고리즘 활용은 아직 탐구가 필요한 영역으로 남아있으며, 특히 딥러닝을 현실에 적용했을 때 발생할 수 있는 여러 가지 문제들은 미지수이다. 이에 따라 본 연구에서는 다이렉트 마케팅 응답모델, 고객이탈분석, 대출 위험 분석 등의 주요한 분류 문제인 이진분류에 딥러닝을 적용할 수 있을 것인지 그 가능성을 실험을 통해 확인하였다. 실험에는 어느 포르투갈 은행의 텔레마케팅 응답여부에 대한 데이터 집합을 사용하였으며, 전통적인 인공신경망인 Multi-Layer Perceptron, 딥러닝 알고리즘인 CNN과 RNN을 변형한 Long Short-Term Memory, 딥러닝 모형에 많이 활용되는 Dropout 기법 등을 이진 분류 문제에 활용했을 때의 성능을 비교하였다. 실험을 수행한 결과 CNN 알고리즘은 비즈니스 데이터의 이진분류 문제에서도 MLP 모형에 비해 향상된 성능을 보였다. 또한 MLP와 CNN 모두 Dropout을 적용한 모형이 적용하지 않은 모형보다 더 좋은 분류 성능을 보여줌에 따라, Dropout을 적용한 CNN 알고리즘이 이진분류 문제에도 활용될 수 있는 가능성을 확인하였다.