Patent Issued for Electronic auscultation and improved identification of auscultation audio samples (USPTO 11751774): UnitedHealth Group Incorporated

2023 OCT 02 (NewsRx) -- By a News Reporter-Staff News Editor at Insurance Daily News -- UnitedHealth Group Incorporated (Minnetonka, Minnesota, United States) has been issued patent number 11751774, according to news reporting originating out of Alexandria, Virginia, by NewsRx editors.

The patent’s inventors are Boss, Gregory J. (Saginaw, MI, US), Gordon, Marilyn L. (Cherry Hill, NJ, US), Muse, Jon Kevin (Thompsons Station, TN, US), Ravindranathan, Rama S. (Edison, NJ, US).

This patent was filed on November 12, 2020 and was published online on September 12, 2023.

From the background information supplied by the inventors, news correspondents obtained the following quote: “Auscultation is the listening to the internal sounds of the body for the purposes of, for example, examining the circulatory and/or respiratory systems. As auscultation often depends on the medically trained ear of a physician, using auscultation to perform regular monitoring of patients, especially patients who are at home or otherwise not in a clinical setting, is generally not attempted.”

Supplementing the background information on this patent, NewsRx reporters also obtained the inventors’ summary information for this patent: “In general, embodiments of the present invention provide methods, apparatus, systems, computing entities, and/or the like for performing electronic auscultation and processing captured auscultation data samples to determine whether any suspect heart or lung sounds are present. In various embodiments, the captured auscultation data samples are processed to identify any suspect heart or lung sounds present in the captured auscultation data samples. In various embodiments, the identified suspect heart or lung sounds may be used to generate an automated diagnosis and/or provided through a dashboard, for example, for physician review.

“In various embodiments, sensors may be placed on a patient’s body. For example, the sensors may include audio sensors configured to record and/or digitize sound, pulse and/or heart rate sensors configured to detect a patient’s pulse, stretch sensors configured to monitor the expansion and contraction of a patient’s chest as the patient inhales and exhales, and/or the like. In an example embodiment, the sensors may be incorporated into and attached to a wearable device, such as a vest and/or the like to enable easy, reproducible, and accurate placement of the sensors on the patient’s body (e.g., chest, back, and/or the like). The sensors may then be used to capture one or more instances of data samples. The data samples may include information/data indicative of a patient’s heart rate (e.g., pulse and/or the like) and audio samples of heart and/or lung sounds of the patient. The audio samples may be processed to generate heart sound samples and/or lung sound samples (e.g., by removing the lung sounds from an audio sample to generate a heart sound sample or by removing the heart sounds form an audio sample to generate a lung sound sample). By mixing and/or combining different audio samples that were captured simultaneously by different audio sensors on the patient’s body in an out of phase manner, any suspect sounds present may be isolated and analyzed. For example, any suspect sounds present in the audio samples may be identified and/or classified.

“In various embodiments, when a suspect sound is identified in an audio sample, additional instances of data samples may be captured. For example, instructions may be provided to a patient to say a particular word, breathe or move their body in a particular manner, assume a particular body position, and/or the like. For example, the instructions may be configured to enable improved detection and/or determination of details regarding the suspect sound. The additional instances of data samples may be analyzed to isolate any suspect sounds present therein. The suspect sounds may then be stored (e.g., in a memory, database, electronic health record, and/or the like) and/or provided for physician review.

“In accordance with an aspect of the present disclosure, a method for identifying suspect heart or lung sounds is provided, the method comprising: obtaining, by processing circuitry, an instance of data samples, the instance of data samples comprising at least two audio samples and respiratory and/or cardiac cycle data, the audio samples comprising auscultation sounds of an individual; based on the respiratory and/or cardiac cycle data, processing the at least two audio samples to generate (a) at least two lung sound samples or (b) at least two heart sound samples; identifying a primary sound sample and at least one secondary sound sample from the (a) at least two lung sound samples or (b) at least two heart sound samples; generating at least one of (a) a modified secondary sound sample by changing a phase of the at least one secondary sound sample by 180 degrees or (b) a modified baseline sample by changing a phase of a baselines sample accessed from memory by 180 degrees; combining the at least one of (a) the modified secondary sound sample or (b) the modified baseline sample with the primary sound sample to generate an isolated sound sample; and determining whether the isolated sound sample includes a suspect sound.

“In accordance with another aspect of the present disclosure, a method for generating an isolated sound sample is provided, the method comprising: identifying, by a processing element, a suspect sound present in an auscultation audio sample captured by one or more sensors on a patient; identifying, by the processing element, one or more contexts for capturing additional instances of data samples based on the suspect sound; causing, by the processing element, an output device to provide instructions regarding the one or more contexts; obtaining, by the processing element, one or more additional instances of data samples, each of the one or more additional instances of data samples captured while the patient was performing one of the one or more contexts; analyzing, by the processing elements at least one of the one or more additional instances of data samples to generate an isolated sound sample, the analyzing of the one or more additional instances of data samples comprising: identifying a primary sound sample and at least one secondary sound sample from auscultation audio samples of the at least one of the one or more additional instances of data samples; generating a modified secondary sound sample or a modified baseline sample by changing the phase of the at last one secondary sound sample or a baseline sample by 180 degrees; and combining the modified secondary sound sample or the modified baseline sample with the primary sound sample to generate an isolated sound sample.”

The claims supplied by the inventors are:

“1. A computer-implemented method comprising: receiving, by one or more processors, an instance of data samples, the instance of data samples comprising (a) at least two audio samples and (b) at least one of respiratory data or cardiac cycle data, the at least two audio samples comprising auscultation sounds of an individual; generating, by the one or more processors, at least one of (a) at least two lung sound samples based at least in part on the at least two audio samples or (b) at least two heart sound samples based at least in part on the at least two audio samples by removing at least one of heart sounds or lung sounds from the at least two audio samples based at least in part on the at least one of the respiratory data or the cardiac cycle data; identifying, by the one or more processors, a primary sound sample and at least one secondary sound sample from at least one of (a) the at least two lung sound samples or (b) the at least two heart sound samples; at least one of (a) generating a modified secondary sound sample by changing a phase of the at least one secondary sound sample or (b) accessing a baseline sample from memory and generating a modified baseline sample by changing a phase of the baseline sample; generating an isolated sound sample by combining the primary sound sample with at least one of (a) the modified secondary sound sample or (b) the modified baseline sample; and determining whether the isolated sound sample includes a suspect heart or lung sound.

“2. The method of claim 1, wherein the at least two audio samples were captured simultaneously.

“3. The method of claim 1, wherein (a) the cardiac cycle data comprises pulse data and (b) the respiratory cycle data comprises a stretch sensor signal indicating expansion and contraction of a patient’s chest due to inhalation and exhalation.

“4. The method of claim 1, wherein determining whether the isolated sound sample includes the suspect heart or lung sound comprises identifying the suspect heart or lung sound in the isolated sound sample based at least in part on at least one of one or more frequency criteria, timing criteria, or consistency criteria.

“5. The method of claim 4, wherein the one or more frequency criteria comprise one or more frequency bandpass filters.

“6. The method of claim 1, further comprising adding the isolated sound sample to itself prior to determining whether the isolated sound sample includes the suspect heart or lung sound.

“7. The method of claim 1, further comprising: identifying the suspect heart or lung sound in the isolated sound sample; and storing (a) an identifier of the suspect heart or lung sound and (b) at least one of the instance of data samples or the isolated sound sample.

“8. The method of claim 7, wherein the suspect heart or lung sound and the at least one of the instance of data samples or the isolated sound sample are stored in association with metadata corresponding to a monitoring session in which the instance of data samples were captured for a patient.

“9. The method of claim 1, wherein the determination of whether the isolated sound sample comprises the suspect heart or lung sound is performed by a machine learning-trained model.

“10. The method of claim 1, wherein the modified secondary sound sample is generated by an operational amplifier.

“11. The method of claim 1, wherein (a) when the modified secondary sound sample is combined with the primary sound sample, no alignment of the modified secondary sound sample and the primary sound sample is performed and (b) when the modified baseline sample is combined with the primary sound sample, an alignment of the modified baseline sample and the primary sound sample is performed.

“12. The method of claim 1, wherein the suspect heart or lung sound is generated by the functioning of a respective one of a patient’s heart or lungs.

“13. The method of claim 1, further comprising: responsive to determining that the isolated sound sample comprises the suspect heart or lung sound, identifying instructions to provide for capturing of an additional instance of data samples; providing the instructions and capturing the additional instance of data samples; generating an additional isolated sound sample by analyzing the additional instance of data samples; and providing the additional isolated sound sample to be at least one of (a) stored in the memory or (b) graphically or audibly provided to a user.

“14. The method of claim 13, wherein the instructions indicate a context that a patient should perform during the capturing of the additional instance of data samples.

“15. The method of claim 14, wherein the context comprises at least one of (a) a body position, (b) a movement, © a breathing pattern, or (d) a word to speak.

“16. The method of claim 1, further comprising, when the modified secondary sound sample is combined with the primary sound sample, storing the isolated sound sample as a baseline sample in the memory.

“17. A computing apparatus comprising memory and one or more processors communicatively coupled to the memory, the one or more processors configured to: receive an instance of data samples, the instance of data samples comprising (a) at least two audio samples and (b) at least one of respiratory data or cardiac cycle data, the at least two audio samples comprising auscultation sounds of an individual; generate at least one of (a) at least two lung sound samples based at least in part on the at least two audio samples or (b) at least two heart sound samples based at least in part on the at least two audio samples by removing at least one of heart sounds or lung sounds from the at least two audio samples based at least in part on the at least one of the respiratory data or the cardiac cycle data; identify a primary sound sample and at least one secondary sound sample from at least one of (a) the at least two lung sound samples or (b) the at least two heart sound samples; at least one of (a) generate a modified secondary sound sample by changing a phase of the at least one secondary sound sample or (b) access a baseline sample from memory and generating a modified baseline sample by changing a phase of the baseline sample; generate an isolated sound sample by combining the primary sound sample with at least one of (a) the modified secondary sound sample or (b) the modified baseline sample; and determine whether the isolated sound sample includes a suspect heart or lung sound.

“18. The computing apparatus of claim 17, wherein determining whether the isolated sound sample includes the suspect heart or lung sound comprises identifying the suspect heart or lung sound in the isolated sound sample based at least in part on at least one of one or more frequency criteria, timing criteria, or consistency criteria.

“19. The computing apparatus of claim 18, wherein the one or more frequency criteria comprise one or more frequency bandpass filters.

“20. One or more non-transitory computer-readable storage media including instructions that, when executed by one or more processors, cause the one or more processors to: receive an instance of data samples, the instance of data samples comprising (a) at least two audio samples and (b) at least one of respiratory data or cardiac cycle data, the at least two audio samples comprising auscultation sounds of an individual; generate at least one of (a) at least two lung sound samples based at least in part on the at least two audio samples or (b) at least two heart sound samples based at least in part on the at least two audio samples by removing at least one of heart sounds or lung sounds from the at least two audio samples based at least in part on the at least one of the respiratory data or the cardiac cycle data; identify a primary sound sample and at least one secondary sound sample from at least one of (a) the at least two lung sound samples or (b) the at least two heart sound samples; at least one of (a) generate a modified secondary sound sample by changing a phase of the at least one secondary sound sample or (b) access a baseline sample from memory and generating a modified baseline sample by changing a phase of the baseline sample; generate an isolated sound sample by combining the primary sound sample with at least one of (a) the modified secondary sound sample or (b) the modified baseline sample; and determine whether the isolated sound sample includes a suspect heart or lung sound.”

For the URL and additional information on this patent, see: Boss, Gregory J. Electronic auscultation and improved identification of auscultation audio samples. U.S. Patent Number 11751774, filed November 12, 2020, and published online on September 12, 2023. Patent URL (for desktop use only): https://ppubs.uspto.gov/pubwebapp/external.html?q=(11751774)&db=USPAT&type=ids

(Our reports deliver fact-based news of research and discoveries from around the world.)