Neural network pattern recognition of lingual-palatal pressure for automated detection of swallow.

We describe a novel device and method for real-time measurement of lingual-palatal pressure and automatic identification of the oral transfer phase of deglutition. Clinical measurement of the oral transport phase of swallowing is a complicated process requiring either placement of obstructive sensors or sitting within a fluoroscope or articulograph for recording. Existing detection algorithms distinguish oral events with EMG, sound, and pressure signals from the head and neck, but are imprecise and frequently result in false detection. We placed seven pressure sensors on a molded mouthpiece fitting over the upper teeth and hard palate and recorded pressure during a variety of swallow and non-swallow activities. Pressure measures and swallow times from 12 healthy and 7 Parkinson's subjects provided training data for a time-delay artificial neural network to categorize the recordings as swallow or non-swallow events. User-specific neural networks properly categorized 96 % of swallow and non-swallow events, while a generalized population-trained network was able to properly categorize 93 % of swallow and non-swallow events across all recordings. Lingual-palatal pressure signals are sufficient to selectively and specifically recognize the initiation of swallowing in healthy and dysphagic patients.

Targeted transtracheal stimulation for vocal fold closure.

Paralysis of the structures in the head and neck due to stroke or other neurological disorder often causes dysphagia (difficulty in swallowing). Patients with dysphagia have a significantly higher incidence of aspiration pneumonia and death. The recurrent laryngeal nerve (RLN), which innervates the intrinsic laryngeal muscles that control the vocal folds, travels superiorly in parallel to the trachea in the tracheoesophageal groove. This study tests the hypothesis that functional electrical stimulation (FES) applied via transtracheal electrodes can produce controlled vocal fold adduction. Bipolar electrodes were placed at 15° intervals around the interior mucosal surface of the canine trachea, and current was applied to the tissue while electromyography (EMG) from the intrinsic laryngeal muscles and vocal fold movement visualization via laryngoscopy were recorded. The lowest EMG thresholds were found at an average location of 100° to the left of the ventral midsagittal line and 128° to the right. A rotatable pair of bipolar electrodes spaced 230° apart were able to stimulate bilaterally both RLNs in every subject. Laryngoscopy showed complete glottal closure with transtracheal stimulation in six of the eight subjects, and this closure was maintained under simultaneous FES-induced laryngeal elevation. Transtracheal stimulation is an effective tool for minimally invasive application of FES to induce vocal fold adduction, providing an alternative mechanism to study airway protection.

Selective intraoperative stimulation of the human larynx.

OBJECTIVES/HYPOTHESIS: Laryngeal contraction normally entails activation of mutually cooperative intrinsic laryngeal muscles (ILMs). Unfortunately, standard stimulating methods do not completely mimic the normal ongoing synchrony between the muscles. We submit that this problem can be addressed by modifying the stimulating waveform. STUDY DESIGN: This study extends prior canine research to the human using quasitrapezoidal (QT) currents instead of standard square waves, which while valuable incompletely express normal ILM interactions. METHODS: In two patients undergoing laryngectomy for cancer, the recurrent laryngeal nerve on the uninvolved side received QT pulses (4 Hz, 60-2,000 µA, 100-500 µs width, 0-500-µs decay) via a bipolar cuff electrode. Pairs of needle electrodes placed into the posterior cricoarytenoideus (PCA), lateral cricoarytenoideus (LCA), and thyroarytenoideus (TA) were used to record electromyography amplitudes, and waveforms were analyzed by a specially designed computer program. RESULTS: With activity from square waves serving as control, we observed statistically significant (P < .05) shifts in mutual relationships among PCA, LCA, and TA for an array of specific QT configurations. CONCLUSIONS: Our preliminary data on selective ILM manipulation offer promise for improved dynamic control of faulty laryngeal contraction patterns.