Selecting a region of interest for extracting physiological parameters from a video of a subject

• US 9,986,923 B2
• Filed: 01/09/2015
• Issued: 06/05/2018
• Est. Priority Date: 01/09/2015
• Status: Active Grant

• Current Assignee: Xerox Corporation (Xerox Holdings Corp.)
• Original Assignee: Xerox Corporation (Xerox Holdings Corp.)
• Inventors: Mestha, Lalit Keshav, Hoover, Martin Edward, Kyal, Survi
• Primary Examiner(s): Fernandez, Katherine
• Assistant Examiner(s): Yang, Yi-Shan
• Application Number:
  US14/593,271
• Publication Number:
  US 20160198965A1
• Time in Patent Office: 1,243 Days
• Field of Search: None
• CPC Class Codes: A61B 2576/00 Medical imaging apparatus involving image processing or analysis A61B1/00009, A61B6/52 and A61B8/52 take precedence; image processing per se G06T; details of algorithms for analysing medical images G06T7/0012
• CPC Class Codes: A61B 5/0013 Medical image data A61B1/00011, A61B6/56, A61B8/56 take precedence
• CPC Class Codes: A61B 5/02433 for infrared radiation
• CPC Class Codes: A61B 5/7203 for noise prevention, reduction or removal
• 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 5/725 using specific filters therefor, e.g. Kalman or adaptive filters A61B5/316 take precedence
• CPC Class Codes: A61B 5/748 Selection of a region of interest, e.g. using a graphics tablet
• CPC Class Codes: G16H 30/40 for processing medical images, e.g. editing

Abstract

What is disclosed is a system and method for selecting a region of interest for extracting physiological parameters from a video of a subject. In one embodiment the present method involves performing the following. First, time-series signals are received which have been generated by having processing image frames of a video of a subject captured using a single band video camera with a bandpass filter with a pass band in a wavelength range of 495-565 nm and/or 800-1000 nm. The regions of interest are areas where a plethysmographic signal can be detected by the camera. Each time-series signal is associated with a different region of interest. A signal strength is then calculated for each of the time-series signals. The region associated with the time-series signal having a highest signal strength is selected. The time-series signal associated with the selected region can be processed to extract a videoplethysmographic (VPG) signal containing physiological parameters.

Claims

1. A system for selecting a region of exposed skin to extract at least one physiological parameter from a video of a subject, the system comprising:

a single band video camera with a bandpass filter for acquiring video of the subject, the video comprising a plurality of time-sequential image frames of an area of exposed skin of the subject, each of the time-sequential image frames comprising a plurality of pixels, the bandpass filter having a pass band in a wavelength range of 495-565 nm and/or 800-1000 nm; and

a processor executing machine readable instructions which configure the processor to;

analyze the time-sequential image frames to identify a plurality of regions of exposed skin of the subject where a plethysmographic signal can be detected;

for each of the plurality of regions of exposed skin;

process the pixels to obtain an associated time-series signal corresponding to relative blood volume changes in blood vessels close to a surface of the exposed skin within the region of interest;

for each of the obtained time-series signals;

compute a power spectral density of the time-series signal;

identify a dominant frequency;

identify harmonic frequencies of the dominant frequency;

define a window of frequencies about the dominant frequency, frequencies within the window collectively defining a band of interest; and

calculate a signal strength for the time-series signal over the band of interest;

determine which of the time-series signals has a highest signal strength;

select, automatically, one of the plurality of regions of exposed skin of the subject associated with the time-series signals having the highest signal strength; and

extract at least one physiological parameter from the selected one of the plurality of regions of exposed skin.

2. The system of claim 1, wherein the power spectral density is computed by at least one of:

a non-parametric spectral density estimation on a filtered time-series signal, and a parametric spectral density estimation on a filtered time-series signal.

3. The system of claim 1, wherein the pulse harmonic strength (PHS) at least comprises:

4. The system of claim 1, wherein the processor is further configured to detrend at least one of the time-series signals to remove non-stationary components.

5. The system of claim 1, further comprises determining whether a subject is having an atrial fibrillation event by

defining a size N of a batch of image frames such that N_min≤

N≤

N_max, where N_min is a minimum size of a batch of image frames and N_max is a maximum size of a batch of image frames; and

for each batch of image frames of size N;

processing this batch of image frames to isolate pixels associated with said region of exposed skin;

processing said isolated pixels to obtain a time-series signal for this batch;

extracting a VPG signal from said time-series signal;

computing a power spectral density across all frequencies within said VPG signal;

calculating a pulse harmonic strength for this VPG signal;

comparing the pulse harmonic strength (PHS) to a discrimination threshold obtained from a Receiver Operating Characteristic (ROC) curve constructed from health vitals obtained for a group of cardiac patients, said discrimination threshold discriminating between low PHS values and high PHS values; and

classifying, as a result of said comparison, that said subject is one of;

in atrial fibrillation, and in sinus rhythm, wherein when PHS is below said threshold said subject is classified as being in atrial fibrillation, and in normal sinus rhythm otherwise.