Fig 1.
Parallel recording of acoustic and HSV data with subsequent extraction of signals.
Table 1.
Number of combined HSV-acoustic datasets and subject range of age for each group (healthy females (NF), females with FD (FDF), healthy males (NM) and males with FD (FDM)).
Fig 2.
Parameter calculation in four steps.
(A) Segmentation of the glottal area between the vocal folds and subdivision in left and right half. (B) Extraction of GAWR (blue), GAWL (red) and GAWT (black) and synchronous audio recording. (C) Detection of maximum based cycles in all signals (all GAWs use cycles based on GAWT). (D) Calculation of 14 acoustic parameters, 25 GAW parameters and 10 symmetry, i.e. GAWL and GAWR, based parameters.
Table 2.
Summarized parameter information.
Fig 3.
Calculation of Pearson Correlation Coefficients (PCC) and Distance Correlation Coefficients (DCC) between parameters forming Pearson / distance correlation matrices.
False discovery rate (FDR) is set to 5% for males and females independently; i.e. FDR is set to 5% for combined subject groups NF & FDF and NM & FDM.
Fig 4.
Distribution of parameter values for (A) acoustic based CPP in females and (B) GAW-based PQ[Std] in males.
In the violet sections of the histograms blue and red histograms are overlapping.
Fig 5.
Correlation between GAW and acoustic F0 [Mean] in (A) females and (B) males with fitted line (black).
Table 3.
Parameters correlated with GAW-based or acoustic based F0 [Mean].
Fig 6.
Relevant parameter relations with highest difference between PCC and DCC in (A) females (acoustic based F0 [Mean] versus GAW-based PhAI [Std]) and (B) males (acoustic based WMCMean versus GAW-based F0 [Mean]).
Fitted are the linear regression line (black, dashed), a second degree polynomial (continuous, black) and a RANSAC regression line (green).
Table 4.
GAW-based parameters correlated with the given acoustic based parameter (NNE, CPP, WMCmax, WMCmean AVI and MJit).