Machine learning models for automated request processing
US11848097B2 · kind B2 · utility
Assignee
Inventors
Key dates
| Filing date | Dec 17, 2020 |
| Grant date | Dec 19, 2023 |
| Priority date | — |
| Expiry date | Feb 5, 2042 |
Classification
- Technology area (CPC G)Physics
- CPC primaryG06T2207/20081
- WIPO fieldMedical technology
- WIPO sectorInstruments
Abstract
A computerized method of automatically processing a medical imaging record using a machine learning model includes training a machine learning model prediction engine with historical medical imaging records, receiving a first medical imaging record from a first system, applying a set of specified approval criteria to the first medical imaging record to determine a provisional outcome, and in response to the provisional outcome being negative, selectively identifying an exception to the provisional outcome in response to input received by a user interface. In response to the exception not being identified, the method includes inputting a feature vector based on the first medical imaging record to the machine learning model prediction engine to generate a likelihood estimate, comparing the generated likelihood estimate to a target threshold, and in response to the generated likelihood estimate being greater than the target threshold, transmitting a signal indicating approval to the first system.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.