Investigation of Vocal Jitter and Glottal Flow Spectrum as Possible Cues for Depression and Near-Term Suicidal Risk

Abstract

Among the many clinical decisions that psychiatrists must make, assessment of a patient's risk of committing suicide is definitely among the most important, complex, and demanding. When reviewing his clinical experience, one of the authors observed that successful predictions of suicidality were often based on the patient's voice independent of content. The voices of suicidal patients judged to be high-risk near-term exhibited unique qualities, which distinguished them from nonsuicidal patients. We investigated the discriminating power of two excitation-based speech parameters, vocal jitter and glottal flow spectrum, for distinguishing among high-risk near-term suicidal, major depressed, and nonsuicidal patients. Our sample consisted of ten high-risk near-term suicidal patients, ten major depressed patients, and ten nondepressed control subjects. As a result of two sample statistical analyses, mean vocal jitter was found to be a significant discriminator only between suicidal and nondepressed control groups (p < 0.05). The slope of the glottal flow spectrum, on the other hand, was a significant discriminator between all three groups (p < 0.05). A maximum likelihood classifier, developed by combining the a posteriori probabilities of these two features, yielded correct classification scores of 85% between near-term suicidal patients and nondepressed controls, 90% between depressed patients and nondepressed controls, and 75% between near-term suicidal patients and depressed patients. These preliminary classification results support the hypothesized link between phonation and near-term suicidal risk. However, validation of the proposed measures on a larger sample size is necessary.

Vocal biomarkers of depression based on motor incoordination

“…Classes of potential biomarkers of growing interest are based on vocal characteristics, which have been shown to change with a patient's mental condition and emotional state [3,[6][7][8][9][10][11][12][13][14]. Although there has been significant effort in using potential vocal biomarkers for emotion classification, there has been little or no exploiting of the effects of incoordination in the vocal modality (or in other modalities such as facial expression) that result from a depressed state.…”

Feature Extraction Method for Predicting Depression by Frequency Domain Analysis

“…A related study [4] reported that the slope that is obtained from the glottal flow spectrum and jitter are the best features to separate the depression from the normal group.…”

“…2) Intensity mean and range: For analyzing the glottal flow spectrum, we extracted features based on the method of glottal flow spectrum proposed in a previous study [4]. While their study measured the slope and correlation in the 300-3000Hz region, we separate the frequency regions into 300-1000Hz, 1000-2000Hz, and 2000-3000Hz, and extract the slope and correlation in each region.…”

Comparative Analysis of Vocal Characteristics in Speakers with Depression and High-Risk Suicide

“…The estimation procedure of studied features can be described in details as follows: First the whole concatenated segments of voiced speech corresponding to individual speaker were windowed into the consecutive segments with a frame length of 25.6ms, estimating the Logarithm of the Discrete Fourier Transform for all segments, calculating the energy of the log-magnitude spectrum filtered out via a 16-triangular Band-pass filterbank with center frequencies respective to the Mel-frequency scale frequency response, computing the Inverse Discrete Fourier transform (IDFT) which represent for the sixteen-order cepstral coefficients corresponding to the vocal-tract response in term of frequency characteristic response of the filter, the second major part of the speech production system. Next step is a procedure of parameter estimation to determine the GSS parameter for all voiced speech segments [6]. The complete procedure of estimation of GSS parameter is depicted in Fig.…”

“…In formerly published papers [2]- [6], the analytical techniques have been designed and developed to determine if the subjects were categorized into any of the following patient groups: Control, Non-suicidal Depressed or High-risk Suicidal. In those studies [1], [7] the vocal cues have been properly used in term of indicator as assistive tool in diagnosing the underlying symptom in patient by experienced clinicians but these skills are not widespread in clinical use.…”