Enhancing the Performance of Pathological Voice Quality Assessment System Through the Attention-Mechanism Based Neural Network.

OBJECTIVE: Doctors, nowadays, primarily use auditory-perceptual evaluation, such as the grade, roughness, breathiness, asthenia, and strain scale, to evaluate voice quality and determine the treatment. However, the results predicted by individual physicians often differ, because of subjective perceptions, and diagnosis time interval, if the patient's symptoms are hard to judge. Therefore, an accurate computerized pathological voice quality assessment system will improve the quality of assessment. METHOD: This study proposes a self_attention-based system, with a deep learning technology, named self_attention-based bidirectional long-short term memory (SA BiLSTM). Different pitches [low, normal, high], and vowels [/a/, /i/, /u/], were added into the proposed model, to make it learn how professional doctors evaluate the grade, roughness, breathiness, asthenia, and strain scale, in a high dimension view. RESULTS: The experimental results showed that the proposed system provided higher performance than the baseline system. More specifically, the macro average of the F1 score, presented as decimal, was used to compare the accuracy of classification. The (G, R, and B) of the proposed system were (0.768±0.011, 0.820±0.009, and 0.815±0.009), which is higher than the baseline systems: deep neural network (0.395±0.010, 0.312±0.019, 0.321±0.014) and convolution neural network (0.421±0.052, 0.306±0.043, 0.3250±0.032) respectively. CONCLUSIONS: The proposed system, with SA BiLSTM, pitches, and vowels, provides a more accurate way to evaluate the voice. This will be helpful for clinical voice evaluations and will improve patients' benefits from voice therapy.

Incorporation of shear wave elastography into a prediction model in the assessment of cervical lymph nodes.

RATIONALE AND OBJECTIVES: To assess the performance of shear wave elastography (SWE) and an extended model in predicting malignant cervical lymph nodes (LNs). MATERIALS AND METHODS: 109 patients who underwent ultrasound (US) and SWE before needle biopsy were enrolled. The optimal cutoff value of elasticity indices (EIs) was determined by receiver operating characteristic (ROC) curves. The c-statistic, net reclassification improvement (NRI) and integrated discrimination improvement (IDI) were used to compare extended model and traditional one. RESULTS: Malignant LNs had higher EIs than benign nodes (p < 0.001). The optimal cutoff point was 42 kilopascal, corresponding to 83.3% sensitivity, 64.7% specificity, and 68.8% overall accuracy. A multivariable logistic regression analysis confirmed that EI was an independent predictor for malignancy. The new extended prediction model had a positive NRI (0.96) and IDI (0.10) for predicting malignant neck LNs. Nevertheless, the c-statistic was not significantly different between the two models. CONCLUSION: The parameter of SWE theoretically improve the model performance. However, its real clinical impact is minor, as the parameters of US-based model is already very robust. SWE can be considered as an adjunctive quantitative tool beyond conventional US examination.

Assessment of Patient Outcomes After Adjuvant Vocal Fold Steroid Injection for Fibrosis After Microlaryngeal Surgery.

IMPORTANCE: Microlaryngeal surgery (MLS) is one of the most common procedures in laryngology; however, in cases of postoperative fibrosis, few remedial options are available. OBJECTIVE: To investigate the outcomes of vocal fold steroid injection (VFSI) as an adjuvant treatment for fibrosis after MLS. DESIGN, SETTING, AND PARTICIPANTS: Case series of 228 consecutive patients who underwent MLS for benign vocal lesions between January 2014 and December 2016 at a tertiary medical center. Adjuvant VFSI was performed in 25 patients (11%) with postoperative fibrosis, defined as severely decreased or absent mucosal wave on videolaryngostroboscopy (VLS) lasting for longer than 1 month postoperatively. Data were analyzed from July 1, 2018 to April 30, 2019. INTERVENTION: All adjuvant VFSIs were performed in the office under local anesthesia and were repeated monthly if the status of the mucosal wave did not return to normal or mildly decreased on the follow-up VLS. MAIN OUTCOMES AND MEASURES: The outcome parameters, namely, the 10-item Voice Handicap Index (VHI-10), smoothed cepstral peak prominence (CPP) of recorded voice samples, and mucosal wave status on VLS, were measured before and after MLS and after adjuvant VFSI. RESULTS: In this case series of 228 patients (72 men and 156 women; mean [SD] age, 42.8 [12.3] years), 25 had a decreased or absent mucosal wave after MLS, indicating postoperative fibrosis. Of these patients, 22 presented with deep-seated lesions that were removed using a microflap approach. After 1 course of adjuvant VFSI in 17 patients and 2 serial monthly injections in 8 patients, VFSI was associated with improved VHI-10 scores (mean, 24.4 points preoperatively, 22.1 points postoperatively, and 12.9 points after VFSI) that were comparable to the scores of the other 203 patients without vocal fold fibrosis after MLS (mean, 24.8 points preoperatively, 11.7 points postoperatively). Adjuvant VFSI was also associated with improved CPP (mean, 4.04 preoperatively, 4.23 postoperatively, and 5.11 after VFSI) that was comparable to the CPP of the other patients without postoperative fibrosis (mean, 4.33 preoperatively and 5.15 postoperatively). Overall, 19 (76%) of 25 patients with vocal fold fibrosis after MLS had normal (n = 11) or mildly decreased (n = 8) mucosal wave after adjuvant VFSI. CONCLUSIONS AND RELEVANCE: Adjuvant VFSI appeared to be associated with improvements in mucosal wave, voice quality, and subjective perception of dysphonia in patients with fibrosis after MLS, with outcomes comparable to those of the other patients without fibrosis. Adjuvant VFSI may be considered a rescue technique, reducing the rate of fibrosis after MLS from 11% to 3%.

Assessment of D-methionine protecting cisplatin-induced otolith toxicity by vestibular-evoked myogenic potential tests, ATPase activities and oxidative state in guinea pigs.

To date, inadequate study has been devoted to the toxic vestibular effects caused by cisplatin. In addition, no electrophysiological examination has been conducted to assess cisplatin-induced otolith toxicity. The purposes of this study are thus two-fold: 1) to determine whether cervical vestibular-evoked myogenic potentials (VEMPs) and ocular VEMPs are practical electrophysiological methods of testing for cisplatin-induced otolith toxicity and 2) to examine if D-methionine (D-met) pre-injection would protect the otolith organs against cisplatin-induced changes in enzyme activities and/or oxidative status. Guinea pigs were intraperitoneally treated once daily with the following injections for seven consecutive days: sterile 0.9% saline control, cisplatin (5 mg/kg) only, D-met (300 mg/kg) only, or a combination of d-met (300 mg/kg) and cisplatin (5 mg/kg), respectively, with a 30 minute window in between. Each animal underwent the oVEMP and cVEMP tests before and after treatment. The changes in the biochemistry of the otolith organs, including membranous Na(+), K(+)-ATPase and Ca(2+)-ATPase, lipid peroxidation (LPO) levels and nitric oxide (NO) levels, were also evaluated. In the cisplatin-only treated guinea pigs, the mean amplitudes of the oVEMP tests were significantly (p<0.05) decreased when compared to the other three groups. In guinea pigs receiving both D-met and cisplatin, the amplitudes of their oVEMP tests were significantly larger (p<0.05) than those of the cisplatin-only group, but smaller (p<0.05) than those of the saline control or D-met-only group. However, no significant difference of the amplitudes of cVEMP tests was noted among the four groups. In comparison with the other three groups, the cisplatin-only group had the lowest (ps<0.05) mean Na(+), K(+)-ATPase and Ca(2+)-ATPase, and the highest (ps<0.05) LPO and NO levels. The oVEMP tests were feasible for the evaluation of cisplatin-related otolith dysfunction. D-Met attenuated the reduced ATPase activities and increased oxidative stress induced by cisplatin toxicity in the otolith organs.

Real-time ultrasound elastography: an assessment of enlarged cervical lymph nodes.

OBJECTIVES: To determine the efficacy of real-time elastography (RTE), compared with our previously proposed prediction model, in the detection of malignancy in cervical lymph nodes (LNs). METHODS: One hundred and thirty-one patients underwent ultrasound-guided fine needle aspiration biopsy (ultrasound FNAB) after ultrasound and RTE evaluation. The formula of the RTE scoring system was a four-point visual scale, based on a previously determined model. The formula of the prediction model was: [Formula: see text]. An extended model was constructed with four previous predictors and elasticity scores, using a logistic regression model. RESULTS: Final histology revealed 77 benign and 54 malignant LNs. In the elasticity score system, sensitivity was 66.7 %, specificity was 57.1 %, the positive predictive value (PPV) was 52.2 % and the negative predictive value (NPV) was 71.0 %. In the prediction model system, sensitivity was 79.6 %, specificity was 92.2 %, the PPV was 87.8 % and the NPV was 86.6 %. When the extended and the original model were compared, the areas under the receiver operating characteristic curve (c-statistic) was 0.94 and 0.95, respectively (P > 0.05). CONCLUSIONS: Qualitative RTE offers no additional value over conventional ultrasound in predicting malignancy in cervical LNs. KEY POINTS: • An ultrasound system can help in the assessment of cervical lymph nodes. • Grey-scale and power Doppler ultrasound remain fundamental for neck nodal evaluation. • Qualitative real-time elastography provided no additional value compared with current prediction models.