System and method for biometric data capture for event prediction

1. A computer implemented system for generating an output data structure from data streams captured during a medical procedure, the system comprising:

a computer processor operating in conjunction with computer memory and a non-transitory computer readable storage medium, the computer processor configured to:

receive a biometric data stream from a biometric sensor coupled to a body of a healthcare practitioner, the biometric sensor adapted to capture electrocardiography (ECG) data;

generate, from processing the biometric data stream, one or more heart rate variability (HRV) data values, the HRV data values representing a variation in time between heartbeats of the healthcare practitioner and time-synchronized to timestamps of captured video or audio data streams;

identify, one or more abnormality-related durations of time during which the HRV data values are greater or lower than a pre-defined threshold data value indicative of potentially elevated stress levels of the healthcare practitioner; and

generate one or more time-based metadata tags indicative of the one or more abnormality-related durations of time for appending to the captured video or audio data streams, the one or more time-based metadata tags encapsulated into the output data structure;

wherein the computer processor is further configured to modify one or more capture characteristics of the capture of the captured video or audio data streams, modifications of the one or more capture characteristics including at least an increase in resolution or bitrate for capturing an increased data volume or load of data relating to the medical procedure during the one or more abnormality-related durations of time indicative of the potentially elevated stress levels of the healthcare practitioner, and wherein the system is a local computer server operating in conjunction with a centralized computer server, and wherein the computer processor is further configured to request additional bandwidth resources for transmission of the captured video or audio data streams to the centralized computer server, the increased data volume or load adapted to support generation of one or more prediction data objects representative of one or more predicted characteristics or incidents relating to the medical procedure that occur during the one or more abnormality-related durations of time; and

wherein the computer processor is further configured to modify the one or more capture characteristics of the capture of the captured video or audio data streams, the modifications of the one or more capture characteristics including at least temporarily activating one or more additional video or audio capture devices during the one or more abnormality-related durations of time.

2. The computer implemented system of claim 1 , wherein the pre-defined threshold data value is a baseline value established in respect of the healthcare practitioner, and the one or more abnormality-related durations of time are established where the HRV data values are indicative of a low variance that is at least one standard deviation below the baseline value.

3. The computer implemented system of claim 1 , wherein the modifications of the one or more capture characteristics shifts an operation mode of the capture characteristics from a normal operation mode to an intensive mode.

4. The computer implemented system of claim 1 , wherein the system is the local computer server operating in conjunction with the centralized computer server; and

wherein the computer processor is further configured to request additional processing resources to be allocated by the centralized computer server for automated analysis of the captured video or audio data streams at durations of time marked by the one or more time-based metadata tags.

5. The computer implemented system of claim 4 , wherein the centralized computer server is configured for extracting machine learning input features from the captured video or audio data streams and the one or more time-based metadata tags, and to process, using a trained machine learning model data architecture, the machine learning input features to generate one or more prediction data objects representative of one or more predicted characteristics or incidents relating to the medical procedure.

6. The computer implemented system of claim 5 , wherein the trained machine learning model data architecture is configured to operate in real or near-real time using the additional processing resources such that when the one or more prediction data objects indicate that a subset of one or more predicted characteristics or incidents relating to the medical procedure are occurring, the centralized computer server generates an alert control command data signal, and wherein the alert control command data signal causes an actuation of a tactile alert, a visual alert, or an audible alert.

7. The computer implemented system of claim 1 , wherein the output data structure includes the captured video or audio data streams augmented by a separate time-synchronized stream comprising the one or more time-based metadata tags indicative of the one or more abnormality-related durations of time.

8. The computer implemented system of claim 1 , wherein the computer processor is further configured to modify the one or more capture characteristics of the capture of the captured video or audio data streams, the modifications of the one or more capture characteristics including at least a change in data encoding during the one or more abnormality-related durations of time.

9. The computer implemented system of claim 8 , wherein the change in data encoding during the one or more abnormality-related durations of time includes a change in irreversible compression characteristics to reduce a proportion of data loss incurred during the data encoding relative to the data encoding of the captured video or audio data streams during times outside of the one or more abnormality-related durations of time.

10. The computer implemented system of claim 1 , wherein the one or more additional video or audio capture devices include at least a wide angle camera or an ambient microphone.

11. A computer implemented method to generate an output data structure from data streams captured during a medical procedure, the method comprising:

receiving a biometric data stream from a biometric sensor coupled to a body of a healthcare practitioner, the biometric sensor adapted to capture electrocardiography (ECG) data;

generating, from processing the biometric data stream, one or more heart rate variability (HRV) data values, the HRV data values representing a variation in time between heartbeats of the healthcare practitioner and time-synchronized to timestamps of captured video or audio data streams;

identifying, one or more abnormality-related durations of time during which the HRV data values are greater or lower than a pre-defined threshold data value indicative of potentially elevated stress levels of the healthcare practitioner;

generating one or more time-based metadata tags indicative of the one or more abnormality-related durations of time for appending to the captured video or audio data streams, the one or more time-based metadata tags encapsulated into the output data structure;

modifying one or more capture characteristics of the capture of the captured video or audio data streams, modifications of the one or more capture characteristics including at least an increase in resolution or bitrate for capturing an increased data volume or load of data relating to the medical procedure during the one or more abnormality-related durations of time indicative of the potentially elevated stress levels of the healthcare practitioner, and wherein system capturing the data streams during the medical procedure is a local computer server operating in conjunction with a centralized computer server, and wherein the computer processor of the is further configured to request additional bandwidth resources for transmission of the captured video or audio data streams to the centralized computer server, the increased data volume or load adapted to support generation of one or more prediction data objects representative of one or more predicted characteristics or incidents relating to the medical procedure that occur during the one or more abnormality-related durations of time; and

modifying the one or more capture characteristics of the capture of the captured video or audio data streams, the modifications of the one or more capture characteristics including at least temporarily activating one or more additional video or audio capture devices during the one or more abnormality-related durations of time.

12. The computer implemented method of claim 11 , wherein the pre-defined threshold data value is a baseline value established in respect of the healthcare practitioner, and the one or more abnormality-related durations of time are established where the HRV data values are indicative of a low variance that is at least one standard deviation below the baseline value.

13. The computer implemented method of claim 11 , wherein the modifications of the one or more capture characteristics shifts an operation mode of the capture characteristics from a normal operation mode to an intensive mode.

14. The computer implemented method of claim 11 , wherein the method is by the computer processor a local computer server operating in conjunction with a centralized computer server; and

wherein the computer processor is further configured to request additional processing resources to be allocated by the centralized computer server for automated analysis of the captured video or audio data streams at durations of time marked by the one or more time-based metadata tags.

15. The computer implemented method of claim 14 , wherein the centralized computer server is configured for extracting machine learning input features from the captured video or audio data streams and the one or more time-based metadata tags, and to process, using a trained machine learning model data architecture, the machine learning input features to generate one or more prediction data objects representative of one or more predicted characteristics or incidents relating to the medical procedure.

16. The computer implemented method of claim 15 , wherein the trained machine learning model data architecture is configured to operate in real or near-real time using the additional processing resources such that when one or more prediction data objects indicate that a subset of one or more predicted characteristics or incidents relating to the medical procedure are occurring, the centralized computer server generates an alert control command data signal, and wherein the alert control command data signal causes an actuation of a tactile alert, a visual alert, or an audible alert.

17. The computer implemented method of claim 11 , wherein the output data structure includes the captured video or audio data streams augmented by a separate time-synchronized stream comprising the one or more time-based metadata tags indicative of the one or more abnormality-related durations of time.

18. The computer implemented method of claim 11 , wherein the computer processor is further configured to modify the one or more capture characteristics of the capture of the captured video or audio data streams, the modifications of the one or more capture characteristics including at least a change in data encoding during the one or more abnormality-related durations of time.

19. The computer implemented method of claim 18 , wherein the change in data encoding during the one or more abnormality-related durations of time includes a change in irreversible compression characteristics to reduce a proportion of data loss incurred during the data encoding relative to the data encoding of the captured video or audio data streams during times outside of the one or more abnormality-related durations of time.

20. A non-transitory computer readable medium storing machine interpretable instructions, which when executed by a computer processor, cause the computer processor to perform a computer implemented method to generate an output data structure from data streams captured during a medical procedure, the method comprising:

receiving a biometric data stream from a biometric sensor coupled to a body of a healthcare practitioner, the biometric sensor adapted to capture electrocardiography (ECG) data;

generating, from processing the biometric data stream, one or more heart rate variability (HRV) data values, the HRV data values representing a variation in time between heartbeats of the healthcare practitioner and time-synchronized to timestamps of captured video or audio data streams;

identifying, one or more abnormality-related durations of time during which the HRV data values are greater or lower than a pre-defined threshold data value indicative of potentially elevated stress levels of the healthcare practitioner; and

generating one or more time-based metadata tags indicative of the one or more abnormality-related durations of time for appending to the captured video or audio data streams, the one or more time-based metadata tags encapsulated into the output data structure;

modifying one or more capture characteristics of the capture of the captured video or audio data streams, the modifications of the one or more capture characteristics including at least an increase in resolution or bitrate for capturing an increased data volume or load of data relating to the medical procedure during the one or more abnormality-related durations of time indicative of the potentially elevated stress levels of the healthcare practitioner, and wherein system capturing the data streams during the medical procedure is a local computer server operating in conjunction with a centralized computer server, and wherein the computer processor is further configured to request additional bandwidth resources for transmission of the captured video or audio data streams to the centralized computer server, the increased data volume or load adapted to support generation of one or more prediction data objects representative of one or more predicted characteristics or incidents relating to the medical procedure that occur during the one or more abnormality-related duration of time; and

modifying the one or more capture characteristics of the capture of the captured video or audio data streams, the modifications of the one or more capture characteristics including at least temporarily activating one or more additional video or audio capture devices during the one or more abnormality-related durations of time.