About Helios Neural Network
Helios Neural Network is a data distribution platform focused on storing and indexing data required for Artificial Intelligence to grow and learn. Instead of storing the data in expensive, traditional closed inaccessible databases, HNN stores data in nodes, provided by the users of Helios Neural Network.
The Helios Neural Network Blockchain stores only the contracts formed between actors defining their involvement in the Helios Neural Network. The Helios Neural network is a custom built blockchain. It features our custom proof of work application that rewards miners for storing, indexing and processing data on their local computers. The distribution and control of the network is managed by custom smart contracts between the miners, the data providers and the data consumers. The Helios Neural Network will fundamentally be able to store and process any indexable data however the platform is being tailored to provide optimum performance for Artificial Intelligence and Machine Learning applications.
The Helios Neural Network is powered by a peer to peer network of Helios Nodes. The Helios node stores and indexes encrypted data on a partition of a node hosts hard drive. The node application enables hosts on the network to transfer data, negotiate contracts, verify the integrity of the data blobs and receive payment for services rendered. To participate in the network nodes must create a keypair which in turn forms the wallet address of the node. All transactions are signed by the node using the wallet address and will in turn validated all signed transactions processed by the node. This enforces accountability on the network and reduces the effectiveness of attacks to the network.
Helios Health Connect
Helios Health connect is Health industry specific Helios Connect application that natively support anonymization of data, integration with existing IHE standards such as Dicom and HL7. XDS will be added in a later version.
Helios Health Connect can act as Dicom Service Class Provider (SCP) to receive Dicom images. When images are sent to the Helios Health Connect SCP the Dicom header is read, with all identifying data removed, data is temporarily stored in a local cache until the report has been received via HL7. The relevant information is added to the image/s manifest ready to be stored and processed on the Helios Neural Network. Once storage has been confirmed on the Helios Neural Network the data will be removed from the local cache.
To retrieve data from the Helios Neural network a dataset is built from the subscribed data within the Helios Health Connect application. At this point the data will be retrieved to the local cache where Helios Health Connect will act as a Dicom Service Class User (SCU) and send the data to the receiving application. In parallel the relevant report data will be used to generate an Observation Result (ORU) HL7 message to provide context to the Dicom images being received.