Abnormal pattern detection processing method and system

• US 7,162,061 B1
• Filed: 01/24/2000
• Issued: 01/09/2007
• Est. Priority Date: 01/22/1999
• Status: Expired due to Fees

• Current Assignee: Fujifilm Corporation (Fujifilm Holdings Corporation)
• Original Assignee: Fuji Photo Film Co Limited (Fujifilm Holdings Corporation)
• Inventors: Takeo, Hideya
• Primary Examiner(s): Mehta; Bhavesh M.
• Assistant Examiner(s): Edwards; Patrick L.
• Application Number:
  US09/489,846
• Time in Patent Office: 2,542 Days
• Field of Search: 382/128, 382/131, 382/132, 378/37, 128/922
• US Class Current: 382/128
• CPC Class Codes: G06T 2207/30004 Biomedical image processing
• CPC Class Codes: G06T 7/0012 Biomedical image inspection

Abstract

The result of detection processing by abnormal pattern detection processor means 30 is stored in a server 40. In addition, the output image on an image output device 440 is read for pattern, and the supporting contents, such as the result of judgment by a doctor (which, when differing from the result of detection processing, corresponds to the corrected result), the comment, and the assessment category, are stored in the server 40 in conformity with the BI-RADS proposed by the ACR. The result of pathologic assessment corresponding to the abnormal pattern is related to the supporting contents and stored in the server 40. The statistical processing in conformity with the follow-up method of the BI-RADS is performed to find the specificity, the cancer detection rate, etc.

Claims

1. An abnormal pattern detection processing method comprising:

detecting an abnormal pattern in an image, based on inputted image information;

processing the detected abnormal pattern;

displaying the detected abnormal pattern;

correcting the processed and displayed abnormal pattern, for each of a plurality of items of the inputted image information;

relating a result of the processed and displayed abnormal pattern to a result of the corrected abnormal pattern, for each of the plurality of items of the inputted image information, for each patient; and

storing the plurality of processed abnormal pattern results and the plurality of corrected abnormal pattern results for each patient, wherein

said correcting step includes canceling a pattern which is not abnormal within the processed and displayed abnormal pattern, and adding an abnormal pattern which was not detected by the abnormal pattern detection; and

said relating step includes determining a cancelled abnormal pattern as an FP, a non-cancelled abnormal pattern as a TP, and an added abnormal pattern as an FN.

2. An abnormal pattern detection processing method according to claim 1, wherein quantitative evaluation of the detection processing is performed, on the basis of said stored plurality of processed abnormal pattern results and said stored plurality of corrected abnormal pattern results.

3. The method of claim 2, wherein the corrected abnormal pattern results comprise a determination of whether the processed abnormal pattern corresponds to at least one of a true positive, false positive, true negative and false negative.

4. The method of claim 3, wherein the quantitative evaluation comprises a ratio of a number of true results relative to a number of true and false results.

5. The method of claim 4, wherein a sensitivity of the quantitative evaluation is determined by the ratio of true positives to a sum of true positives and false negatives.

6. The method of claim 4, wherein a specificity of the quantitative evaluation is determined by the ratio of true negatives to a sum of true negatives and false positives.

7. The method of claim 4, wherein a positive predictive value of the quantitative evaluation is determined by the ratio of true positives to a sum of true positives and false positives.

8. An abnormal pattern detection processing method according to claim 1, wherein the processing automatically determines whether the abnormal pattern exists or not, based on a result of the detection of said abnormal pattern in said image.