A novel automated detection system for swallowing sounds during eating and speech under everyday conditions

Abstract

The wave analysis of swallowing sounds has been receiving attention because the recording process is easy and non-invasive. However, up until now, an expert has been needed to visually examine the entire recorded wave to distinguish swallowing from other sounds. The purpose of this study was to establish a methodology to automatically distinguish the sound of swallowing from sound data recorded during a meal in the presence of everyday ambient sound. Seven healthy participants (mean age: 26·7 ± 1·3 years) participated in this study. A laryngeal microphone and a condenser microphone attached to the nostril were used for simultaneous recording. Recoding took place while participants were taking a meal and talking with a conversational partner. Participants were instructed to step on a foot pedal trigger switch when they swallowed, representing self-enumeration of swallowing, and also to achieve six additional noise-making tasks during the meal in a randomised manner. The automated analysis system correctly detected 342 out of the 352 self-enumerated swallowing events (sensitivity: 97·2%) and 479 out of the 503 semblable wave periods of swallowing (specificity: 95·2%). In this study, the automated detection system for swallowing sounds using a nostril microphone was able to detect the swallowing event with high sensitivity and specificity even under the conditions of daily life, thus showing potential utility in the diagnosis or screening of dysphagic patients in future studies.

High-Resolution Cervical Auscultation Signal Features Reflect Vertical and Horizontal Displacements of the Hyoid Bone During Swallowing

“…Less invasive methods to assess specific aspects of dysphagia have been investigated in previous studies such as ultrasonography to assess soft tissue function [30], manometry to assess pressure generation and propagation from the oral cavity through the esophagus [31], [32], and electromyography to assess the timing and sequence of muscle activation [33]. Cervical auscultation (CA) which uses stethoscopes to observe sounds emanating from the pharyngeal mechanism to infer about swallow physiology [16], [34]–[41], enables an examiner to “hear”a swallow, however it has been shown repeatedly to provide inconsistent information leading to subjective interpretations by the examiner due limitations of the human auditory system and to variations in instrument design since stethoscopes are not designed to transmit pharyngeal sounds [42], [43]. Despite its poor precision in clinical practice, CA has intrigued investigators including ourselves with interests in signal processing, who hypothesize that the acoustic-vibratory information arising from oropharyngeal physiology contains useful diagnostic clues as to the nature of dysphagia.…”

A Comprehensive Assessment Protocol for Swallowing (CAPS): Paving the Way towards Computer-Aided Dysphagia Screening

“…Dry (saliva) swallowing identification featured in the majority of wearable systems that assessed swallowing [ 19 ]. For non-swallowing, Skowronski et al [ 30 ] proposed a non-swallowing protocol with tasks including yawning, sniffing, tongue movement, humming, throat clearing, coughing, and speech, while Fukuike et al [ 22 ] added gargling, sighing, and sipping tea. Talking, reading, and speech have not been standardized in previous studies, though they might be examined by machine learning.…”

“…Moreover, the protocol design is devoted to a computer-aided dysphagia screening tool using wearable technology and machine learning techniques or other classifier algorithms. In such a case, the signals of sensors corresponding to each swallowing maneuver can either be manually labelled by the professional watching the video recording, or be labelled by asking the participant to press a button or pedal during the swallowing episode [ 21 , 22 ]. The tasks during the assessment phase can be repeated according to the need for data augmentation.…”

Validation of the optimal site in the neck region for detecting swallowing sounds

“…Therefore, swallowing sound recorded at site 11 could be used in future studies to provide functional information concerning the rhinopharynx. In previous studies, a ready‐made laryngeal microphone was used to detect swallowing sound . This microphone was attached to the position over the lateral border of the trachea immediately inferior to the cricoid cartilage, as recommended by Takahashi et al .…”