Machine learning models for automated diagnosis of disease database entities

Inventors

REICHER, Joshua • MUELLY, Michael

Assignees

Imvaria Inc

Abstract

A method of automated diagnosis of disease database entities includes receiving a case processing request via an input application programming interface (API), extracting image data from the case processing request including at least one medical scan image of the patient, selecting at least a portion of the medical scan image(s) according to specified selection criteria, normalizing the selected at least a portion of the medical scan image(s), supplying the selected at least a portion of the medical scan image(s) to a machine learning model to generate a target medical condition prediction output, wherein the target medical condition prediction output is indicative of a likelihood that a patient will experience a future disease diagnosis event corresponding to the target medical condition, and automatically transmitting the target medical condition prediction output as an electronic transmission via an output API to a provider system associated with the patient.

Core Innovation

The invention provides an end-to-end automated diagnosis method for disease database entities. The method receives a case processing request via an input application programming interface (API), where the request includes at least one medical scan image and at least one medical data entry associated with a patient. The method extracts image data and at least one of text data and lab data associated with a medical condition of the patient.

The method selects at least a portion of the medical scan image(s) according to specified selection criteria, normalizes the selected portion(s), and converts the selected portion(s) to at least one mathematical representation for processing by a machine learning model implementation. The machine learning model generates a target medical condition prediction output indicative of a likelihood that a patient will experience a future disease diagnosis event corresponding to the target medical condition.

The prediction output is automatically transmitted as an electronic transmission via an output API to a provider system associated with the patient. The disclosed automation may be implemented as SaMD without a visual UI. The workflow also selects and validates or parses DICOM metadata, normalizes CT into a 3D volumetric representation, and performs 3D deep learning using a convolutional neural network without human CT segmentation.

The invention addresses automated disease diagnosis for disease database entities, including interstitial lung disease and idiopathic pulmonary fibrosis, by integrating medical scan images and associated text and or lab data with an automated image-selection and processing pipeline. The background context is to enable an automated, API-based workflow that ingests, processes, and outputs diagnosis likelihood predictions to provider systems.

Claims Coverage

The partial claims include two independent claims. Across these claims, the coverage centers on an API-based automated diagnostic and processing pipeline that extracts image and associated text or lab data, selects and normalizes scan portions, converts selected portions to mathematical representations for machine learning, generates a future-diagnosis-event likelihood output, and automatically stores labeled case data in a case storage database.

Overall, the independent claims cover an automated, API-based workflow that ingests scan images with associated text or lab condition data, selects and normalizes scan portions, converts the selected portions into mathematical representations for a machine learning model, and produces an electronic likelihood of a future disease diagnosis event that is automatically transmitted to a provider system; a related independent claim additionally stores labeled case data in a case storage database, with scan selection criteria that may include slice thickness, image reconstruction kernel, and manufacturer.

Stated Advantages

Documented Applications