Artificial neural network classification of pharyngeal high‐resolution manometry with impedance data

Abstract

Purpose To use classification algorithms to classify swallows as safe, penetration, or aspiration based on measurements obtained from pharyngeal high-resolution manometry (HRM) with impedance. Study design Case series evaluating new method of data analysis. Method Multilayer perceptron (MLP), an artificial neural network (ANN), was evaluated for its ability to classify swallows as safe, penetration, or aspiration. Data were collected from 25 disordered subjects swallowing 5 or 10 ml boluses. Following extraction of relevant parameters, a subset of the data was used to train the models and the remaining swallows were then independently classified by the ANN. Results A classification accuracy of 89.4±2.4% was achieved when including all parameters. Including only manometry-related parameters yielded a classification accuracy of 85.0±6.0% while including only impedance-related parameters yielded a classification accuracy of 76.0±4.9%. Receiver operating characteristic (ROC) analysis yielded areas under the curve (AUC) of 0.8912 for safe, 0.8187 for aspiration, and 0.8014 for penetration. Conclusions Classification models show high accuracy in classifying swallows from dysphagic patients as safe or unsafe. HRM-impedance with ANN represents one method which could be used clinically to screen for patients at risk for penetration or aspiration.

Multimodal Swallowing Evaluation with High-Resolution Manometry Reveals Subtle Swallowing Changes in Early and Mid-Stage Parkinson Disease

“…The output of HRM is objective pressure data with high temporal and spatial resolution. HRM has the capability to determine objective changes in swallowing pressure in patients with dysphagia [18–20]. HRM data is objective with good temporal and spatial resolution and thus can reveal subtle changes in swallowing-related pressure and timing events in patients in the early stages of PD, before signs or symptoms of overt dysphagia appear.…”

Implementation of High-resolution Manometry in the Clinical Practice of Speech Language Pathology

“…Automated analysis holds the promise of standardizing measures, with algorithms that can maximize use of spatiotemporal data in representing swallowing physiology [21, 30, 31]. The value of impedance data in the pharynx may be greater when combined with manometric measures in composite analysis [29, 32]. Ongoing research is needed to provide the evidence base that can define the role of HRM in swallowing care delivery.…”

Classification of High‐Resolution Manometry Data According to Videofluoroscopic Parameters Using Pattern Recognition

“…Secondly, the classification accuracy of 84.2% observed for penetration-aspiration was based on a binary classification scheme (normal versus penetration+aspiration) and was lower than our previously reported value of 89.4% obtained using a ternary scheme (normal versus penetration versus aspiration). 13 This disparity may be due to the absence of impedance measurement, which is useful for detecting bolus residue. 6 Future studies using combined HRM-impedance to identify abnormalities observed on videofluoroscopy may produce even stronger correlations than those observed here.…”

“…10,11 We previously applied ANNs to classify HRM data into normal and disordered swallows 12 as well as identify penetration and aspiration. 13 While application to identify clinically significant events currently assessed on videofluoroscopy such as aspiration is an important step in the development of HRM, videofluoroscopy can assess much more than aspiration status.…”