A robust tool to compare pre- and post-surgical voice quality.

Claudia Manfredi, Riccardo Canalicchio, Giulio Cecconi, Giovanna Cantarella

Research output: Contribution to journalArticlepeer-review


Assessing voice quality by means of objective parameters is of great relevance for clinicians. A large number of indexes have been proposed in literature and in commercially available software tools. However, clinicians commonly resort to a small subset of such indexes, due to difficulties in managing set up options and understanding their meaning. In this paper, the analysis has been limited to few but effective indexes, devoting great effort to their robust and automatic evaluation. Specifically, fundamental frequency (F0), along with its irregularity (Jitter (J) and Relative Average Perturbation (RAP)), noise and formant frequencies, are tracked on voiced parts of the signal only. Mean and std values are also displayed. The underlying high-resolution estimation procedure is further strengthened by an adaptive estimation of the optimal length of signal frames for analysis, linked to varying signal characteristics. Moreover, the new tool allows for automatic analysis of any kind of signal, both as far as F0 range and sampling frequency are concerned, no manual setting being required to the user. This makes the tool feasible for application by non-expert users, also thanks to its simple interface. The proposed approach is applied here to patients suffering from cysts and polyps that underwent micro-laryngoscopic direct exeresis (MLSD).

ASJC Scopus subject areas

  • Computer Vision and Pattern Recognition
  • Signal Processing
  • Biomedical Engineering
  • Health Informatics


Dive into the research topics of 'A robust tool to compare pre- and post-surgical voice quality.'. Together they form a unique fingerprint.

Cite this