A new study proposes integrating blockchain technology with the Internet of Things to manage decentralized medical records. The researchers argue that combining blockchain with IoT can provide privacy, security, and tamper-proof data storage while enabling remote patient monitoring. To avoid bloating the blockchain, the system stores only the medical reports’ hashes on-chain and keeps the actual reports on an IPFS repository. A Node.js intermediary ensures only authenticated requests write to the ledger and coordinates IPFS interactions.
The architecture uses Hyperledger Fabric with RAFT consensus to support high throughput and privacy. IPFS is installed as a local node to manage heavy medical reports in formats such as PDF, DICOM, and JPEG. The client SDK interacts with the smart contract, and the content identifiers (hashes) from IPFS are stored on the ledger for retrieval. RPM UNIT employs an ESP32 microcontroller to gather heart rate, SpO2, and body temperature via MAX30100 and DS18B20 sensors.
The RPM unit periodically transmits vitals data to a Node.js API, which then writes the information to the blockchain ledger while coordinating IPFS interactions for reports when needed. Experiment I (Reliability) tests involved generating a large number of transactions. The Caliper benchmarking tool produced 1,500 transactions with four parallel workers, each submitting at 25 transactions per second, totaling 100 TPS, and all transactions were committed to the ledger.
In a real RPM scenario, the RPM unit transmitted 480 transactions over eight hours and all were recorded without loss, with vitals data captured and processed through the ESP32-based sensor setup. Experiment II (Performance) evaluated throughput and latency under heavier parallel loads. The setup used two, four, and eight workers corresponding to 40, 80, and 160 TPS, respectively. Write throughput peaked at 91.4 TPS when attempting 94 TPS, while read throughput reached 160.2 TPS with a constant read latency of 0.01 seconds.
The observed latencies varied, with a minimum around 0.24 seconds at 40 TPS and a maximum near 0.45 seconds at higher write loads; read operations consistently imposed lower latency than writes. The maximum single JSON transaction size noted was 4 KB. For testing, ECG data from a Kaggle dataset was uploaded to IPFS to generate and verify medical report hashes.
The study concludes that storing both patient vitals and medical reports on the blockchain ledger is feasible and beneficial for privacy and data integrity in remote patient monitoring. Looking ahead, the authors suggest evaluating memory usage, energy consumption, and incorporating additional sensors such as ECG and EEG to enhance the system’s capabilities.
Leave a Reply