Assessing Airflow Sensitivity to Healthy and Diseased Lung Conditions in a Computational Fluid Dynamics Model Validated In Vitro.

Computational models are useful for understanding respiratory physiology. Crucial to such models are the boundary conditions specifying the flow conditions at truncated airway branches (terminal flow rates). However, most studies make assumptions about these values, which are difficult to obtain in vivo. We developed a computational fluid dynamics (CFD) model of airflows for steady expiration to investigate how terminal flows affect airflow patterns in respiratory airways. First, we measured in vitro airflow patterns in a physical airway model, using particle image velocimetry (PIV). The measured and computed airflow patterns agreed well, validating our CFD model. Next, we used the lobar flow fractions from a healthy or chronic obstructive pulmonary disease (COPD) subject as constraints to derive different terminal flow rates (i.e., three healthy and one COPD) and computed the corresponding airflow patterns in the same geometry. To assess airflow sensitivity to the boundary conditions, we used the correlation coefficient of the shape similarity (R) and the root-mean-square of the velocity magnitude difference (Drms) between two velocity contours. Airflow patterns in the central airways were similar across healthy conditions (minimum R, 0.80) despite variations in terminal flow rates but markedly different for COPD (minimum R, 0.26; maximum Drms, ten times that of healthy cases). In contrast, those in the upper airway were similar for all cases. Our findings quantify how variability in terminal and lobar flows contributes to airflow patterns in respiratory airways. They highlight the importance of using lobar flow fractions to examine physiologically relevant airflow characteristics.

Decoding phonation with artificial intelligence (DeP AI): Proof of concept.

OBJECTIVE: Acoustic analysis of voice has the potential to expedite detection and diagnosis of voice disorders. Applying an image-based, neural-network approach to analyzing the acoustic signal may be an effective means for detecting and differentially diagnosing voice disorders. The purpose of this study is to provide a proof-of-concept that embedded data within human phonation can be accurately and efficiently decoded with deep learning neural network analysis to differentiate between normal and disordered voices. METHODS: Acoustic recordings from 10 vocally-healthy speakers, as well as 70 patients with one of seven voice disorders (n = 10 per diagnosis), were acquired from a clinical database. Acoustic signals were converted into spectrograms and used to train a convolutional neural network developed with the Keras library. The network architecture was trained separately for each of the seven diagnostic categories. Binary classification tasks (ie, to classify normal vs. disordered) were performed for each of the seven diagnostic categories. All models were validated using the 10-fold cross-validation technique. RESULTS: Binary classification averaged accuracies ranged from 58% to 90%. Models were most accurate in their classification of adductor spasmodic dysphonia, unilateral vocal fold paralysis, vocal fold polyp, polypoid corditis, and recurrent respiratory papillomatosis. Despite a small sample size, these findings are consistent with previously published data utilizing deep neural networks for classification of voice disorders. CONCLUSION: Promising preliminary results support further study of deep neural networks for clinical detection and diagnosis of human voice disorders. Current models should be optimized with a larger sample size. LEVELS OF EVIDENCE: Level III.

Body Surface Mapping of T-wave Alternans Depends on the Distribution of Myocardial Scarring.

T-Wave alternans (TWA) testing using 12-lead electrocardiogram/Frank leads is emerging as an important non-invasive biomarker to identify patients at high risk of Sudden Cardiac Death (SCD). Cardiac scarring is very common among cardiomyopathy patients; however, its influence on the body surface distribution of TWA has not yet been defined. Our objective was to perform a simulation study in order to determine whether cardiac scarring affects the distribution of TWA on thorax such that the standard leads fail to detect TWA in some of cardiomyopathy patients; thereby producing a false-negative test. Developing such a novel lead configuration could improve TWA quantification and potentially optimize electrocardiogram (ECG) lead configuration and risk stratification of SCD in cardiomyopathy patients. The simulation was performed in a 1500-node heart model using ECGSIM. TWA was mimicked by simulating action potential duration alternans in the ventricles. Cardiac scarring with different sizes were simulated by manipulating the apparent velocity, transmembrane potential and transition zone at varied locations along the left ventricular posterior wall. Our simulation study showed that the location of maximum TWA depends on the location and size of the myocardium scarring in patients with cardiomyopathy, which can give rise to false-negative TWA signal detection using standard clinical leads. The TWA amplitude generally increased with the increment of scar size (P<0.00001). We found one specific location (a non-standard lead) that consistently appeared as the top five maximum TWA leads and could be considered as an additional lead to improve the outcome of the TWA testing in cardiomyopathy patients.

Compliance and quality of life in patients on prescribed voice rest.

OBJECTIVES: To determine patient compliance with voice rest and the impact of voice rest on quality of life (QOL). STUDY DESIGN: Prospective. SETTING: University hospital. SUBJECTS AND METHODS: Demographics, self-reported compliance, QOL impact on a 100-mm visual analog scale (VAS), and communication methods were collected from 84 participants from 2 academic voice centers. RESULTS: Of 84 participants, 36.9% were men, 63.1% were women, and 64.3% were singers. The mean age of participants was 47.2 years. The mean duration of voice rest was 8.8 days (range, 3-28), and the median was 7 days. Overall compliance was 34.5%. Postoperative voice rest patients were more compliant than non-postoperative patients (42.4% vs 16.0%, P = .04, χ(2)). Voice rest had an impact on QOL (mean ± SD, 68.5 ± 27.7). Voice rest also had a greater impact on singers than nonsingers (mean VAS 77.2 vs 63.6, P = .03, t test) and on those age <60 years than those age ≥ 60 years (mean VAS 74.4 vs 46.7, P < .001, t test). More talkative patients and those with longer periods of voice rest had worse QOL scores (Spearman correlation = 0.35, P = .001 and Spearman correlation = 0.24, P = .03, respectively). Restrictions in personal and social life were noted in 36.9% of patients, 46.4% were unable to work, 44.0% felt frustrated, and 38.1% reported feeling handicapped while on voice rest. CONCLUSIONS: Given poor patient compliance and the significant impact of voice rest on QOL, further studies are warranted to examine the efficacy of voice rest and factors that may contribute to patient noncompliance with treatment.