US 12,481,961 B2
Systems and methods for using federated learning in healthcare model development
Yuval Baror, Ramat Hasharon (IL); Ittai Dayan, Swampscott, MA (US); and Yaron Blinder, Haifa (IL)
Assigned to Rhino Federated Computing, Inc., Boston, MA (US)
Filed by Rhino Federated Computing, Inc., Boston, MA (US)
Filed on Mar. 8, 2023, as Appl. No. 18/180,710.
Claims priority of provisional application 63/269,053, filed on Mar. 9, 2022.
Prior Publication US 2023/0290456 A1, Sep. 14, 2023
Int. Cl. G06F 17/00 (2019.01); G06F 21/62 (2013.01); G06Q 10/101 (2023.01); G16H 10/60 (2018.01)
CPC G06Q 10/101 (2013.01) [G06F 21/6254 (2013.01); G16H 10/60 (2018.01)] 15 Claims
OG exemplary drawing
 
1. A federated learning system for model development comprising:
a server accessible by each of a first and second client agent, the server comprising instructions which, when executed by one or more processors, cause the server to perform a process operable to:
create a project based on an indication received from a first user device;
receive collaborator information from the first user device, the collaborator information identifying one or more collaborators at a second site;
receive a model configuration from the first user device;
receive a request to import a first cohort from the first client agent residing on a first network associated with the first site, wherein the first cohort was generated by the first client agent accessing a first dataset, pseudonymizing or deidentifying it if it contains protected health information (PHI) or personal identifiable information (PII);
receive a request to import a second cohort from the second client agent residing on a second network associated with the second site, wherein the second cohort was generated by the second client agent accessing a second dataset, pseudonymizing or deidentifying it if it contains PHI or PII; and
initiate, via one or more containers, a federated learning process based on the model configuration using the first and second cohorts as training data, the server operating as a federated server and the first and second client agents operating as federated learning clients.