Computer systems and methods for machine-learning based treatment modeling for oncology based on inconsistent ras biomarker detection data records

Patent metadata

Field	Value
Publication number	US-2022399126-A1
Application number	US-202117344510-A
Country	US
Kind code	A1
Filing date	Jun 10, 2021
Priority date	Jun 10, 2021
Publication date	Dec 15, 2022
Grant date	—

Abstract

To enable automated processing of certain observation data by machine-learning based models, as well as to ensure those machine-learning based models are trained utilizing reliable and consistent data, independently generated observation data records each comprising biomarker mutation indicators, are utilized to generate a model input data set by applying biomarker mutation indicator-based filters, including intra-date filters and inter-date filters to identify and rectify observation data records that do not satisfy applicable biomarker mutation indicators. By identifying and rectifying the observation data records that do not satisfy applicable biomarker mutation indicators, a clean, model input data set is generated that is then utilized to generate a severity score for the patient.

First claim

That which is claimed:

1 . A computer-implemented method for automatically modeling severity attributes of a colon cancer or rectal cancer treatment utilizing a plurality of independently generated observation data records for a patient, the method comprising:
receiving a plurality of independently generated observation data records each comprising a biomarker mutation indicator relating to RAS, KRAS, or NRAS and wherein KRAS and NRAS are subsets of RAS; generating a model input data set comprising a subset of the plurality of independently generated observation data records at least in part by: applying a plurality of biomarker mutation indicator-based filters comprising:
an intra-date filter configured to identify a plurality of observation data records having a shared date-of-service and to eliminate one or more observation data records failing to satisfy the intra-date filter; and
an inter-date filter configured to identify a plurality of observation data records across a plurality of dates of service and to eliminate one or more observation data records failing to satisfy the inter-date filter;
wherein each of the plurality of biomarker mutation indicator-based filters are configured to identify observation data records failing to satisfy the biomarker mutation indicator-based filters across RAS, KRAS, and NRAS biomarker mutation indicators;
after applying the plurality of biomarker mutation indicator-based filters to identify the subset of the plurality of independently generated observation data records, generating a derived biomarker mutation indicator for each of the subset of the plurality of independently generated observation data records, wherein the derived biomarker mutation indicator is a derived RAS biomarker mutation indicator generated based at least in part on the biomarker mutation indicator included within a respective observation data record; providing the model input data set to a machine-learning severity model configured to generate severity data based at least in part on the derived RAS biomarker mutation indicators of the observation data records within the model input data set; and generating, via the machine-learning severity model, severity data relating to the patient.

2 . The computer-implemented method of claim 1 , wherein generating a model input data set comprises generating a flat model input data file comprising each of the subset of the plurality of observation data records.

3 . The computer-implemented method of claim 1 , wherein the intra-date filter is configured to eliminate one or more observation data records failing to satisfy at least one intra-date filter configuration selected from:
(A) the intra-date filter is configured to identify a first observation data record having the shared date-of-service and having a positive result indicator of at least one of RAS, KRAS, or NRAS and a second observation data record having the shared date-of-service and having a negative result indicator of at least one of RAS, KRAS, or NRAS and to eliminate the first observation data record and the second observation data record as failing to satisfy the intra-date filter; (B) the intra-date filter is configured to identify a first observation data record having the shared date-of-service and having a positive result indicator of RAS, a second observation data record having the shared date-of-service and having a negative result indicator of KRAS, and a third observation data record having the shared date-of-service and having a negative result indicator of NRAS and to eliminate the first observation data record, the second observation data record, and the third observation data record as failing to satisfy the intra-date filter; or (C) the intra-date filter is configured to identify a first clinical record having the shared date-of-service and having a negative result indicator of RAS and a second observation data record having the shared-date-of-service and having a positive result indicator of at least one of KRAS or NRAS and to eliminate the first observation data record and the second observation data record as failing to satisfy the intra-date filter.

4 . The computer-implemented method of claim 1 , wherein the inter-date filter is configured to eliminate one or more observation data records failing to satisfy at least one inter-date filter configuration selected from:
(A) the inter-date filter is configured to identify a first observation data record having a first date-of-service and having a positive result indicator of at least one RAS, KRAS, or NRAS and a second observation data record having a second date-of-service occurring after the first date-of-service and having a negative result indicator of at least one of RAS, KRAS, or NRAS and to eliminate the first observation data record and the second observation data record as failing to satisfy the inter-date filter; (B) the inter-date filter is configured to identify a first observation data record having a first date-of-service having a positive result indicator of at least one NRAS or KRAS and a second observation data record having a second date-of-service occurring after the first-date-of-service and having a negative result indicator of RAS and to eliminate the first observation data record and the second observation data record as failing to satisfy the inter-date filter; or (C) the inter-date filter is configured to identify a first observation data record having a first date-of-service and having a positive result indicator of at least one of RAS, KRAS, or NRAS and having a second observation data record having a second date-of-service occurring after the first date-of-service and having a negative result indicator of at least one of KRAS or NRAS and to eliminate the second observation data record as failing to satisfy the output filter.

5 . The computer-implemented method of claim 1 , wherein generating a model input data set is performed in accordance with a relevant data pre-processing methodology relating to colon cancer or rectal cancer, and wherein the method further comprises retrieving the relevant data pre-processing methodology from a plurality of data pre-processing methodologies based at least in part on the plurality of independently generated observation data records prior to generating the model input data set.

6 . The computer-implemented method of claim 1 , wherein:
the intra-date filter is configured to eliminate one or more observation data records failing to satisfy the intra-date filter relating to one of RAS, KRAS, or NRAS; and the inter-date filter is configured to eliminate one or more observation data records failing to satisfy the inter-date filter relating to one of RAS, KRAS, or NRAS.

7 . The computer-implemented method of claim 1 , further comprising applying a preliminary filter criteria before generating the model input data set, wherein the preliminary filter criteria comprise one or more of:
a date-based filter criterion for selecting independently generated observation data records for further analysis as generated within a defined date range; a data source filter criterion for selecting independently generated observation data records for further analysis as generated by one or more defined data sources; or a data content filter criterion for selecting independently generated observation data records for further analysis as containing an identifier selected from a plurality of available identifiers eligible for further analysis.

8 . The computer-implemented method of claim 1 , wherein the machine-learning severity model is a linear regression model.

9 . A system comprising one or more memory storage areas and one or more processors for automatically modeling severity attributes

Assignees

• Optum Inc

Inventors

• John Christine M
• Ochs Lorre Ann
• Richards Lynn Anne
• Guo Fang
• Fan Liping
• Vaynerman Julia

Classifications

• G16H50/30

  for calculating health indices; for individual health risk assessment · CPC title

• G16H50/20 Primary

  for computer-aided diagnosis, e.g. based on medical expert systems · CPC title

• G16H20/00

  ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance · CPC title

• G16H50/50 Primary

  for simulation or modelling of medical disorders · CPC title

• G16H50/70

  for mining of medical data, e.g. analysing previous cases of other patients · CPC title