Hierarchical modeling in medical abnormality detection
원문보기
IPC분류정보
국가/구분
United States(US) Patent
등록
국제특허분류(IPC7판)
G06K-009/00
A61B-006/00
출원번호
UP-0054600
(2005-02-08)
등록번호
US-7653227
(2010-02-24)
발명자
/ 주소
Krishnan, Sriram
Bi, Jinbo
Rao, R. Bharat
출원인 / 주소
Siemens Medical Solutions USA, Inc.
인용정보
피인용 횟수 :
14인용 특허 :
15
초록▼
Hierarchal modeling is used to distinguish one state or class from three or more classes. In a first stage, a normal or other class is distinguished from a diseased or other groups of classes. If the results of the first stage classification indicate diseased or data within the groups of different c
Hierarchal modeling is used to distinguish one state or class from three or more classes. In a first stage, a normal or other class is distinguished from a diseased or other groups of classes. If the results of the first stage classification indicate diseased or data within the groups of different classes, a subsequent stage of classification is performed. In a subsequent stage of classification, the data is classified to distinguish one or more other classes from the remaining classes. Using two or more stages, medical information is classified by eliminating one or more possible classes in each stage to finally identify a particular class most appropriate or probable for the data.
대표청구항▼
We claim: 1. A method for modeling in medical abnormality detection, the method comprising: obtaining medical data representing characteristics of a patient and corresponding to one of at least five possible ranked states, the at least five possible ranked states being a first, a second, fourth and
We claim: 1. A method for modeling in medical abnormality detection, the method comprising: obtaining medical data representing characteristics of a patient and corresponding to one of at least five possible ranked states, the at least five possible ranked states being a first, a second, fourth and fifth state; hierarchically classifying with a processor the medical data between the first state and a group of at least the second, third, fourth and fifth states; hierarchically classifying with the processor the medical data between the second state and a group of at least the third, fourth and fifth states; hierarchically classifying with the processor the medical data between the third state and a group of at least the fourth and fifth states; hierarchically classifying with the processor the medical data between the fourth state and at least the fifth state; and displaying an output indicating an actual state, determined from the classifying acts, of the patient. 2. The method of claim 1 wherein the at least five possible ranked states comprise five cardiac wall motion states, the first state being a normal state. 3. The method of claim 2 wherein classifying between the first state and the group of at least the second, third, fourth and fifth state comprises classifying between the normal state and all disease states, the disease states including the second and third states of the five cardiac wall motion states, and wherein classifying between the second and a group of at least the third, fourth and fifth states comprises classifying between one or more of the disease states and one or more different ones of the disease states, the disease states comprising hypokinesia, akinesia, dyskinesia, and aneurysm. 4. The method of claim 2 wherein classifying between the first state and the group of at least the second, third, fourth and fifth states comprises classifying between the normal state and all disease states, the disease states including the second and third states of the five cardiac wall motion states, and wherein classifying between the second state and a group of at least the third, fourth and fifth states comprises classifying between the second state and three other disease states; further comprising: classifying between the third state and two other disease states; and classifying between the fourth state and the fifth state. 5. The method of claim 1 wherein classifying between the first state and the group of at least the second, third, fourth and fifth states is performed before classifying between the second state and a group of at least the third, fourth and fifth states, classifying between the first state and the group of at least the second, third, fourth and fifth states being operable to rule out the first state from possible states. 6. The method of claim 1 wherein classifying between the first state and the group of at least the second, third, fourth and fifth states comprises classifying between (a) another group comprising the first state and a fourth state and (b) the group of at least the second, third, fourth and fifth states. 7. The method of claim 1 wherein a number of classifying acts is one less than a total number of possible ranked states. 8. The method of claim 1 wherein classifying between the first state and the group of at least the second, third, fourth and fifth states is performed as a function of a different process or parameter than classifying between the second state and a group of at least the third, fourth and fifth states. 9. A system for modeling in medical abnormality detection, the method comprising: a memory operable to store medical data representing one of at least three possible ranked states; a processor operable to apply to the medical data a first classifier in a hierarchal model, the first classifier operable to distinguish between first and second groups of states of the at least three possible ranked states, and the processor operable to apply to the medical data a second classifier in the hierarchal model, the second classifier operable to distinguish between third and fourth groups of states of the at least three possible states, the third and fourth groups being sub-sets of the second group of states and each being free of states of the first group of states; and a display for displaying an output indicating an actual state, determined from the classifiers, of the patient. 10. The system of claim 9 wherein the at least three possible ranked states comprise cardiac wall motion states. 11. The system of claim 9 wherein the first group comprises a normal state, the second group comprises at least first and second disease states, the third group comprises at least the first disease state and the fourth group comprises at least the second disease state. 12. The system of claim 9 wherein the processor is operable to apply the first classifier before the second classifier and operable to apply at least a third classifier in the hierarchal model after the second classifier. 13. The system of claim 9 wherein the first group consists of a first state and the second group comprises the third and fourth groups. 14. The system of claim 9 wherein the first classifier is different than the second classifier. 15. A method for detecting a medical abnormality, the method comprising: applying with a processor a hierarchal model of at least four classifiers to medical data representing a patient, the first classifier operable to distinguish between a normal state and disease states, the second classifier operable to distinguish between a first disease state and at least a second disease state, the third classifier operable to distinguish between the second disease state and at least a third disease state, the fourth classifier operable to distinguish between the third disease state and at least a fourth disease state; identifying which of the normal state, first disease state, second disease state, third disease state, and fourth disease state is represented by the medical data as a function of the applying; and outputting the identified slate of the patient. 16. The method of claim 15 wherein applying comprises distinguishing between the normal state, the first disease state comprising a hypokinesia state, and the second disease state comprising an akinesia state, the distinguishing being performed with sequential application of the first and second classifiers. 17. The method of claim 15 wherein the first classifier is operable to distinguish between (a) the normal state and (b) all disease states, the disease states including hypokinesia, akinesia, dyskinesia, and aneurysm states, the second classifier operable to distinguish between (a) the hypokinesia state and (b) akinesia, dyskinesia, and aneurysm states; wherein the third classifier is operable to distinguish between (a) the akinesia state and (b) dyskinesia mad aneurysm states, and the fourth classifier is operable to distinguish between (a) the dyskinesia state and (b) the aneurysm state; and wherein the first, second, third and fourth classifiers are applied in sequential order, the later occurring applications only being performed where the previous applications indicated (b) states. 18. The method of claim 15 wherein applying the hierarchal model comprises applying the first classifier prior to the second classifier, and applying the second classifier only if the first classifier indicates the medical data to not represent a normal state. 19. In a computer readable storage media having stored therein data representing instructions executable by a programmed processor for detecting medical abnormality, the storage media comprising instructions for: applying a hierarchal model of at least four classifiers to medical data, the first classifier operable to distinguish between a normal state and disease states, the second classifier operable to distinguish between a first disease state and at least a second disease state, the third classifier operable to distinguish between the second disease state and at least a third disease state, the fourth classifier operable to distinguish between the third disease state and at least a fourth disease state; and identifying which of the normal state, first disease state, second disease state, third disease state, and fourth disease state is represented by the medical data as a function of the applying. 20. The instructions of claim 19 wherein applying comprises distinguishing between the normal state, the first disease state comprising a hypokinesia state, and the second disease state comprising a akinesia state, the distinguishing being performed with sequential application of the first and second classifiers. 21. The instructions of claim 19 wherein the first classifier is operable to distinguish between (a) the normal state and (b) all disease states, the disease states including hypokinesia, akinesia, dyskinesia, and aneurysm states, the second classifier operable to distinguish between (a) the hypokinesia state and (b) akinesia, dyskinesia, and aneurysm states; wherein the third classifier is operable to distinguish between (a) the akinesia state and (b) dyskinesia and aneurysm states, and the fourth classifier is operable to distinguish between (a) the dyskinesia state and (b) the aneurysm state; and wherein the first, second, third and fourth classifiers are applied in sequential order, the later occurring applications only being performed where the previous applications indicated (b) states. 22. The instructions of claim 19 wherein applying the hierarchal model comprises applying the first classifier prior to the second classifier, and applying the second classifier only if the first classifier indicates the medical data to not represent a normal state.
연구과제 타임라인
LOADING...
LOADING...
LOADING...
LOADING...
LOADING...
이 특허에 인용된 특허 (15)
Barnhill Stephen D. ; Zhang Zhen, Computer assisted methods for diagnosing diseases.
Hans Torp NO; Bjorn Olstad NO; Andreas Heimdal NO; Steinar Bjaerum NO, Method and apparatus for providing real-time calculation and display of tissue deformation in ultrasound imaging.
Robergé, James; Soble, Jeffrey; Donati, Alessandro Davide, Method and system for generation of medical reports from data in a hierarchically-organized database.
Armstrong E. Glenn ; Petry Christoph ; Wu Alan ; Wagner Gerald, Reflex algorithm for early and cost effective diagnosis of myocardial infractions suitable for automated diagnostic platforms.
DelloStritto, James J.; Craw, Chad E.; Chung, Song Y.; Blaszak, Ronald J.; LoMascolo, Frank, Application generator for a dynamic medical object information base.
DelloStritto, James J.; Blaszak, Ronald J.; Craw, Chad E.; Gondek, Cory Russell; LoMascolo, Frank; Petersen, Eric G.; West, Kenneth G.; Goldfain, Albert; Narayan, Mahesh; Chung, Song Y., Communication of information between a plurality of network elements.
DelloStritto, James J.; Blaszak, Ronald J.; Craw, Chad E.; Gondek, Cory Russell; LoMascolo, Frank; Petersen, Eric G.; West, Kenneth G.; Goldfain, Albert; Narayan, Mahesh; Chung, Song Y., Communication of information between a plurality of network elements.
DelloStritto, James J.; Blaszak, Ronald James; Chung, Song Yong; Craw, Chad Everett; Goldfain, Albert; Gondek, Cory Russell; LoMascolo, Frank; Narayan, Mahesh; Petersen, Eric G.; West, Kenneth G., Communication of information between a plurality of network elements.
Craw, Chad E.; DelloStritto, James J.; Chung, Song Y.; Blaszak, Ronald J.; LoMascolo, Frank, Device data sheets and data dictionaries for a dynamic medical object information base.
Shekhar, Raj; Castro-Pareja, Carlos; Dandekar, Omkar, Method and apparatus for accelerated elastic registration of multiple scans of internal properties of a body.
Punithakumar, Kumaradevan; Li, Shuo; Ben Ayed, Ismail; Ross, Ian; Islam, Ali; Rankin, Richard, Methods, apparatus and articles of manufacture to process cardiac images to detect heart motion abnormalities.
Punithakumar, Kumaradevan; Li, Shuo; Ben Ayed, Ismail; Ross, Ian; Islam, Ali; Rankin, Richard, Methods, apparatus and articles of manufacture to process cardiac images to detect heart motion abnormalities.
Nawana, Namal; Horton, William C.; Frasier, William J.; Serhan, Hassan A.; O'Neil, Michael; Sommerich, Robert E.; DiPietro, Jennifer; Nordstrom, Christopher; Parsons, Matthew; Martin, Thomas, Systems and methods for surgical and interventional planning, support, post-operative follow-up, and functional recovery tracking.
Nawana, Namal; Horton, William C.; Reinhardt, Max; Bellas, Jonathan; Cranson, Cody; DiPietro, Jennifer; Fowler, Mary Louise; Frasier, William J.; Griffin, John Paul; Hall, Mark T.; Konieczynski, David D.; LeClerc, Michelle; Martin, Thomas; Nordstrom, Christopher; O'Neil, Michael; Paiva, James; Parsons, Matthew; Pavento, Nicholas; Raymond, Douglas; Roveda, James J.; Selover, Sean P.; Serhan, Hassan A.; Slivka, Michael A.; Sommerich, Robert E., Systems and methods for surgical and interventional planning, support, post-operative follow-up, and, functional recovery tracking.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.