Self-improving classification system
AssigneeMedtronic, Inc.
InventorsOrhan Soykan, Daisy Phan Cross
Life of this patent
Estimating the term…
Abstract
A self-improving classification system classifies specimens based on class identifiers. The system stores specimen profiles in a database that is updated with additional specimen profiles and with follow-up data that corrects classification of specimens that were initially incorrectly classified. Algorithms use the updated database to discover new class identifiers, modify thresholds of known class identifiers, and drop unnecessary class identifiers to improve classification of specimens.
Claims25
A self-improving method of classifying a specimen, the method comprising:
- extracting one or more selected from the group consisting of proteomic information and genomic information from the specimen to determine at least one proteomic parameter or genomic parameter;
- using a computer to develop an algorithm for classifying specimens based on class identifiers, wherein the class identifiers include at least one physiological or demographic parameter and one or more selected from the group consisting of proteomic parameters and genomic parameters;
- using a computer to generate a specimen profile from the specimen by analyzing the class identifiers of the specimen and storing the specimen profile in a database;
- classifying the specimen into a class based on the specimen profile and the class identifiers using the algorithm;
- determining if reclassification of the specimen is necessary;
- refining the algorithm based on a refinement of the class identifiers;
- and
- reclassifying the specimen if necessary.
(canceled)
(canceled)
14-33. (canceled)
A self-improving system of classifying specimens, the system comprising:
- a database containing specimen profiles that are clustered into classes based on class identifiers, wherein class identifiers include at least one physiological or demographic parameter and one or more selected from the group consisting of proteomic parameters and genomic parameters;
- a means for classifying specimens using a computer and a first algorithm based on the class identifiers, and placing the specimens in classes based on the class identifiers;
- and
- a means for improving the first algorithm based on follow-up data regarding accuracy of classification of the specimens.
36-38. (canceled)
A self-improving method of identifying patients who are candidates for treatment with a medical device, the method comprising:
- creating a database;
- gathering data from patients, the data including class identifiers wherein the class identifiers include at least one physiological or demographic parameter and one or more selected from the group consisting of proteomic parameters and genomic parameters;
- analyzing the data using a computer and an algorithm, the algorithm identifying patients who are candidates for treatment with the medical device;
- adding the data to the database;
- correcting identification of the patients based upon follow-up data indicating a response of the patients to the medical device;
- and
- updating the database and the algorithm using corrected identification of the patients.
The self-improving method of 39, wherein the follow-up data further comprises:
- an operational history of the medical device.
46-54. (canceled)
Dependencies are inferred from the claim text.