Methods and devices using swallowing accelerometry signals for swallowing impairment detection

Abstract

A method classifies vibrational data acquired for a swallowing event to identify a possible swallowing impairment in a candidate. The method includes receiving axis-specific vibrational data for an anterior-posterior (A-P) axis and a superior-inferior (S-I) axis and representative of the swallowing event, for example from a sensor operatively coupled to a processing module that is a local or remote computing device. A portion of the axis-specific vibrational data for the A-P axis can be combined with a portion of the axis-specific vibrational data for the S-I axis on the processing module using one or more of linear combination, squared (power) sum, moving window correlation of the two signals, local minimum or local maximum of the two signals, and trigonometric relation. The method can include outputting from the processing module a classification of the swallowing event based on the combined vibrational data.

1 . A device for identifying a possible swallowing impairment in a candidate during execution of a swallowing event, the device comprising:
 an accelerometer configured to acquire axis-specific vibrational data along an anterior-posterior (A-P) axis and a superior-inferior (S-I) axis of the candidate's throat, the axis-specific vibrational data is representative of the swallowing event; and   a processing module that is a local or remote computing device operatively coupled to the accelerometer, the processing module configured for processing the axis-specific data to (i) combine at least a portion of a first signal comprising the axis-specific vibrational data acquired along the A-P axis with at least a portion of a second signal comprising the axis-specific vibrational data acquired along the S-I axis using at least one process selected from the group consisting of linear combination, squared (power) sum, moving window correlation of the two signals, local minimum or local maximum of the two signals, and trigonometric relation and (ii) classify the swallowing event as one of a plurality of classifications based on the combined vibrational data, the plurality of classifications comprising a first classification indicative of normal swallowing and a second classification indicative of possibly impaired swallowing.   
     
     
         2 . The device of  claim 1 , wherein the processing module is configured to extract meta-features from the combined vibrational data. 
     
     
         3 . The device of  claim 2 , wherein the processing module is configured to use the meta-features extracted from the combined vibrational data to classify the swallowing event. 
     
     
         4 . The device of  claim 3 , wherein the processing module is configured to compare the meta-features extracted from the combined vibrational data to preset classification criteria to classify the swallowing event. 
     
     
         5 . The device of  claim 1 , wherein the second classification is indicative of at least one of a swallowing safety impairment or a swallowing efficiency impairment. 
     
     
         6 . The device of  claim 1 , wherein the second classification is indicative of at least one of penetration or aspiration, and the processing module is configured to further classify the swallowing event as a first event indicative of a safe event or a second event indicative of an unsafe event. 
     
     
         7 . The device of  claim 1 , wherein the processing module is configured to classify multiple successive swallowing events by classifying the combined vibrational data for each of the successive swallowing events as indicative of one of the first classification or the second classification. 
     
     
         8 . The device of  claim 1 , wherein the processing module uses a non-segmented spectrogram for pre-processing of at least one of (i) the axis-specific vibrational data along the A-P axis, (ii) the axis-specific vibrational data along the S-I axis or (iii) the combined vibrational data. 
     
     
         9 . A method for classifying cervical accelerometry data acquired for a swallowing event to identify a possible swallowing impairment in a candidate, the method comprising:
 receiving axis-specific vibrational data for an anterior-posterior (A-P) axis and a superior-inferior (S-I) axis and representative of the swallowing event, a processing module that is a local or remote computing device operatively coupled to an accelerometer receives the axis-specific vibrational data from the accelerometer;   combining at least a portion of a first signal comprising the axis-specific vibrational data for the A-P axis with at least a portion of a second signal comprising the axis-specific vibrational data for the S-I axis using at least one process selected from the group consisting of linear combination, squared (power) sum, moving window correlation of the two signals, local minimum or local maximum of the two signals, and trigonometric relation, the processing module forms the combined vibrational data; and   outputting a classification of the swallowing event as one of a plurality of classifications based on the combined vibrational data, the plurality of classifications comprising a first classification indicative of normal swallowing and a second classification indicative of possibly impaired swallowing, and the processing module outputs the classification.   
     
     
         10 . The method of  claim 9 , comprising extracting meta-features from the combined vibrational data on the processing module. 
     
     
         11 . The method of  claim 10 , comprising using the meta-features extracted from the combined vibrational data to classify the swallowing event on the processing module. 
     
     
         12 . The method of  claim 11 , comprising comparing the meta-features extracted from the combined vibrational data to preset classification criteria to classify the swallowing event on the processing module. 
     
     
         13 . The method of  claim 12 , wherein the preset classification criteria are defined for each of swallowing safety and swallowing efficiency. 
     
     
         14 . The method of  claim 13 , wherein the preset classification criteria are defined by features previously extracted and classified from a known training data set. 
     
     
         15 . The method of  claim 9 , wherein the second classification is indicative of at least one of a swallowing safety impairment or a swallowing efficiency impairment. 
     
     
         16 . The method of  claim 9 , wherein the second classification is indicative of at least one of penetration or aspiration, and the method comprises further classifying the swallowing event as a first event indicative of a safe event or a second event indicative of an unsafe event. 
     
     
         17 . The method of  claim 9 , comprising classifying successive swallowing events by classifying the combined vibrational data for each of the successive swallowing events as indicative of one of the first classification or the second classification. 
     
     
         18 . The method of  claim 9 , comprising using a non-segmented spectrogram for pre-processing of at least one of (i) the axis-specific vibrational data along the A-P axis and the S-I axis or (ii) the combined vibrational data