Automated diagnosis based at least in part on pulse waveforms
First Claim
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.
0 Assignments
0 Petitions
Accused Products
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.
13 Citations
21 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. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18)
-
-
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. - View Dependent Claims (20, 21)
-
Specification