Method and system for automatic classification of lesions in breast MRI

• US 9,218,656 B2
• Filed: 10/02/2008
• Issued: 12/22/2015
• Est. Priority Date: 10/05/2007
• Status: Active Grant

• Current Assignee: Siemens Medical Solutions USA Incorporated (Siemens AG)
• Original Assignee: Siemens Medical Solutions USA Incorporated (Siemens AG)
• Inventors: Valadez, Gerardo Hermosillo
• Primary Examiner(s): Le, Long V
• Assistant Examiner(s): MCDONALD, KATHERINE M
• Application Number:
  US12/244,049
• Publication Number:
  US 20090093711A1
• Time in Patent Office: 2,637 Days
• Field of Search: 600/407, 600/420, 600/300, 600/410, 382/128, 382/130, 382/224
• US Class Current: 1/1
• CPC Class Codes: A61B 5/055 involving electronic [EMR] or nuclear [NMR] magnetic resonance, e.g. magnetic resonance imaging
• CPC Class Codes: A61B 5/7264 Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems neural networks per se G06N3/00; expert systems per se G06N5/00
• CPC Class Codes: A61B 5/7267 involving training the classification device
• CPC Class Codes: G06T 2207/10088 Magnetic resonance imaging [MRI]
• CPC Class Codes: G06T 2207/30068 Mammography; Breast
• CPC Class Codes: G06T 7/0012 Biomedical image inspection
• CPC Class Codes: G16H 30/40 for processing medical images, e.g. editing
• CPC Class Codes: G16H 50/20 for computer-aided diagnosis, e.g. based on medical expert systems
• CPC Class Codes: G16H 50/30 for calculating health indices; for individual health risk assessment
• CPC Class Codes: G16H 50/70 for mining of medical data, e.g. analysing previous cases of other patients
• CPC Class Codes: G16Z 99/00 Subject matter not provided for in other main groups of this subclass

Abstract

A method for classifying a lesion in an MRI includes acquiring a pre-contrast MR image. A magnetic contrast agent is administered. A set of post-contrast MR images are acquired. The acquired pre-contrast and post-contrast MR images are displayed. A manually entered value is received for a size of a lesion manually identified within the pre-contrast and post-contrast MR images. A manually entered value for an absorption/washout profile is received of the manually identified lesion. A risk of malignancy for the manually identified lesion is automatically determined based on the received size value and absorption/washout profile.

Claims

1. A method for classifying a lesion in an MRI, comprising:

acquiring, by an MR imager, a pre-contrast MR image;

administering a magnetic contrast agent;

acquiring, by the MR imager, a set of post-contrast MR images;

displaying the acquired pre-contrast and post-contrast MR images;

receiving, in a computer system, a plurality of manually entered factors pertaining to a lesion manually identified within the pre-contrast and post-contrast MR images, wherein the factors pertaining to the lesion include a manually entered value for a size (s) of a lesion manually identified within the pre-contrast and post-contrast MR images, and a manually entered value for an absorption/washout profile (t) of the manually identified lesion; and

automatically determining, in the computer system, a risk of malignancy for the manually identified lesion based on the received plurality of factors pertaining to the lesion by normalizing a value of (s·

t)³;

wherein the absorption/washout profile (t) are expressed as;

1 when the lesion exhibits a progressive absorption,

2 when the lesion exhibits a plateau, and

3 when the lesion exhibits a washout.

2. The method of claim 1, wherein the acquired pre-contrast and post contrast MR images comprise a dynamic contrast enhanced MRI (DCE-MRI).

3. The method of claim 1, wherein the magnetic contrast agent includes a gadolinium compound.

4. The method of claim 1, wherein the value for the size of the lesion and the absorption/washout profile of the lesion are determined by a medical practitioner upon reading the displayed acquired pre-contrast and post-contrast MR images.

5. The method of claim 1, wherein the risk of malignancy R_c is automatically determined from the received size value (s) and absorption/washout profile (t) according to the equation:

6. The method of claim 5, wherein c is 23 millimeters.

7. The method of claim 5, wherein c is determined by a training algorithm that uses a set of training data of lesions that includes characterizations of lesion sizes, absorption/washout profiles, and determinations as to whether the lesions turned out to be benign or malignant.

8. A method for classifying a lesion in an MRI, comprising:

receiving, in a computer system, a manually entered value for a size (s) of a lesion manually identified within DCE-MR images, wherein the lesion has been diagnosed as having an uncertain probability of malignancy;

receiving, in the computer system, a manually entered value for an absorption/washout profile (t) of the manually identified lesion;

automatically determining, in the computer system, a risk of malignancy for the manually identified lesion based on the received size value and absorption/washout profile by normalizing a value of (s·

t)³; and

recharacterizing, in the computer system, the diagnosis of the lesion as either being benign or malignant based on the automatically determined risk of malignancy;

wherein the absorption/washout profile (t) are expressed as;

1 when the lesion exhibits a progressive absorption,

2 when the lesion exhibits a plateau, and

3 when the lesion exhibits a washout.

9. The method of claim 8, wherein the DCE-MR images were acquired along with the administration of a magnetic contrast agent including a gadolinium compound.

10. The method of claim 8, wherein the value for the size of the lesion and the absorption/washout profile of the lesion are determined by a medical practitioner upon reading the DCE-MR images.

11. The method of claim 8, wherein the risk of malignancy is automatically determined from the received size value (s) and absorption/washout profile (t) according to the equation:

12. The method of claim 11, wherein c is 23 millimeters.

13. The method of claim 11, wherein c is determined by a training algorithm that uses a set of training data of lesions that includes characterizations of lesion sizes, absorption/washout profiles, and determinations as to whether the lesions turned out to be benign or malignant.

14. The method of claim 8, wherein the characterization of the lesion as having an uncertain probability of malignancy includes a classification of the lesion as a BIRADS level 3 lesion and the recharacterization of the diagnosis includes either downgrading the lesion to BIRADS level 2, upgrading the lesion to BIRADS level 4 or keeping the lesion as BIRADS level 3.

15. A computer system comprising:

a processor; and

a non-transitory program storage device readable by the computer system, tangibly embodying a program of instructions executable by the processor to perform method steps for method for classifying a lesion in an MRI, the method comprising;

receiving a manually entered value for a size (s) of a lesion manually identified within DCE-MR images, wherein the lesion has been diagnosed as having an uncertain probability of malignancy;

receiving a manually entered value for an absorption/washout profile (t) of the manually identified lesion;

automatically determining a risk of malignancy for the manually identified lesion based on the received size value and absorption/washout profile by normalizing a value of (s·

t)³; and

recharacterizing the diagnosis of the lesion as either being benign or malignant based on the automatically determined risk of malignancy;

wherein the absorption/washout profile (t) are expressed as;

1 when the lesion exhibits a progressive absorption,

2 when the lesion exhibits a plateau, and

3 when the lesion exhibits a washout.

16. The computer system of claim 15, wherein the risk of malignancy R_c is automatically determined from the received size value (s) and absorption/washout profile (t) according to the equation:

17. The computer system of claim 16 wherein c is 23 millimeters.

18. The computer system of claim 16 wherein c is determined by a training algorithm that uses a set of training data of lesions that includes characterizations of lesion sizes, absorption/washout profiles, and determinations as to whether the lesions turned out to be benign or malignant.