Automated diagnosis based at least in part on pulse waveforms

• US 10,299,735 B2
• Filed: 08/21/2015
• Issued: 05/28/2019
• Est. Priority Date: 08/22/2014
• Status: Expired due to Fees

• Current Assignee: Pulse Tectonics, LLC
• Original Assignee: Pulse Tectonics, LLC
• Inventors: Furness, III, Thomas Adrian, Melville, Ross, Doane, Robert, Pal, Brian
• Primary Examiner(s): Downey, John R
• Application Number:
  US15/505,844
• Publication Number:
  US 20170258336A1
• Time in Patent Office: 1,376 Days
• CPC Class Codes: A61B 5/0004 characterised by the type of physiological signal transmitted
• CPC Class Codes: A61B 5/024 Detecting, measuring or recording pulse rate or heart rate A61B5/0205, A61B5/021 take precedence
• CPC Class Codes: A61B 5/02438 with portable devices, e.g. worn by the patient
• CPC Class Codes: A61B 5/026 Measuring blood flow A61B3/1233, A61B3/1241 take precedence
• CPC Class Codes: A61B 5/318 Heart-related electrical modalities, e.g. electrocardiography [ECG]
• CPC Class Codes: A61B 5/6824 Arm or wrist
• CPC Class Codes: A61B 5/6826 Finger
• CPC Class Codes: A61B 5/7225 Details of analog processing, e.g. isolation amplifier, gain or sensitivity adjustment, filtering, baseline or drift compensation A61B5/30, A61B5/316 take precedence; transformers for specific applications H01F38/00; amplifiers H03F
• CPC Class Codes: A61B 5/7246 using correlation, e.g. template matching or determination of similarity
• CPC Class Codes: A61B 5/7264 Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems neural networks per se G06N3/00; expert systems per se G06N5/00
• CPC Class Codes: A61B 5/7275 Determining trends in physiological measurement data; Predicting development of a medical condition based on physiological measurements, e.g. determining a risk factor
• CPC Class Codes: A61B 5/7278 Artificial waveform generation or derivation, e.g. synthesising signals from measured signals A61B5/327 takes precedence
• CPC Class Codes: A61B 5/7475 User input or interface means, e.g. keyboard, pointing device, joystick
• CPC Class Codes: G16H 50/20 for computer-aided diagnosis, e.g. based on medical expert systems

Abstract

A front-end system collects diagnostically relevant information or data, including pulse waveform information, and optionally timing information, symptom information, and/or diagnostic information. One or more sensors or transducers detect pulse waveforms in a consistent manner, across various subjects and practitioners. A back-end system stores diagnostically relevant information and diagnostic information, typically for a large population of subjects and practitioners. The back-end system generates system generated diagnoses and/or suggested remedial actions based on submitted inquiries or requests. The back-end system can advantageously employ machine learning techniques to identify correlations between defining characteristics or features of pulse waveforms and diagnoses, over a large number of samples submitted by a substantial number of practitioners, for instance of traditional Chinese medicine. The back-end system can also employ primary and/or secondary symptoms in identifying correlations.

Claims

1. A diagnostic system, comprising:

at least one processor circuit;

at least one nontransitory memory communicatively coupled the at least one processor circuit and which stores at least one of processor executable instructions or data, execution of which causes the at least processor circuit to;

for each of a number of visits by each of a plurality of subjects, receive respective pulse signal information representative of;

i) a first pulse signal waveform captured from the respective subject at a first applied pressure at a first location, ii) at least a second pulse signal waveform captured from the respective subject a second applied pressure at the first location, the second applied pressure different than the first applied pressure, and iii) at least one value indicative of at least one of the first or the second applied pressures, and receive diagnosis information that includes at least a primary diagnosis associated with the respective visit by the respective subject;

for each of the number of visits by each of the plurality of subjects, receive timing information which represents an event in a cardiac cycle that is sensed separately from the plurality of pulse waveforms;

store the received respective pulse signal information, timing information, and diagnosis information to the at least one nontransitory memory; and

perform machine learning on the pulse waveform, the timing information, and diagnosis information during a training time to determine one or more correlations between various ones of a number of defining characteristics of representations of the first and at least the second pulse signal waveforms and the diagnosis information.

2. The diagnostic system of claim 1 wherein for each of the number of visits by each of the plurality of subjects, receives respective pulse signal information representative of:

iii) a third pulse signal waveform captured from the respective subject at a third applied pressure at the first location, the third applied pressured different from the first and the second applied pressures.

3. The diagnostic system of claim 2 wherein for a given one of the visits of a given one of the subjects, the received respective pulse signal information further represents a fourth, a fifth and a sixth pulse signal waveform captured from the respective subject at a first, a second and a third applied pressure, respectively, at a second location along a given artery, the second location spaced from the first location.

4. The diagnostic system of claim 3 wherein for a given one of the visits of a given one of the subjects, the first, the second and the third pulse signal waveforms are captured temporally proximate to one another.

5. The diagnostic system of claim 3 wherein for a given one of the visits of a given one of the subjects, the first, the second, the third, the fourth, the fifth and the sixth pulse signal waveforms are captured temporally proximate to one another.

6. The diagnostic system of claim 2 wherein for the given one of the visits of the given one of the subjects, the first, the second and the third pulse signal waveforms represent pulse waveforms captured at a first location along the radial artery proximate a wrist bone on a thumb side of a hand of the subject at the first, the second and the third applied pressures, respectively, and the fourth, the fifth, and the sixth pulse signal waveforms represent pulse waveforms captured at a second location along the radial artery proximate the ulnar, at the first, the second and the third applied pressures, respectively, the second location spaced from the first location.

7. The diagnostic system of claim 6 wherein for a given one of the visits of a given one of the subjects, the received respective pulse signal information further represents a seventh, an eight and a ninth pulse signal waveform captured from the respective subject at a first, a second and a third applied pressure, respectively, at a third location along the radial artery, the third location closely spaced from the first and the second locations.

8. The diagnostic system of claim 7 wherein for a given one of the visits of a given one of the subjects, the first, the second, the third, the fourth, the fifth, the sixth, the seventh, the eight, and the ninth pulse signal waveforms are captured substantially concurrently with one another.

9. The diagnostic system of claim 6 wherein the first, the second and the third pulse signal waveforms each represents respective ones of pulse waveforms captured through a skin of the given subject.

10. The method of claim 1 wherein for at least a subset of the number of visits by at least a subset of the plurality of subjects the received diagnosis information includes at least a secondary diagnosis associated with the respective visit by the respective subject.

11. The diagnostic system of claim 1 wherein for at least a subset of the number of visits by at least a subset of the plurality of subjects, receives at least a primary symptom associated with the respective visit by the respective subject.

12. The diagnostic system of claim 11 wherein for at least a subset of the subset of visits by at least a subset of the subset of subjects, receives at least a secondary symptom associated with the respective visit by the respective subject.

13. The diagnostic system of claim 11 wherein for at least a subset of the number of visits by at least a subset of the plurality of subjects the received diagnosis information includes at least a secondary diagnosis associated with the respective visit by the respective subject.

14. The diagnostic system of claim 1 wherein, during a runtime, the runtime following the training time, for each of a plurality of additional visits, the at least one processor circuit receives pulse waveform information for the respective visit and respective subject, and determines a diagnosis based at least in part on the received pulse waveform information and the determined correlations, and provides the determined diagnosis.

15. The diagnostic system of claim 1 wherein, during a runtime, the runtime following the training time, for each of a plurality of additional visits, the at least one processor circuit receives pulse waveform information and symptom information for the respective visit and respective subject, and determines the diagnosis based at least in part on the received pulse waveform information, the symptom information and the determined correlations, and provides the determined diagnosis.

16. The diagnostic system of claim 1 wherein the timing information represents an electrocardiogram (EKG) signal sensed with an EKG transducer, the pulse signal waveforms being delayed with respect to the electrocardiogram (EKG) signal by a known range of time.

17. The diagnostic system of claim 16 wherein the at least one processor circuit further identifies at least one identifying characteristic in at least one of the pulse signal waveforms based at least in part on the timing information, for each of at least a subset of the number of visits by each of at least a subset of the plurality of subjects.

18. The diagnostic system of claim 1 wherein the received timing information is electrocardiogram timing information which represents an event in a cardiac cycle that is sensed separately from the plurality of pulse waveforms;

and the at least one processor circuit further identifies at least one identifying characteristic in at least one of the pulse signal waveforms based at least in part on the timing information, for each of at least a subset of the number of visits by each of at least a subset of the plurality of subjects.

19. A method operation in a diagnostic system that includes at least one processor circuit and at least one nontransitory memory communicatively coupled the at least one processor circuit and which stores at least one of processor executable instructions or data, the method comprising:

for each of a number of visits by each of a plurality of subjects, receiving respective pulse signal information representative of;

i) a first pulse signal waveform captured from the respective subject at a first applied pressure at a first location, ii) at least a second pulse signal waveform captured from the respective subject a second applied pressure at the first location, the second applied pressure different than the first applied pressure, and iii) at leave one value indicative of at least one of the first or the second applied pressures, receiving timing information which represents an event in a cardiac cycle that is sensed separately from the plurality of pulse waveforms and receiving diagnosis information that includes at least a primary diagnosis associated with the respective visit by the respective subject;

storing the received respective pulse signal information, timing information and diagnosis information to the at least one nontransitory memory; and

determining one or more correlations between various ones of a number of defining characteristics of representations of the first and at least the second pulse signal waveforms and the diagnosis information.

20. The method of claim 19, further comprising:

performing machine learning on the pulse waveform and diagnosis information during a training time, by the at least one processor circuit, to determine one or more correlations between various ones of a number of defining characteristics of representations of the first and at least the second pulse signal waveforms and the diagnosis information.

21. The method of claim 20 wherein the timing information represents an electrocardiogram (EKG) signal sensed with an EKG transducer, the pulse signal waveforms being delayed with respect to the electrocardiogram (EKG) signal by a known range of time.