Jennifer L. Spielman

Suitability of Dysphonia Measurements for Telemonitoring of Parkinson's Disease

In this paper, we present an assessment of the practical value of existing traditional and nonstandard measures for discriminating healthy people from people with Parkinsons disease (PD) by detecting dysphonia. We introduce a new measure of dysphonia, pitch period entropy (PPE), which is robust to many uncontrollable confounding effects including noisy acoustic environments and normal, healthy variations in voice frequency. We collected sustained phonations from 31 people, 23 with PD. We then selected ten highly uncorrelated measures, and an exhaustive search of all possible combinations of these measures finds four that in combination lead to overall correct classification performance of 91.4%, using a kernel support vector machine. In conclusion, we find that nonstandard methods in combination with traditional harmonics-to-noise ratios are best able to separate healthy from PD subjects. The selected nonstandard methods are robust to many uncontrollable variations in acoustic environment and individual subjects, and are thus well suited to telemonitoring applications.

Novel Speech Signal Processing Algorithms for High-Accuracy Classification of Parkinson's Disease

There has been considerable recent research into the connection between Parkinsons disease (PD) and speech impairment. Recently, a wide range of speech signal processing algorithms (dysphonia measures) aiming to predict PD symptom severity using speech signals have been introduced. In this paper, we test how accurately these novel algorithms can be used to discriminate PD subjects from healthy controls. In total, we compute 132 dysphonia measures from sustained vowels. Then, we select four parsimonious subsets of these dysphonia measures using four feature selection algorithms, and map these feature subsets to a binary classification response using two statistical classifiers: random forests and support vector machines. We use an existing database consisting of 263 samples from 43 subjects, and demonstrate that these new dysphonia measures can outperform state-of-the-art results, reaching almost 99% overall classification accuracy using only ten dysphonia features. We find that some of the recently proposed dysphonia measures complement existing algorithms in maximizing the ability of the classifiers to discriminate healthy controls from PD subjects. We see these results as an important step toward noninvasive diagnostic decision support in PD.

Testing objective measures of motor impairment in early Parkinson's disease: Feasibility study of an at-home testing device.

We tested the feasibility of a computer based at‐home testing device (AHTD) in early‐stage, unmedicated Parkinsons disease (PD) patients over 6 months. We measured compliance, technical reliability, and patient satisfaction to weekly assessments of tremor, small and large muscle bradykinesia, speech, reaction/movement times, and complex motor control. relative to the UPDRS motor score. The AHTD is a 6.5″ × 10″ computerized assessment battery. Data are stored on a USB memory stick and sent by internet to a central data repository as encrypted data packets. Although not designed or powered to measure change, the study collected data to observe patterns relative to UPDRS motor scores. Fifty‐two PD patients enrolled, and 50 completed the 6 month trial, 48 remaining without medication. Patients complied with 90.6% of weekly 30‐minute assessments, and 98.5% of data packets were successfully transmitted and decrypted. On a 100‐point scale, patient satisfaction with the program at study end was 87.2 (range: 80–100). UPDRS motor scores significantly worsened over 6 months, and trends for worsening over time occurred for alternating finger taps (P = 0.08), tremor (P = 0.06) and speech (P = 0.11). Change in tremor was a significant predictor of change in UPDRS (P = 0.047) and was detected in the first month of the study. This new computer‐based technology offers a feasible format for assessing PD‐related impairment from home. The high patient compliance and satisfaction suggest the feasibility of its incorporation into larger clinical trials, especially when travel is difficult and early changes or frequent data collection are considered important to document.

The effects of intensive voice treatment on facial expressiveness in Parkinson disease: preliminary data.

ObjectiveThe purpose of the present retrospective study was to examine the effects of intensive voice therapy on facial expression in Parkinson disease. BackgroundParkinson disease (PD) often presents with symptoms that reduce communicative effectiveness on multiple levels, including decreased vocal loudness and reduced facial mobility. Recent advances in voice treatment have provided the first short- and long-term efficacy data indicating improvements in voice and speech following intensive voice therapy (Lee Silverman Voice Treatment [LSVT]). Anecdotal reports from both clinicians and patients indicate that the LSVT also has a positive impact on facial expression. These observations suggest a need to investigate more directly the effects of voice therapy on facial movement and expressiveness in PD. MethodForty-four individuals with idiopathic PD participated in this study. Video data were taken from recordings of individuals with PD who had received either one month of phonation-based treatment (LSVT) or respiratory treatment (RT) as part of a large treatment efficacy study designed to examine the effects of different types of therapy on speech and voice in PD. Twenty-second video samples of all subjects taken before and after treatment were paired and played at random without sound to trained raters, who judged each pair of video clips for facial mobility and engagement. All recordings were made while subjects were engaged in conversational speech. ResultsInter-rater reliability was extremely high (0.90) for both the rating of facial mobility and engagement. Overall, members of the LSVT group received more ratings of increased facial mobility (P = 0.036) and engagement (P = 0.056) following treatment relative to members of the RT group. In addition, the extent of change for facial mobility after treatment was perceived as greater (P = 0.05) for the LSVT group than for the RT group. ConclusionsThese results indicate that intensive voice therapy may have a positive effect on facial expressivity in PD. Such findings lend support to contemporary theories relating multiple expressive modalities (e.g., voice, face, and gesture) and suggest that targeting voice may be an effective and efficient way to influence expressive output in general.

Intensive voice treatment (LSVT®LOUD) for Parkinson's disease following deep brain stimulation of the subthalamic nucleus.

PURPOSE Intensive voice therapy (LSVT(®)LOUD) can effectively manage voice and speech symptoms associated with idiopathic Parkinson disease (PD). This small-group study evaluated voice and speech in individuals with and without deep brain stimulation of the subthalamic nucleus (STN-DBS) before and after LSVT LOUD, to determine whether outcomes for surgical subjects were comparable to non-surgical cohorts. METHODS Eight subjects with PD (four with STN-DBS and four without) received LSVT LOUD four times a week for four weeks. Four additional subjects with PD remained untreated. Voice intensity (SPL), Vowel Articulation Index (VAI), the Voice Handicap Index (VHI), and a structured interview were evaluated before and after treatment and again six months later. RESULTS Both treated groups showed significant increases in SPL from pre to post and six-month follow up. VAI was significantly higher for the treated groups compared to the untreated subjects at follow up. Several treated individuals had significant clinical improvement in VHI scores, particularly within the LSVT-DBS group. Treated individuals reported improvements in voice and speech in structured interviews; however, answers suggest more variable long-term maintenance within the LSVT-DBS group. The untreated group exhibited no significant changes in any measure throughout the study. CONCLUSIONS Results support LSVT LOUD for treating voice and speech in individuals with PD following STN-DBS surgery. However, modifications may be required to maintain functional improvements. LEARNING OUTCOMES As a result of this activity, the participant will be able to (1) describe how deep brain stimulation of the subthalamic nucleus may affect voice and speech in Parkinson disease; (2) describe the effects of intensive voice therapy (LSVT(®)LOUD) on people with PD both with and without STN-DBS; (3) describe how individuals with STN-DBS maintained treatment effects over time.

The inability to produce soft voice (IPSV): A tool to detect vocal change in school-teachers

Purpose. This study investigated whether clinicians could detect voice changes reported by teachers, from self-ratings teachers conducted of their inability to produce soft voice (IPSV). Methods. Ten teachers wore a vocal dosimeter and completed daily IPSV ratings approximately every 2 hours for 14 days. Following the 2 weeks of dosimetry, two speech clinicians specialized in voice rated the teachers’ IPSV from dosimeter recordings. Teacher and clinician ratings were compared for each participant. Results. Although agreement between teacher and clinician ratings was not significant, descriptive analyses demonstrated an average difference score of 1.7 (SD 1.4) between teacher and clinician ratings. Conclusions. This study supports the potential usefulness of the IPSV as a simple tool to detect voice changes in oneself or others.

Acoustical Analysis of Posed Prosodic Expressions: Effects of Emotion and Sex

Prosodic expression is an important channel of emotional communication and can be assessed through computerized acoustical analysis. Fundamental Frequency (F0) is the most commonly studied acoustic measure and considered the best index of perceived pitch. In this study, we examined two factors that can influence prosodic expression, sex and emotion type, A special feature is the use of a neutral expression as a control measure. We also described procedures for eliciting posed prosodic expression via an experimental task from the New York Emotion Battery. Subjects were healthy men (n=10) and women (n=9), matched for. age (M = 29.2 yr.) and education (M= 15.6 yr). Subjects were asked to intone neutral-content sentences with happy, sad, and neutral prosody. F0 mean and standard deviation were measured using the Computerized Speech Lab program. Initial findings indicated that women produced significantly higher F0 values than did men and that happy sentences were produced with significantly higher F0 values than were sad sentences. When semitone conversions were applied and neutral prosody was subtracted out, differences remained for emotion type but not for sex. Findings are discussed in terms of implications for the assessment and treatment of prosody in clinical populations.

Speech treatment in Parkinson's disease: Randomized controlled trial (RCT): SPEECH TREATMENT IN PARKINSON'S DISEASE: RCT

As many as 89% of people with Parkinsons disease (PD) develop speech disorders.