top of page


Designed, in the first instance, for the Healthcare sector, EICON AI provides a compelling implementation of an Edge Computing architecture to the AI Algorithm provider: manage the algorithms, deployment configurations, workflows, routing, etc. in the Cloud; deploy a fully configured agent ("CAN") to each Healthcare site to execute the algorithm, manage its results, and provide full audit trail of all data transformations and user activities.

Following is a high-level overview of core capabilities and optional additional modules that make up the EICON AI solution.


EICON REACH (AI) uses the core Edge Computing architecture of EICON REACH to facilitate AI-algorithm-based medical image data analysis at the edge, i.e. in the hospital setting. EICON AI is a fully configured instance of EICON REACH.

EICON AI is built entirely on the EICON REACH architecture and is the most complete demonstration, to date, of the power and potential that Edge Computing and EICON REACH represent.

We  have leveraged EICON REACH to create a comprehensive, end-to-end, fully automated AI algorithm execution platform that it is controlled and managed in the Cloud but executed in the Healthcare setting, adjacent to the data.

  • A highly innovative solution with enormous potential

  • On a Cloud-based instance of EICON REACH HUB

    • Manage your algorithms

    • Manage configurations and routing rules

    • Set up data pre-processing procedures

    • Configure Clinical Access Nodes (CANs) for distribution to target destinations in the Healthcare setting

    • Establish the communication initiation protocol - HUB to CAN or CAN to HUB (can be either but not both)

    • Send algorithm, configurations, routing rules to CAN

Once fully configured on the HUB, the CAN is deployed to any number of target destinations.

NOTE: The CAN is "minimally smart", i.e., it relies on the HUB for almost all its behaviors

  • Once deployed at the healthcare setting, the CAN will

    • Receive algorithm(s), configurations, workflows, routing rules from HUB​
    • ​Receive data from a data source, e.g., any DICOM node - an MR machine, a PACS system, etc

    • Route the scan through any pre-processing rules received from the HUB

    • Route the scan to the algorithm, located on the CAN

    • Receive results from the algorithm

    • Return results to sender

    • Transfer Audit Trail of all data transformations to HUB

On the HUB, review operational processing reports, audit trail, etc

bottom of page