System and method for estimating extracorporeal blood volume in a physical sample

• US 9,652,655 B2
• Filed: 07/09/2012
• Issued: 05/16/2017
• Est. Priority Date: 07/09/2011
• Status: Active Grant

• Current Assignee: Gauss Surgical, Gauss Surgical, Inc. (Stryker Corporation)
• Original Assignee: Gauss Surgical, Inc. (Stryker Corporation)
• Inventors: Satish, Siddarth, Zandifar, Ali, Miller, Kevin
• Primary Examiner(s): Newman, Michael A.
• Application Number:
  US13/544,664
• Publication Number:
  US 20130010094A1
• Time in Patent Office: 1,772 Days
• Field of Search: 382128, 382134, 382171, 382209, 382190, 604404
• CPC Class Codes: A61B 5/02042 Determining blood loss or bleeding, e.g. during a surgical procedure
• CPC Class Codes: G06F 18/00 Pattern recognition
• CPC Class Codes: G06T 2207/10024 Color image
• CPC Class Codes: G06T 2207/30004 Biomedical image processing
• CPC Class Codes: G06T 5/00 Image enhancement or restoration
• CPC Class Codes: G06T 7/62 of area, perimeter, diameter or volume

Abstract

One method for estimating the extracorporeal blood volume in a portion of a physical sample includes: extracting a feature from a portion of an image of the sample; tagging the portion of the image of the sample with a blood volume indicator according to the extracted feature; and estimating the extracorporeal blood volume in at least the portion of the physical sample, associated with the portion of the image of the sample, according to the blood volume indicator.

Claims

1. A method for estimating the extracorporeal blood volume in a portion of a physical sample, the method comprising:

extracting a feature from a portion of an image of the sample;

tagging the portion of the image of the sample with a blood volume indicator according to the extracted feature; and

estimating the extracorporeal blood volume in at least the portion of the physical sample associated with the portion of the image of the sample, according to the blood volume indicator;

wherein extracting the feature from the portion of the image of the sample comprises extracting the feature that comprises a color intensity value in a component space; and

wherein tagging the portion of the image of the sample with the blood volume indicator comprises transforming the extracted feature into the blood volume indicator according to a parametric model.

6. The method of claim 1, wherein estimating the extracorporeal blood volume in at least the portion of the sample comprises estimating the total extracorporeal blood volume in the physical sample based upon aggregated blood volume indicator tags of the portion of the image of the sample and substantially all other portions of the image.

7. The method of claim 6, further comprising estimating patient blood loss by aggregating the estimated total extracorporeal blood volume in the physical sample and estimated total extracorporeal blood volumes in additional physical samples in additional images.

8. The method of claim 6, further comprising estimating a volume of non-blood fluid in the physical sample by subtracting the weight of the estimated total extracorporeal blood volume in the physical sample from an entered weight of the physical sample.

33. The method of claim 1, further comprising identifying the physical sample in the image as a type of absorbent gauze sponge.

34. The method of claim 33, further comprising indexing a sample count for the physical sample that is identified as an absorbent gauze sponge.

39. The method of claim 1, wherein tagging the portion of the image of the sample with the blood volume indicator comprises tagging the portion of the image of the sample with an estimated hemoglobin mass.

40. The method of claim 39, wherein estimating the extracorporeal blood volume in at least the portion of the physical sample comprises estimating the extracorporeal blood volume in at least the portion of the physical sample according to the hemoglobin mass and an estimated hematocrit of the blood in the physical sample.

2. A method for estimating the extracorporeal blood volume in a portion of a physical sample, the method comprising:

extracting a feature from a portion of an image of the sample;

tagging the portion of the image of the sample with a blood volume indicator according to the extracted feature; and

estimating the extracorporeal blood volume in at least the portion of the physical sample associated with the portion of the image of the sample, according to the blood volume indicator;

wherein estimating the extracorporeal blood volume in at least a portion of the physical sample comprises associating a physical dimension of the physical sample with the portion of the image of the sample by transforming the portion of the image according to at least one of an estimated distance and an estimated angle between the capture origin of the image and the physical sample.

3. The method of claim 2, wherein extracting the feature from the portion of the image of the sample comprises extracting a feature selected from the group consisting of a redness value in the red component space, a blueness value in the blue component space, a greenness value in the green component space, a surface area of the physical sample, a surface area of a bloodied region of the physical sample, a pixel count of the portion of the image associated with the portion of the physical sample, and a color intensity value of the portion of the image associated with an unsoiled portion of the sample.

9. The method of claim 2, further comprising capturing the image through an optical sensor incorporated into a mobile electronic device.

10. The method of claim 9, further comprising displaying, on a display incorporated into the mobile electronic device, the estimated extracorporeal blood volume in at least the portion of the physical sample.

11. The method of claim 2, further comprising updating a digital medical record of a patient with the estimated blood volume in at least the portion of the physical sample.

18. The method of claim 2, further comprising identifying the physical sample in the image as a type of absorbent gauze sponge.

19. The method of claim 18, further comprising indexing a sample count for the physical sample that is identified as an absorbent gauze sponge.

20. The method of claim 2 wherein tagging the portion of the image of the sample with the blood volume indicator comprises tagging the portion of the image of the sample with an estimated hemoglobin mass.

21. The method of claim 20, wherein estimating the extracorporeal blood volume in at least the portion of the physical sample comprises estimating the extracorporeal blood volume in at least the portion of the physical sample according to the hemoglobin mass and an estimated hematocrit of the blood in the physical sample.

22. The method of claim 2, wherein estimating the extracorporeal blood volume in at least the portion of the sample comprises estimating the total extracorporeal blood volume in the physical sample based upon aggregated blood volume indicator tags of the portion of the image of the sample and substantially all other portions of the image.

23. The method of claim 22, further comprising estimating patient blood loss by aggregating the estimated total extracorporeal blood volume in the physical sample and estimated total extracorporeal blood volumes in additional physical samples in additional images.

24. The method of claim 22, further comprising estimating a volume of non-blood fluid in the physical sample by subtracting the weight of the estimated total extracorporeal blood volume in the physical sample from an entered weight of the physical sample.

35. The method of claim 2, wherein tagging the portion of the image of the sample with the blood volume indicator comprises comparing the portion of the image of the sample with a template image of known extracorporeal blood volume indicator and associating the portion of the image of the sample with the blood volume indicator that is correlated with the known extracorporeal blood volume indicator of the template image.

36. The method of claim 35, wherein comparing the portion of the image of the sample with the template image comprises pairing the portion of the image of the sample and the template image via template matching.

37. The method of claim 36, wherein comparing the portion of the image of the sample with the template image comprises selecting the template image from a library of template images of known extracorporeal blood volume indicators.

38. The method of claim 36, wherein comparing the portion of the image of the sample with the template image comprises calculating the absolute difference in pixel intensity for a plurality of pixels in each of the portion of the image of the sample and the template image.

41. The method of claim 2, wherein extracting the feature from the portion of the image of the sample comprises extracting the feature that comprises a color intensity value in a component space, and wherein tagging the portion of the image of the sample with the blood volume indicator comprises comparing the color intensity value of the portion of the image of the sample with intensity values of color models paired with known blood volume indicators.

42. The method of claim 41, wherein tagging the portion of the image of the sample with the blood volume indicator comprises generating the blood volume indicator tag that is a composite of known blood volume indicators of a plurality of color models.

4. A method for estimating the extracorporeal blood volume in a portion of a physical sample, the method comprising:

extracting a feature from a portion of an image of the sample;

tagging the portion of the image of the sample with a blood volume indicator according to the extracted feature; and

estimating the extracorporeal blood volume in at least the portion of the physical sample associated with the portion of the image of the sample, according to the blood volume indicator;

wherein tagging the portion of the image with the blood volume indicator comprises classifying the physical sample according to a non-image feature, selecting a blood volume indicator model based upon the classification of the sample, and generating the blood volume indicator tag by passing the extracted feature into the selected blood volume indicator model.

5. The method of claim 4, wherein the non-image feature is selected from the group consisting of current patient intravascular hematocrit, estimated patient intravascular hematocrit, historic patient intravascular hematocrit, weight of the sample, clinician-estimated sample blood volume, an ambient lighting condition, type of the sample, properties of the sample, a patient vital sign, patient medical history, an identity of a surgeon, and a type of surgery.

25. The method of claim 4, wherein extracting the feature from the portion of the image of the sample comprises extracting the feature that comprises a color intensity value in a component space.

26. The method of claim 4, further comprising identifying the physical sample in the image as a type of absorbent gauze sponge.

27. The method of claim 26, further comprising indexing a sample count for the physical sample that is identified as an absorbent gauze sponge.

28. The method of claim 4 wherein tagging the portion of the image of the sample with the blood volume indicator comprises tagging the portion of the image of the sample with an estimated hemoglobin mass.

29. The method of claim 28, wherein estimating the extracorporeal blood volume in at least the portion of the physical sample comprises estimating the extracorporeal blood volume in at least the portion of the physical sample according to the hemoglobin mass and an estimated hematocrit of the blood in the physical sample.

30. The method of claim 4, wherein estimating the extracorporeal blood volume in at least the portion of the sample comprises estimating the total extracorporeal blood volume in the physical sample based upon aggregated blood volume indicator tags of the portion of the image of the sample and substantially all other portions of the image.

31. The method of claim 30, further comprising estimating patient blood loss by aggregating the estimated total extracorporeal blood volume in the physical sample and estimated total extracorporeal blood volumes in additional physical samples in additional images.

32. The method of claim 30, further comprising estimating a volume of non-blood fluid in the physical sample by subtracting the weight of the estimated total extracorporeal blood volume in the physical sample from an entered weight of the physical sample.

12. A system for determining the extracorporeal blood volume in a physical sample, the system comprising:

an optical sensor that captures an image of the physical sample;

a processor that extracts a feature from a portion of an image of the sample, tags the portion of the image of the sample with a blood volume indicator according to the extracted feature, and estimates the extracorporeal blood volume in at least a portion of the physical sample, identified in the portion of the image of the sample, according to the blood volume indicator; and

a display that depicts the estimated blood volume in at least the portion of the physical sample.

13. The system of claim 12, wherein the optical sensor is a camera, and wherein the processor further determines at least one of the angle of the camera relative to the physical sample and the distance between the camera and the physical sample at the time of the image was captured.

14. The system of claim 12, wherein the optical sensor is configured to be mounted overhead a surgical operating table.

15. The system of claim 12, further comprising a handheld housing that contains the optical sensor, the processor, and the display.

16. The system of claim 12, further comprising a wireless communication module that communicates the estimated blood volume in the portion of the physical sample to a remote server that stores an electronic medical record of a patient.

17. The system of claim 12, further comprising a data storage module configured to store a library of template images of known blood volume indicators, wherein the processor is configured to access a template image from the data storage module and to compare the template image with the portion of the image of the sample according to the extracted feature to generate the blood volume indicator tag.