Method and system for intelligent linking of medical data

Patent metadata

Field	Value
Publication number	US-9370305-B2
Application number	US-201214005399-A
Country	US
Kind code	B2
Filing date	Feb 13, 2012
Priority date	Mar 16, 2011
Publication date	Jun 21, 2016
Grant date	Jun 21, 2016

Abstract

Described herein are systems and methods for intelligently combining medical findings received across different modalities. The system comprises an extraction module extracting contextual information from an image of an area of interest including annotations, a feature selection module building a current feature vector using the extracted contextual information and the annotations, and a referencing engine computing a similarity score between the current feature vector and a prior feature vector of a prior image. The method comprises extracting contextual information from an image of an area of interest including annotations, building a current feature vector using the extracted contextual information and the annotations, and computing a similarity score between the current feature vector and a prior feature vector of a prior image.

First claim

What is claimed is:

1. A system, comprising:
an extraction module extracting contextual information from an image of an area of interest of a patient including annotations;
a feature selection module building a current feature vector using the extracted contextual information and the annotations;
a referencing engine computing a similarity score between the current feature vector and a prior feature vector of a prior image of the patient; and
a presentation engine providing a display based on the similarity score,
wherein the similarity score (D) is computed as a weighted (m i ) sum of distances (d i ) between each element of the current feature vector and the prior feature vector, wherein:
D=Σm i *d i ; and

wherein distances (d i ) of annotation features (a i ) and (b i ), and contextual features (a′ i ) and (b′ i ) are computed as:
d i =e a i −b i *e a′ i −b′ i .

2. The system of claim 1 , wherein the presentation engine displays matching annotations between the image and the prior image.

3. The system of claim 1 , wherein the presentation engine receives input from a user via a user interface and adjusts the image based on the input.

4. The system of claim 3 , wherein the input received from the user includes at least one of an image selection, an adjusted detail, an assessment, and a request for additional information.

5. The system of claim 1 , wherein the annotations include at least one of an image modality reading, a laterality reading, a depth reading, a location reading, a type reading, a shape reading, and a measurement reading.

6. The system of claim 1 , wherein the contextual data includes at least one of a relative distance to nipple, a relative distance to skin surface, a relative distance to lesion, a volume estimate, a type likelihood, and a shape likelihood.

7. The system of claim 1 , wherein the area of interest is a lesion from a breast cancer patient.

8. The system of claim 1 , wherein the image and the prior image are of the same area of interest using different imaging protocols, and wherein the imaging protocols includes at least one of ultrasonic imaging, magnetic resonance imaging (“MRI”) imaging, and mammography (“MG”) imaging.

9. A method, comprising:
extracting, by an extraction module, contextual information from an image of an area of interest of a patient including annotations;
building, a feature selection module, a current feature vector using the extracted contextual information and the annotations;
computing, by a referencing engine, a similarity score between the current feature vector and a prior feature vector of a prior image of the patient, wherein the similarity score (D) is computed as a weighted (m i ) sum of distances (d i ) between each element of the current feature vector and the prior feature vector; and
displaying, by a presentation engine, a presentation based on the similarity score, wherein:
D=Σm i *d i ; and

wherein distances (d i ) of annotation features (a i ) and (b i ), and contextual features (a′ i ) and (b′ i ) are computed as:
d i =e a i −b i *e a′ i −b′ i .

10. The method of claim 9 , further comprising:
displaying, by a presentation engine, matching annotations between the image and the prior image.

11. The method of claim 9 , further comprising:
receiving input from a user via a user interface; and
adjusting the image based on the input.

12. The method of claim 11 , wherein the input received from the user includes at least one of an image selection, an adjusted detail, an assessment, and a request for additional information.

13. The method of claim 9 , wherein the annotations include at least one of an image modality reading, a laterality reading, a depth reading, a location reading, a type reading, a shape reading, and a measurement reading.

14. The method of claim 9 , wherein the contextual data includes at least one of a relative distance to nipple, a relative distance to skin surface, a relative distance to lesion, a volume estimate, a type likelihood, and a shape likelihood.

15. The method of claim 9 , wherein the area of interest is a lesion from a breast cancer patient.

16. The method of claim 9 , wherein the image and the prior image are of the same area of interest using different imaging protocols, and wherein the imaging protocols includes at least one of ultrasonic imaging, magnetic resonance imaging (“MRI”) imaging, and mammography (“MG”) imaging.

17. A non-transitory computer readable storage medium including a set of instructions that are executable by a processor, the set of instructions being operable at least to:
extract contextual information from an image of an area of interest of a patient including annotations;
build a current feature vector using the extracted contextual information and the annotations; and
compute a similarity score between the current feature vector and a prior feature vector of a prior image of the patient, wherein the similarity score (D) is computed as a weighted (m i ) sum of distances (d i ) between each element of the current feature vector and the prior feature vector;
provide a display presentation based on the similarity score (D), wherein:
D=Σm i *d i ; and

wherein distances (d i ) of annotation features (a i ) and (b i ), and contextual features (a′ i ) and (b′ i ) are computed as:
d i =e a i −b i *e a′ i −b′ i .

18. The non-transitory computer readable storage medium of claim 17 , wherein the set of instructions are further operable to:
display matching annotations between the image and the prior image;
receive input from a user via a user interface; and
adjust the image based on the input.

19. The non-transitory computer readable storage medium of claim 17 , wherein the annotations include at least one of an image modality reading, a laterality reading, a depth reading, a location reading, a type reading, a shape reading, and a measurement reading.

20. The non-transitory computer readable storage medium of claim 17 , wherein the image and the prior image are of the same area of interest using different imaging protocols, and wherein the imaging protocols includes at least one of ultrasonic imaging, magnetic resonance imaging (“MRI”) imaging, and mammography (“MG”) imaging.

Assignees

• Qian Yuechen
• Sevenster Merlijn
• Koninkl Philips Nv

Inventors

• Qian Yuechen
• Sevenster Merlijn

Classifications

• G06T2207/10116

  X-ray image · CPC title

• G06T7/0014

  using an image reference approach · CPC title

• G06T2207/10136

  3D ultrasound image · CPC title

• G06T2207/30068

  Mammography; Breast · CPC title

• G06T2207/10088

  Magnetic resonance imaging [MRI] · CPC title