Machine Learning Systems, Methods, Components, and Software for Recommending and Ordering Independent Medical Examinations

Abstract

Millions of bodily injury insurance claims are filed yearly to help people who've suffered accidents from work, slip-and-falls, and auto collisions. In processing these claims, insurance companies, TPAs, and law firms routinely hire experts, such as physicians, to conduct independent medical evaluations (IMEs) to assist claims adjusters and attorneys in analyzing the eligibility of claimants for indemnity and medical benefit payments. IMEs typically cost thousands of dollars each. Yet, many are ordered too early, wasting money that could otherwise be used to reduce insurance premiums. To reduce this waste, the inventors devised, among other things, one or more exemplary systems which not only predict the outcomes of IMEs based on claimant medical records and/or or activity data before ordering them, but also presents selected claims and predictions within a graphical user interface that facilitates ordering the IMEs from a list of available physicians.

1 - 30 . (canceled) 
     
     
         31 . A system comprising a non-transitory machine-readable storage medium having instructions stored thereon, which, when executed by one or more processors, cause the system to: obtain a first set of electronic medical records for one or more claimants associated with corresponding filed insurance claims for corresponding bodily injuries, extract a first set of one or more features from the first set of medical records; feed the first set of one or more features into a neural network responsive to the first set of one or more features to output and store in a memory on a communications network a maximum medical improvement prediction for at least a first one of the claimants; generate at least a portion of a graphical user interface display having an insurance claim display region, the claim display region including identifier indicia associated with the insurance claim of the one insurance claimant, an indication of the maximum medical improvement prediction for the one insurance claimant, and a first user selectable control feature operatively associated with the second insurance claimant for causing display of a list of one or more physician identifiers for physicians determined to be qualified and available to perform an independent medical examination of the one insurance claimant based on the indication of the maximum medical improvement prediction, and a second user selectable control feature for electronically ordering an independent medical examination of the one insurance claimant from a selected one of the listed physicians. 
     
     
         32 . The system of  claim 31 , wherein the instructions also cause one or more of the processors to exclude one or more medical records from each of the first sets of medical records from affecting the maximum medical improvement prediction. 
     
     
         33 . The system of  claim 31 : wherein the first set of medical records for each corresponding insurance claimant define a corresponding temporal record sequence; and wherein the instructions further cause one or more of the processors to prevent one or more of the medical records that do not differ sufficiently from an immediately preceding medical record within its corresponding temporal record sequence from affecting the maximum medical improvement prediction. 
     
     
         34 . The system of  claim 31 , wherein the graphical user interface display presents a representation of a prediction confidence indication in association with the claim identifier indicia for the one insurance claimant. 
     
     
         35 . The system of  claim 31 , wherein each listed physician in the graphical user interface display has a predetermined availability to conduct the independent medical examination within a predetermined time frame, such as seven days, from a current date of user selection of the first user selectable control feature. 
     
     
         36 . The system of  claim 31 , wherein the instructions further cause one or more of the processors to: obtain a second set of electronic medical records for two or more claimants associated with corresponding filed insurance claims for corresponding bodily injuries, the medical records for each claimant having a corresponding maximum medical recovery status label indicating affirmatively or negatively whether the records represent maximum medical recovery for his or her bodily injury; extract a second set of one or more features from the second set of medical records; and train the neural network based on the second set of one or more features and corresponding maximum medical recovery status labels. 
     
     
         37 . The system of  claim 31 , wherein the instructions further cause one or more of the processors to cause the neural network to output a prediction of whether a human physician would regard the medical records as indicating that the one claimant has experienced proper or improper medical treatment for his or her bodily injury. 
     
     
         38 . The system of  claim 37 , wherein the proper or improper medical status prediction predicts whether a physician conducting an independent medical examination would consider current activity restrictions for a claimant proper or not; or current medical treatment for the claimed injury proper or not; or a recommended treatment for the claimed injury proper or not; or work disability for the claimed injury proper or not. 
     
     
         39 . The storage medium of  claim 31 , wherein the neural network algorithm has been trained using supervised training. 
     
     
         40 . A non-transitory machine-readable storage medium having instructions stored thereon, which, when executed by one or more processors, cause the system to: obtain a first set of electronic medical records for one or more claimants associated with corresponding filed insurance claims for corresponding bodily injuries, extract a first set of one or more features from the first set of medical records; feed the first set of one or more features into a neural network to output a prediction of whether a physician would deem the medical records as indicating that the claimant has achieved maximum or sub-maximum medical recovery and thus is likely ineligible or likely eligible for continued insurance benefits for the injury. 
     
     
         41 . The storage medium of  claim 40 , wherein the instructions also cause one or more of the processors to exclude one or more medical records from each of the first sets of medical records from affecting the assessment. 
     
     
         42 . The storage medium of  claim 40 : wherein the first set of medical records for each corresponding insurance claimant define a corresponding temporal record sequence; and wherein the instructions further cause one or more of the processors to prevent one or more of the medical records that do not differ sufficiently from an immediately preceding medical record within its corresponding temporal record sequence from affecting the output of the neural network. 
     
     
         43 . The storage medium of  claim 40 , wherein the neural network algorithm has been trained using supervised training. 
     
     
         44 . The storage medium of  claim 40 , wherein the instructions further cause one or more of the processors to: generate at least a portion of a graphical user interface display having an insurance claim display region, the claim display region including identifier indicia associated with the insurance claim of the insurance claimant, an indication of the maximum medical improvement prediction, and a first user selectable control feature operatively associated with the second insurance claimant for causing display of a list of one or more physician identifiers for physicians determined to be qualified and available to perform an independent medical examination of the second insurance claimant within a predetermined time period, and a second user selectable control feature for electronically ordering an independent medical examination of the second insurance claimant from a selected one of the listed physicians. 
     
     
         45 . The storage medium of  claim 44 , wherein each listed physician in the graphical user interface display has a predetermined availability to conduct the independent medical examination within a predetermined time frame, such as seven days, from a current date of user selection of the first user selectable control feature. 
     
     
         46 . The medium of  claim 42 , wherein the instructions further cause one or more of the processors to cause a neural network to output a prediction of whether a physician would regard the medical records as indicating that a proper or improper medical status. 
     
     
         47 . The storage medium of  claim 46 , wherein the proper or improper medical status prediction predicts whether a physician conducting an independent medical examination would consider current activity restrictions for a claimant proper or not; or current medical treatment for the claimed injury proper or not; or a recommended treatment for the claimed injury proper or not; or work disability for the claimed injury proper or not. 
     
     
         48 . A non-transitory machine-readable storage medium having instructions stored thereon, which, when executed by one or more processors, cause the system to: obtain a first set of electronic medical records for one or more claimants associated with corresponding filed insurance claims for corresponding bodily injuries, extract a first set of one or more features from the first set of medical records; feed the first set of one or more features into an artificial intelligence to output two or more physician predictions, with at least of the predictions being whether a physician would deem the medical records as indicating that the claimant has achieved maximum or sub-maximum medical improvement and at least one of the predictions being whether a physician would consider current activity restrictions for a claimant proper or not, or current medical treatment for the claimed injury proper or not, or a recommended treatment for the claimed injury proper or not, or work disability for the claimed injury proper or not. 
     
     
         49 . The storage medium of  claim 48 , wherein the instructions further cause one or more of the processors to: generate at least a portion of a graphical user interface display having an insurance claim display region, the claim display region including identifier indicia associated with the insurance claim of the insurance claimant, an indication of the maximum medical improvement prediction, and a first user selectable control feature operatively associated with the second insurance claimant for causing display of a list of one or more physician identifiers for physicians determined to be qualified and available to perform an independent medical examination of the second insurance claimant within a predetermined time period, and a second user selectable control feature for electronically ordering an independent medical examination of the second insurance claimant from a selected one of the listed physicians. 
     
     
         50 . The storage medium of  claim 49 , wherein the artificial intelligence comprises at least one supervised trained neural network or at least one binary logistical regression algorithm. 
     
     
         51 . The storage medium of claim, wherein the instructions also cause one or more of the processors to exclude one or more medical records from each of the first sets of medical records from affecting at the predictions.