×

System and method for detecting billing errors using predictive modeling

  • US 9,785,983 B2
  • Filed: 06/13/2013
  • Issued: 10/10/2017
  • Est. Priority Date: 06/13/2012
  • Status: Active Grant
First Claim
Patent Images

1. A system for detecting billing errors using artificial intelligence comprising:

  • a computer system in communication with a billing client, said computer system electronically receiving and processing billing information electronically gathered by the billing client over a pre-defined period of time, said computer system configured to include an artificial neural network having an input layer, a plurality of processing elements, and an output layer;

    a billing history database in communication with the computer system and storing the billing information, the computer system processing the billing information to select one or more data fields of the billing information; and

    a billing error detection engine executed by the computer system, said detection engine processing the one or more data fields using one or more predictive models to detect, score, and flag potential billing errors in the billing information, the billing error detection engine executing the following steps;

    a feedback model so that the computer system learns relationships between billing codes present in the billing information,an inpatient model that targets low charges and high charges in inpatient data by filtering for each Diagnosis Related Groups (DRG) the number of visits within a pre-defined threshold and then applying a Principal Component Analysis (PCA) Module to calculate and compare a department-hospital level average with a reconstructed value for new visits, the inpatient model utilizing said artificial neural network of said computer system, said artificial neural network reconstructing charge values and flagging actual charge values for review if a difference between the reconstructed charge value and the actual charge value is above a thresholdan outpatient model that detects missing codes in outpatient data by applying a supervised learning model to learn the probability of a presence of a code using a logistic regression (LR) model for each code to be evaluated, and applying a Decision Tree (DT) model to capture non-linearity between data and their codes and to take into account multiple hospitals,applying a joint-density learning model to learn interdependencies between visit data using a Restricted Boltzmann Machine (RBM) model to compute whether a code should be present and a probability of missing charges, and applying a Gaussian Missing Data (GMD) model to suggest other codes that should be present; and

    executing a cascade model to capture relationship between codes and improve prediction accuracy and performance by (i) applying a normalization model to pre-process outputs of the LR model, the DT model, and the RBM model to calibrate the outputs for consistency, (ii) applying an ensemble model to combine the LR model, the DT model, the RBM model, and the GMD model to generate an ensemble score, and (iii) applying a feedback model to further refine results by receiving as input a predicted code and an ensemble score to generate a probability of code acceptance;

    wherein the computer system transmits the flagged potential billing errors to the billing client for review.

View all claims
  • 9 Assignments
Timeline View
Assignment View
    ×
    ×