1
|
Anwar A, Khalifa Y, Lucatorto E, Coyle JL, Sejdic E. Towards a comprehensive bedside swallow screening protocol using cross-domain transformation and high-resolution cervical auscultation. Artif Intell Med 2024; 154:102921. [PMID: 38991399 DOI: 10.1016/j.artmed.2024.102921] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 01/24/2024] [Revised: 06/17/2024] [Accepted: 06/20/2024] [Indexed: 07/13/2024]
Abstract
High-resolution cervical auscultation (HRCA) is an emerging noninvasive and accessible option to assess swallowing by relying upon accelerometry and sound sensors. HRCA has shown tremendous promise and accuracy in identifying and predicting swallowing physiology and biomechanics with accuracies equivalent to trained human judges. These insights have historically been available only through instrumental swallowing evaluation methods, such as videofluoroscopy and endoscopy. HRCA uses supervised learning techniques to interpret swallowing physiology from the acquired signals, which are collected during radiographic assessment of swallowing using barium contrast. Conversely, bedside swallowing screening is typically conducted in non-radiographic settings using only water. This poses a challenge to translating and generalizing HRCA algorithms to bedside screening due to the rheological differences between barium and water. To address this gap, we proposed a cross-domain transformation framework that uses cycle generative adversarial networks to convert HRCA signals of water swallows into a domain compatible with the barium swallows-trained HRCA algorithms. The proposed framework achieved a cross-domain transformation accuracy that surpassed 90%. The authenticity of the generated signals was confirmed using a binary classifier to confirm the framework's capability to produce indistinguishable signals. This framework was also assessed for retaining swallow physiological and biomechanical properties in the signals by applying an existing model from the literature that identifies the opening and closure of the upper esophageal sphincter. The outcomes of this model showed nearly identical results between the generated and original signals. These findings suggest that the proposed transformation framework is a feasible avenue to advance HCRA towards clinical deployment for water-based swallowing screenings.
Collapse
Affiliation(s)
- Ayman Anwar
- Edward S. Rogers Department of Electrical and Computer Engineering, University of Toronto, Toronto, ON, Canada.
| | - Yassin Khalifa
- Center for Research Computing, University of Pittsburgh, Pittsburgh, PA, USA; Information Technology Analytics, University of Pittsburgh, Pittsburgh, PA, USA; Systems and Biomedical Engineering, Cairo University, Giza, Egypt.
| | - Erin Lucatorto
- Department of Communication Science and Disorders, School of Health and Rehabilitation Sciences, University of Pittsburgh, Pittsburgh, PA, USA.
| | - James L Coyle
- Department of Communication Science and Disorders, School of Health and Rehabilitation Sciences, University of Pittsburgh, Pittsburgh, PA, USA.
| | - Ervin Sejdic
- Edward S. Rogers Department of Electrical and Computer Engineering, University of Toronto, Toronto, ON, Canada; North York General Hospital, Toronto, ON, Canada.
| |
Collapse
|
2
|
Ghasemzadeh H, Hillman RE, Mehta DD. Toward Generalizable Machine Learning Models in Speech, Language, and Hearing Sciences: Estimating Sample Size and Reducing Overfitting. JOURNAL OF SPEECH, LANGUAGE, AND HEARING RESEARCH : JSLHR 2024; 67:753-781. [PMID: 38386017 PMCID: PMC11005022 DOI: 10.1044/2023_jslhr-23-00273] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Grants] [Track Full Text] [Subscribe] [Scholar Register] [Received: 04/25/2023] [Revised: 08/29/2023] [Accepted: 12/19/2023] [Indexed: 02/23/2024]
Abstract
PURPOSE Many studies using machine learning (ML) in speech, language, and hearing sciences rely upon cross-validations with single data splitting. This study's first purpose is to provide quantitative evidence that would incentivize researchers to instead use the more robust data splitting method of nested k-fold cross-validation. The second purpose is to present methods and MATLAB code to perform power analysis for ML-based analysis during the design of a study. METHOD First, the significant impact of different cross-validations on ML outcomes was demonstrated using real-world clinical data. Then, Monte Carlo simulations were used to quantify the interactions among the employed cross-validation method, the discriminative power of features, the dimensionality of the feature space, the dimensionality of the model, and the sample size. Four different cross-validation methods (single holdout, 10-fold, train-validation-test, and nested 10-fold) were compared based on the statistical power and confidence of the resulting ML models. Distributions of the null and alternative hypotheses were used to determine the minimum required sample size for obtaining a statistically significant outcome (5% significance) with 80% power. Statistical confidence of the model was defined as the probability of correct features being selected for inclusion in the final model. RESULTS ML models generated based on the single holdout method had very low statistical power and confidence, leading to overestimation of classification accuracy. Conversely, the nested 10-fold cross-validation method resulted in the highest statistical confidence and power while also providing an unbiased estimate of accuracy. The required sample size using the single holdout method could be 50% higher than what would be needed if nested k-fold cross-validation were used. Statistical confidence in the model based on nested k-fold cross-validation was as much as four times higher than the confidence obtained with the single holdout-based model. A computational model, MATLAB code, and lookup tables are provided to assist researchers with estimating the minimum sample size needed during study design. CONCLUSION The adoption of nested k-fold cross-validation is critical for unbiased and robust ML studies in the speech, language, and hearing sciences. SUPPLEMENTAL MATERIAL https://doi.org/10.23641/asha.25237045.
Collapse
Affiliation(s)
- Hamzeh Ghasemzadeh
- Center for Laryngeal Surgery and Voice Rehabilitation, Massachusetts General Hospital, Boston
- Department of Surgery, Harvard Medical School, Boston, MA
- Department of Communicative Sciences and Disorders, Michigan State University, East Lansing
| | - Robert E. Hillman
- Center for Laryngeal Surgery and Voice Rehabilitation, Massachusetts General Hospital, Boston
- Department of Surgery, Harvard Medical School, Boston, MA
- Speech and Hearing Bioscience and Technology, Division of Medical Sciences, Harvard Medical School, Boston, MA
- MGH Institute of Health Professions, Boston, MA
| | - Daryush D. Mehta
- Center for Laryngeal Surgery and Voice Rehabilitation, Massachusetts General Hospital, Boston
- Department of Surgery, Harvard Medical School, Boston, MA
- Speech and Hearing Bioscience and Technology, Division of Medical Sciences, Harvard Medical School, Boston, MA
- MGH Institute of Health Professions, Boston, MA
| |
Collapse
|
3
|
Hoffmann J, Roldan-Vasco S, Krüger K, Niekiel F, Hansen C, Maetzler W, Orozco-Arroyave JR, Schmidt G. Pilot Study: Magnetic Motion Analysis for Swallowing Detection Using MEMS Cantilever Actuators. SENSORS (BASEL, SWITZERLAND) 2023; 23:3594. [PMID: 37050654 PMCID: PMC10099077 DOI: 10.3390/s23073594] [Citation(s) in RCA: 2] [Impact Index Per Article: 2.0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Figures] [Subscribe] [Scholar Register] [Received: 02/24/2023] [Revised: 03/22/2023] [Accepted: 03/28/2023] [Indexed: 06/19/2023]
Abstract
The swallowing process involves complex muscle coordination mechanisms. When alterations in such mechanisms are produced by neurological conditions or diseases, a swallowing disorder known as dysphagia occurs. The instrumental evaluation of dysphagia is currently performed by invasive and experience-dependent techniques. Otherwise, non-invasive magnetic methods have proven to be suitable for various biomedical applications and might also be applicable for an objective swallowing assessment. In this pilot study, we performed a novel approach for deglutition evaluation based on active magnetic motion sensing with permanent magnet cantilever actuators. During the intake of liquids with different consistency, we recorded magnetic signals of relative movements between a stationary sensor and a body-worn actuator on the cricoid cartilage. Our results indicate the detection capability of swallowing-related movements in terms of a characteristic pattern. Consequently, the proposed technique offers the potential for dysphagia screening and biofeedback-based therapies.
Collapse
Affiliation(s)
- Johannes Hoffmann
- Department of Electrical and Information Engineering, Faculty of Engineering, Kiel University, 24118 Kiel, Germany
| | - Sebastian Roldan-Vasco
- GITA Lab, Faculty of Engineering, Universidad de Antioquia, Medellín 050010, Colombia
- Faculty of Engineering, Instituto Tecnológico Metropolitano, Medellín 050536, Colombia
| | - Karolin Krüger
- Department of Electrical and Information Engineering, Faculty of Engineering, Kiel University, 24118 Kiel, Germany
| | - Florian Niekiel
- Fraunhofer Institute for Silicon Technology ISIT, 25524 Itzehoe, Germany
| | - Clint Hansen
- Department of Neurology, Kiel University, 24118 Kiel, Germany
| | - Walter Maetzler
- Department of Neurology, Kiel University, 24118 Kiel, Germany
| | - Juan Rafael Orozco-Arroyave
- GITA Lab, Faculty of Engineering, Universidad de Antioquia, Medellín 050010, Colombia
- Pattern Recognition Lab, Friedrich-Alexander-Universität, 91054 Erlangen, Germany
| | - Gerhard Schmidt
- Department of Electrical and Information Engineering, Faculty of Engineering, Kiel University, 24118 Kiel, Germany
| |
Collapse
|
4
|
Khalifa Y, Donohue C, Coyle JL, Sejdic E. Autonomous Swallow Segment Extraction Using Deep Learning in Neck-Sensor Vibratory Signals From Patients With Dysphagia. IEEE J Biomed Health Inform 2023; 27:956-967. [PMID: 36417738 PMCID: PMC10079637 DOI: 10.1109/jbhi.2022.3224323] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/27/2022]
Abstract
Dysphagia occurs secondary to a variety of underlying etiologies and can contribute to increased risk of adverse events such as aspiration pneumonia and premature mortality. Dysphagia is primarily diagnosed and characterized by instrumental swallowing exams such as videofluoroscopic swallowing studies. videofluoroscopic swallowing studies involve the inspection of a series of radiographic images for signs of swallowing dysfunction. Though effective, videofluoroscopic swallowing studies are only available in certain clinical settings and are not always desirable or feasible for certain patients. Because of the limitations of current instrumental swallow exams, research studies have explored the use of acceleration signals collected from neck sensors and demonstrated their potential in providing comparable radiation-free diagnostic value as videofluoroscopic swallowing studies. In this study, we used a hybrid deep convolutional recurrent neural network that can perform multi-level feature extraction (localized and across time) to annotate swallow segments automatically via multi-channel swallowing acceleration signals. In total, we used signals and videofluoroscopic swallowing study images of 3144 swallows from 248 patients with suspected dysphagia. Compared to other deep network variants, our network was superior at detecting swallow segments with an average area under the receiver operating characteristic curve value of 0.82 (95% confidence interval: 0.807-0.841), and was in agreement with up to 90% of the gold standard-labeled segments.
Collapse
|
5
|
Qiao J, Jiang YT, Dai Y, Gong YB, Dai M, Liu YX, Dou ZL. Research on a real-time dynamic monitoring method for silent aspiration after stroke based on semisupervised deep learning: A protocol study. Digit Health 2023; 9:20552076231183548. [PMID: 37434729 PMCID: PMC10331777 DOI: 10.1177/20552076231183548] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/03/2022] [Accepted: 06/05/2023] [Indexed: 07/13/2023] Open
Abstract
Objective This study aims to establish a real-time dynamic monitoring system for silent aspiration (SA) to provide evidence for the early diagnosis of and precise intervention for SA after stroke. Methods Multisource signals, including sound, nasal airflow, electromyographic, pressure and acceleration signals, will be obtained by multisource sensors during swallowing events. The extracted signals will be labeled according to videofluoroscopic swallowing studies (VFSSs) and input into a special dataset. Then, a real-time dynamic monitoring model for SA will be built and trained based on semisupervised deep learning. Model optimization will be performed based on the mapping relationship between multisource signals and insula-centered cerebral cortex-brainstem functional connectivity through resting-state functional magnetic resonance imaging. Finally, a real-time dynamic monitoring system for SA will be established, of which the sensitivity and specificity will be improved by clinical application. Results Multisource signals will be stably extracted by multisource sensors. Data from a total of 3200 swallows will be obtained from patients with SA, including 1200 labeled swallows from the nonaspiration category from VFSSs and 2000 unlabeled swallows. A significant difference in the multisource signals is expected to be found between the SA and nonaspiration groups. The features of labeled and pseudolabeled multisource signals will be extracted through semisupervised deep learning to establish a dynamic monitoring model for SA. Moreover, strong correlations are expected to be found between the Granger causality analysis (GCA) value (from the left middle frontal gyrus to the right anterior insula) and the laryngeal rise time (LRT). Finally, a dynamic monitoring system will be established based on the former model, by which SA can be identified precisely. Conclusion The study will establish a real-time dynamic monitoring system for SA with high sensitivity, specificity, accuracy and F1 score.
Collapse
Affiliation(s)
- Jia Qiao
- Department of Rehabilitation Medicine, The Third Affiliated Hospital of Sun Yat-Sen University
| | - Yuan-tong Jiang
- School of Software Engineering, South China University of Technology
| | - Yong Dai
- Clinical Medical College of Acupuncture-Moxibustion and Rehabilitation, Guangzhou University of Chinese Medicine
| | - Yan-bin Gong
- Department of Computer Science and Engineering, The Hong Kong University of Science and Technology
| | - Meng Dai
- Department of Rehabilitation Medicine, The Third Affiliated Hospital of Sun Yat-Sen University
| | - Yan-xia Liu
- School of Software Engineering, South China University of Technology
| | - Zu-lin Dou
- Department of Rehabilitation Medicine, The Third Affiliated Hospital of Sun Yat-Sen University
| |
Collapse
|
6
|
Schwartz R, Khalifa Y, Lucatorto E, Perera S, Coyle J, Sejdic E. A Preliminary Investigation of Similarities of High Resolution Cervical Auscultation Signals Between Thin Liquid Barium and Water Swallows. IEEE JOURNAL OF TRANSLATIONAL ENGINEERING IN HEALTH AND MEDICINE 2022; 10:4900109. [PMID: 34963825 PMCID: PMC8694539 DOI: 10.1109/jtehm.2021.3134926] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Grants] [Track Full Text] [Download PDF] [Figures] [Subscribe] [Scholar Register] [Received: 06/03/2021] [Revised: 08/27/2021] [Accepted: 12/03/2021] [Indexed: 11/06/2022]
Abstract
Dysphagia, commonly referred to as abnormal swallowing, affects millions of people annually. If not diagnosed expeditiously, dysphagia can lead to more severe complications, such as pneumonia, nutritional deficiency, and dehydration. Bedside screening is the first step of dysphagia characterization and is usually based on pass/fail tests in which a nurse observes the patient performing water swallows to look for dysphagia overt signs such as coughing. Though quick and convenient, bedside screening only provides low-level judgment of impairment, lacks standardization, and suffers from subjectivity. Recently, high resolution cervical auscultation (HRCA) has been investigated as a less expensive and non-invasive method to diagnose dysphagia. It has shown strong preliminary evidence of its effectiveness in penetration-aspiration detection as well as multiple swallow kinematics. HRCA signals have traditionally been collected and investigated in conjunction with videofluoroscopy exams which are performed using barium boluses including thin liquid. An HRCA-based bedside screening is highly desirable to expedite the initial dysphagia diagnosis and overcome all the drawbacks of the current pass/fail screening tests. However, all research conducted for using HRCA in dysphagia is based on thin liquid barium boluses and thus not guaranteed to provide valid results for water boluses used in bedside screening. If HRCA signals show no significant differences between water and thin liquid barium boluses, then the same algorithms developed on thin liquid barium boluses used in diagnostic imaging studies, it can be then directly used with water boluses. This study investigates the similarities and differences between HRCA signals from thin liquid barium swallows compared to those signals from water swallows. Multiple features from the time, frequency, time-frequency, and information-theoretic domain were extracted from each type of swallow and a group of linear mixed models was tested to determine the significance of differences. Machine learning classifiers were fit to the data as well to determine if the swallowed material (thin liquid barium or water) can be correctly predicted from an unlabeled set of HRCA signals. The results demonstrated that there is no systematic difference between the HRCA signals of thin liquid barium swallows and water swallows. While no systematic difference was discovered, the evidence of complete conformity between HRCA signals of both materials was inconclusive. These results must be validated further to confirm conformity between the HRCA signals of thin liquid barium swallows and water swallows.
Collapse
Affiliation(s)
- Ryan Schwartz
- Department of Electrical and Computer EngineeringSwanson School of EngineeringUniversity of Pittsburgh Pittsburgh PA 15261 USA
| | - Yassin Khalifa
- Department of Electrical and Computer EngineeringSwanson School of EngineeringUniversity of Pittsburgh Pittsburgh PA 15261 USA
| | - Erin Lucatorto
- Department of Communication Science and DisordersSchool of Health and Rehabilitation SciencesUniversity of Pittsburgh Pittsburgh PA 15260 USA
| | - Subashan Perera
- Division of Geriatric MedicineDepartment of MedicineUniversity of Pittsburgh Pittsburgh PA 15261 USA
| | - James Coyle
- Department of Communication Science and DisordersSchool of Health and Rehabilitation SciencesUniversity of Pittsburgh Pittsburgh PA 15260 USA
| | - Ervin Sejdic
- Department of Electrical and Computer EngineeringSwanson School of EngineeringUniversity of Pittsburgh Pittsburgh PA 15261 USA
- The Edward S. Rogers Department of Electrical and Computer EngineeringFaculty of Applied Science and EngineeringUniversity of Toronto Toronto ON M5S 2E4 Canada
- North York General Hospital Toronto ON M2K 1E1 Canada
| |
Collapse
|