Image analysis – Applications – Biomedical applications
Reexamination Certificate
1998-01-13
2001-01-30
Au, Amelia (Department: 2723)
Image analysis
Applications
Biomedical applications
C359S391000, C436S063000
Reexamination Certificate
active
06181811
ABSTRACT:
FIELD OF THE INVENTION
The invention relates to a method and apparatus for cytological and biological specimen screening and diagnosis, and more particularly to a method and apparatus for cytological and biological specimen screening and diagnosis based on a slide score and a field of view score from an automated specimen screener.
BACKGROUND OF THE INVENTION
Automated biological and cytological specimen screeners, such as the AutoPap® 300 available from NeoPath Inc. of Redmond, Wash., quantitatively evaluate biological and cytological samples. These automated screeners utilize computer based imaging to evaluate biological and cytological samples such as Pap smears. Automated screeners interact with medical experts, such as cytologists and pathologists, by providing useful information to assist the expert in the identification of abnormal cases and in making a final diagnosis. These systems identify potential abnormal specimens and recommend further review by human cytologists and pathologists. The cytologist or pathologist then diagnoses the potential abnormal specimen. The cytologist or pathologist provides the final diagnosis of the potentially abnormal specimen.
An abnormal sub-population comprises those specimens from a set of specimens to be screened that are abnormal. Given a set of specimens comprising normal and abnormal, two methods are commonly used by automated screeners to assist medical experts in identifying the abnormal sub-population.
The first method employs a system that screens a slide and attempts to identify suspicious objects or cells. The system attempts to identify suspicious objects or cells as those that satisfy a set of predetermined criteria of potential abnormality. Then the system acquires images for further review by human experts. The system selects a subset of objects within each specimen based on the object detection results for further image review by the cytologist or pathologist. This first method allows the expert to avoid reviewing the remaining areas of each slide if the review result of the selected images does not reveal any abnormality.
This first method has many disadvantages. For example, reviewers must review many selected images for every specimen even if the specimen is clearly normal by any measurement. Furthermore, image review of the selected images in a significant percentage of cases is inconclusive. Therefore, the reviewer must resort to examining the specimen microscopically. This first method only performs measurement and recognition of individual cells or objects and does not consider factors such as: the contextual information between objects on the same slide, the global characteristics of the slide such as the staining level and the relationship between the slides of a population under screening. This first method does not provide a test for or information regarding specimen sampling adequacy. This first method does not consider the slide to slide variation problem within a slide set to be screened.
The second method computes a slide score or specimen score. The slide or specimen score is defined as a likelihood of abnormality, based on all the quantitative measurements of the slides, including overall cell detection results, contextual information of the specimen and even the patient risk measurement of the specimen, among others. The second method then determines whether the slide should be reviewed by a human expert by applying a predetermined threshold to the score. The threshold is determined by ranking all slide scores and selecting a score value that enables a certain percentage, e.g., 50%, of higher score slides to be selected for review. The second method selects a subset of all the slides based on the scores for further human review. The goal of the selection is to select a subset of slides that contains all the abnormals of the original population. By doing so, slide review of the remaining slides is no longer needed and the remaining slides may be reported as normal.
The second method allows for effective selection of abnormal slides for review based on the slide score. To facilitate human review of the potential abnormal material on a biological specimen, what is needed is information directing a level of review needed to minimize a false negative ratio while lowering cost.
However, information is needed concerning the location of potential abnormal material residing on the specimen. Without location information, human reviewers still need to search the entire slide and may miss critical material detected by the automated device and dismiss the specimen as a false-positive case. By providing location information, a human reviewer may dismiss a normal slide after reviewing the selected “most suspicious areas” and determining the slide to be a “false-positive”. The second method directs review by a human reviewer thereby improving the sensitivity of human review and the efficiency of review process. This can significantly reduce the workload of the human reviewer.
SUMMARY OF THE INVENTION
The invention provides a method and apparatus for optimizing biological and cytological specimen screening and diagnosis. The invention recommends a slide review process for cytological specimen screening to identify abnormal sub-populations for further review and also diagnosis by a human expert. The invention processes a cytological specimen with an automated screener. Using a slide score generated by the automated screener, the invention determines a slide review process using a slide score classification. The method of the invention improves the overall performance of the screening process as measured by sensitivity to abnormal specimens, and at the same time reduces the work load of the human reviewer. The method of the invention also effectively and smoothly integrates the process of initial screening of the specimen with the process of further review of the specimen and final diagnosis of the specimen.
REFERENCES:
patent: 3699336 (1972-10-01), Ehrlich et al.
patent: 3824393 (1974-07-01), Brain
patent: 3916176 (1975-10-01), Alien et al.
patent: 4045655 (1977-08-01), Suzuki et al.
patent: 4097845 (1978-06-01), Bacus
patent: 4129854 (1978-12-01), Suzuki et al.
patent: 4150360 (1979-04-01), Kopp et al.
patent: 4175860 (1979-11-01), Bacus
patent: 4210419 (1980-07-01), Castleman
patent: 4213036 (1980-07-01), Kopp et al.
patent: 4296373 (1981-10-01), Angel et al.
patent: 4513438 (1985-04-01), Graham et al.
patent: 4523278 (1985-06-01), Reinhardt et al.
patent: 4596464 (1986-06-01), Hoffman et al.
patent: 4672559 (1987-06-01), Jansson et al.
patent: 4690521 (1987-09-01), Saccomanno
patent: 4700298 (1987-10-01), Palcic et al.
patent: 4741043 (1988-04-01), Bacus
patent: 4761075 (1988-08-01), Matsushita et al.
patent: 4807979 (1989-02-01), Saccomanno et al.
patent: 4833382 (1989-05-01), Gibbs
patent: 4907158 (1990-03-01), Kettler et al.
patent: 4965725 (1990-10-01), Rutenberg
patent: 5000554 (1991-03-01), Gibbs
patent: 5031099 (1991-07-01), Kettler
patent: 5072382 (1991-12-01), Kamentsky
patent: 5073857 (1991-12-01), Peters et al.
patent: 5086476 (1992-02-01), Bacus
patent: 5172418 (1992-12-01), Ito et al.
patent: 5218645 (1993-06-01), Bacus
patent: 5235522 (1993-08-01), Bacus
patent: 5257182 (1993-10-01), Luck et al.
patent: 5268966 (1993-12-01), Kasdan
patent: 5287272 (1994-02-01), Rutenberg et al.
patent: 5313532 (1994-05-01), Harvey et al.
patent: 5315700 (1994-05-01), Johnston et al.
patent: 5352613 (1994-10-01), Tafas et al.
patent: 5361140 (1994-11-01), Hayenga et al.
patent: 5428690 (1995-06-01), Bacus et al.
patent: 5473706 (1995-12-01), Bacus et al.
patent: 5544650 (1996-08-01), Boon et al.
patent: 5561556 (1996-10-01), Weissman
patent: 5574594 (1996-11-01), Fowler et al.
patent: 5581487 (1996-12-01), Kelly et al.
patent: 5602674 (1997-02-01), Weissmann et al.
patent: 5627442 (1997-05-01), Engelse et al.
patent: 5655029 (1997-08-01), Rutenberg et al.
patent: 5677966 (1997-10-01), Doerrer et al.
patent: 5715082 (1998-02-01), Saccomanno et al.
patent: 5740270 (1998-04-01), Rutenberg et al.
patent: 5786130 (1998-07-
Ellison Dayle G.
Kuan Chih-Chau L.
Lee Shih-Jong J.
Nelson Larry A.
Oh Seho
Au Amelia
Dastouri Mehrdad
Leone George
Moffa Emil
NeoPath, Inc.
LandOfFree
Method and apparatus for optimizing biological and... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus for optimizing biological and..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for optimizing biological and... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2554319