Unsupervised Learning And Prediction Of Lines Of Therapy From High-Dimensional Longitudinal Medications Data

Abstract

In one aspect, the present disclosure provides a method for labeling one or more medications concurrently administered to a patient as a line of therapy. The method includes identifying medical records of the patient from a plurality of digital records, creating, from the subset of medical records, a plurality of treatment intervals including at least one medication administered to the patient and a time interval, associating medications of the one or more treatments with a respective treatment interval when the administration of the medication falls within the time interval, refining the time interval of a respective treatment interval when a treatment of the one or more treatments falls outside the time interval but within an extension period, identifying one or more potential lines of therapy from the plurality of treatment intervals, and labeling the potential line of therapy having the highest maximum likelihood estimation as the line of therapy.

1 . A method for identifying personalized care recommendations for a patient from one or more lines of therapy identified by an artificial intelligence engine from patient records of a cohort of patients, comprising:
 receiving medical record data related to the patient;   identifying one or more lines of therapy derived from the patient records of the cohort of patients, the one or more lines of therapy identified by the artificial intelligence engine deriving a plurality of curated fields from the patient records, identifying a subset of the curated fields reflecting medical histories of the cohort of patients with respect to one or more disease state diagnoses, one or more patient care events occurring after the disease state diagnoses, and one or more time intervals corresponding to the patient care events,   storing the identified one or more lines of therapy in a datastore;   identifying one or more of the stored lines of therapy as being relevant to a disease state of the patient;   evaluating responses to the one or more patient care events corresponding to the relevant lines of therapy; and   generating and transmitting a report to a user, the report comprising one or more personalized care recommendations based on the identified one or more relevant lines of therapy and the evaluated responses to the one or more patient care events, the personalized care recommendations including an identification of one or more patient care events to apply to the patient.   
     
     
         2 . The method of  claim 1 , wherein the identified one or more patient care events comprises a treatment including one or more of:
 administration of a medication, a drug, a molecule, or a chemical,   implantation of a medical device,   use of a medical device,   use of a biotherapy, a virotherapy, a phage therapy, a phytotherapy, a gene therapy, an epigenetic therapy, a protein therapy, an enzyme replacement therapy, a hormone therapy, a cell therapy, an immunotherapy, an antibody therapy, a nutrition therapy, an electromagnetic radiation therapy, or a radiation therapy,   a surgical procedure, or   a radiosurgery.   
     
     
         3 . The method of  claim 1 , wherein the identified one or more patient care events comprises a clinical trial, a therapy, or an off-label medication or treatment. 
     
     
         4 . The method of  claim 1 , wherein the one or more personalized care recommendations includes a time interval to administer the identified one or more patient care events. 
     
     
         5 . The method of  claim 1 , wherein the report includes analytics relating to the identified one or more patient care events. 
     
     
         6 . The method of  claim 5 , wherein the analytics include one or more of:
 a progression free survival analysis;   an outliers analysis;   an effectiveness of surgery analysis;   an analysis of patient molecular features on the responses to the one or more patient care events;   an adverse events analysis; or   an analysis of effects from duration of the relevant lines of therapy.   
     
     
         7 . The method of  claim 1 , wherein the artificial intelligence engine is unsupervised. 
     
     
         8 . The method of  claim 1 , wherein the report is stored as a webform, and wherein transmitting the report to the user comprises transmitting the webform to the user with instructions to display the webform via a graphical user interface presented on a user device. 
     
     
         9 . The method of  claim 1 , wherein the user is a health care provider. 
     
     
         10 . The method of  claim 1 , wherein the user is the patient. 
     
     
         11 . The method of  claim 1 , wherein disease state of the patient includes cancer, cardiology, depression, mental health, diabetes, infectious disease, epilepsy, dermatology, or autoimmune diseases. 
     
     
         12 . The method of  claim 1 , further comprising:
 receiving a selection of one of the identified one or more patient care events; and   generating instructions to administer the selected patient care event to the patient.   
     
     
         13 . The method of  claim 1 , wherein the step of identifying one or more lines of therapy derived from the patient records of the cohort of patients comprises:
 creating, from the patient records, a plurality of treatment intervals comprising at least one medication administered to the patient and a time interval, wherein a given time interval is associable with more than one treatment interval;   associating medications of the one or more treatments with a respective treatment interval when the administration of the medication falls within the time interval;   extending the time interval of a respective treatment interval to include an earlier or later time interval of another treatment when the another treatment of the one or more treatments falls outside the time interval but within an extension period and wherein an interval superseding event does not take place during the extended time interval;   identifying, via the artificial intelligence engine, one or more potential lines of therapy from the plurality of treatment intervals, wherein each potential line of therapy comprises one or more of the plurality of treatment intervals and a maximum likelihood estimation that the one or more of the plurality of treatment intervals of the potential line of therapy is a line of therapy; and   labeling, via the unsupervised artificial intelligence engine, the potential line of therapy having the highest maximum likelihood estimation as the line of therapy.   
     
     
         14 . The method of  claim 1 , wherein the artificial intelligence engine deriving a plurality of curated fields from the patient records comprises receiving non-curated digital patient records comprising a plurality of different record types or a plurality of records from different sources, combining related portions of the non-curated digital records, and determining that other portions of the non-curated digital records are unrelated to the related portions. 
     
     
         15 . The method of  claim 14 , further comprising:
 digitizing, optical character recognizing, and encoding the curated fields into a structured format.   
     
     
         16 . The method of  claim 1 , further comprising:
 labeling, via the artificial intelligence engine, additional potential lines of therapy during successive time intervals as an incrementally numbered line of therapy from the potential lines of therapy occurring chronologically after each preceding line of therapy and having the respective highest maximum likelihood estimation of the potential lines of therapy in each successive time interval.   
     
     
         17 . A system for identifying personalized care recommendations for a patient from one or more lines of therapy identified by an artificial intelligence engine from patient records of a cohort of patients, comprising:
 at least one computer having at least one processor programmed to:
 receive medical record data related to the patient; 
 identify one or more lines of therapy derived from the patient records of the cohort of patients, the one or more lines of therapy identified by the artificial intelligence engine deriving a plurality of curated fields from the patient records, identifying a subset of the curated fields reflecting medical histories of the cohort of patients with respect to one or more disease state diagnoses, one or more patient care events occurring after the disease state diagnoses, and one or more time intervals corresponding to the patient care events, 
 store the identified one or more lines of therapy in a datastore; 
 identify one or more of the stored lines of therapy as being relevant to a disease state of the patient; 
 evaluate responses to the one or more patient care events corresponding to the relevant lines of therapy; and 
 generate and transmit a report to a user, the report comprising one or more personalized care recommendations based on the identified one or more relevant lines of therapy and the evaluated responses to the one or more patient care events, the personalized care recommendations including an identification of one or more patient care events to apply to the patient. 
   
     
     
         18 . The system of  claim 17 , wherein the processor being programmed to identify one or more lines of therapy derived from the patient records of the cohort of patients comprises the processor being programmed to:
 create, from the patient records, a plurality of treatment intervals comprising at least one medication administered to the patient and a time interval, wherein a given time interval is associable with more than one treatment interval;   associate medications of the one or more treatments with a respective treatment interval when the administration of the medication falls within the time interval;   extend the time interval of a respective treatment interval to include an earlier or later time interval of another treatment when the another treatment of the one or more treatments falls outside the time interval but within an extension period and wherein an interval superseding event does not take place during the extended time interval;   identify, via the artificial intelligence engine, one or more potential lines of therapy from the plurality of treatment intervals, wherein each potential line of therapy comprises one or more of the plurality of treatment intervals and a maximum likelihood estimation that the one or more of the plurality of treatment intervals of the potential line of therapy is a line of therapy; and   label, via the unsupervised artificial intelligence engine, the potential line of therapy having the highest maximum likelihood estimation as the line of therapy.   
     
     
         19 . A computer program product for identifying personalized care recommendations for a patient from one or more lines of therapy identified by an artificial intelligence engine from patient records of a cohort of patients, comprises instructions that, when executed by a processor, cause the processor to:
 receive medical record data related to the patient;   identify one or more lines of therapy derived from the patient records of the cohort of patients, the one or more lines of therapy identified by the artificial intelligence engine deriving a plurality of curated fields from the patient records, identifying a subset of the curated fields reflecting medical histories of the cohort of patients with respect to one or more disease state diagnoses, one or more patient care events occurring after the disease state diagnoses, and one or more time intervals corresponding to the patient care events,   store the identified one or more lines of therapy in a datastore;   identify one or more of the stored lines of therapy as being relevant to a disease state of the patient;   evaluate responses to the one or more patient care events corresponding to the relevant lines of therapy; and   generate and transmit a report to a user, the report comprising one or more personalized care recommendations based on the identified one or more relevant lines of therapy and the evaluated responses to the one or more patient care events, the personalized care recommendations including an identification of one or more patient care events to apply to the patient.   
     
     
         20 . The computer program product of  claim 19 , wherein the instructions causing the processor to identify one or more lines of therapy derived from the patient records of the cohort of patients comprise instructions that cause the processor to:
 create, from the patient records, a plurality of treatment intervals comprising at least one medication administered to the patient and a time interval, wherein a given time interval is associable with more than one treatment interval;   associate medications of the one or more treatments with a respective treatment interval when the administration of the medication falls within the time interval;   extend the time interval of a respective treatment interval to include an earlier or later time interval of another treatment when the another treatment of the one or more treatments falls outside the time interval but within an extension period and wherein an interval superseding event does not take place during the extended time interval;   identify, via the artificial intelligence engine, one or more potential lines of therapy from the plurality of treatment intervals, wherein each potential line of therapy comprises one or more of the plurality of treatment intervals and a maximum likelihood estimation that the one or more of the plurality of treatment intervals of the potential line of therapy is a line of therapy; and   label, via the unsupervised artificial intelligence engine, the potential line of therapy having the highest maximum likelihood estimation as the line of therapy.