Abstract:Testing healthcare Internet of Things (IoT) applications at system and integration levels necessitates integrating numerous medical devices of various types. Challenges of incorporating medical devices are: (i) their continuous evolution, making it infeasible to include all device variants, and (ii) rigorous testing at scale requires multiple devices and their variants, which is time-intensive, costly, and impractical. Our collaborator, Oslo City's health department, faced these challenges in developing automated test infrastructure, which our research aims to address. In this context, we propose a meta-learning-based approach (MeDeT) to generate digital twins (DTs) of medical devices and adapt DTs to evolving devices. We evaluate MeDeT in OsloCity's context using five widely-used medical devices integrated with a real-world healthcare IoT application. Our evaluation assesses MeDeT's ability to generate and adapt DTs across various devices and versions using different few-shot methods, the fidelity of these DTs, the scalability of operating 1000 DTs concurrently, and the associated time costs. Results show that MeDeT can generate DTs with over 96% fidelity, adapt DTs to different devices and newer versions with reduced time cost (around one minute), and operate 1000 DTs in a scalable manner while maintaining the fidelity level, thus serving in place of physical devices for testing.

What problem does this paper attempt to address?

The main problem that this paper attempts to solve is how to efficiently integrate and test multiple evolving medical devices in the system and integration testing of the Internet of Things (IoT) in healthcare applications. Specifically, the paper aims to address the following two main challenges: 1. **Device Evolution (Challenge 1: Evolution)** - Medical devices are constantly undergoing new feature additions, hardware/software upgrades, and service updates, resulting in the existence of multiple versions of devices, each with different hardware/software characteristics. - Incorporating these different - versioned multiple - type devices into the test infrastructure is not feasible because of the large variety and frequent changes of devices. 2. **Large - Scale Testing (Challenge 2: Large Scale Testing)** - Large - scale testing at the system and integration levels requires integrating thousands of evolving medical devices with healthcare IoT applications. - This process is not only labor - intensive and costly, but may also generate excessive requests to the device server during the testing process, leading to service blockage and even device damage. To solve these problems, the authors propose a meta - learning - based method (MeDeT) for generating digital twins (Digital Twins, DTs) of medical devices and enabling these DTs to adapt to device evolution. In this way, DTs can be used to replace physical devices during the testing process, thereby achieving large - scale automated testing and reducing dependence on and potential damage to physical devices. ### Specific Objectives - **Generate Digital Twins (DTs)**: Use meta - learning techniques to generate digital twins of medical devices, ensuring that they can accurately simulate the behavior of actual devices. - **Adapt to Device Evolution**: Enable the generated DTs to quickly adapt to new or improved versions of devices, reducing the time and resource consumption for retraining. - **Support Large - Scale Testing**: Be able to operate a large number of DTs (e.g., 1000) simultaneously to support large - scale system and integration testing. - **Improve Testing Efficiency**: Significantly improve testing efficiency by reducing the time cost required for generating and adapting DTs (e.g., from 16 - 25 minutes to approximately 1 minute). ### Method Overview The core idea of MeDeT is to utilize the few - shot learning ability in meta - learning to quickly generate and adapt DTs. The specific steps include: 1. **Data Generation**: Use the API documents of medical devices to generate simulated data. 2. **Data Preparation**: Pre - process the generated data to ensure its suitability for training machine - learning models. 3. **Meta - Learning**: Create meta - data sets and task sets, determine the neural network model architecture, and train the model through continuous evaluation and adaptation steps. 4. **Build Digital Twins**: Create DTs and their API interfaces according to the device model, serial number, and trained model. 5. **DT Request Handler**: Allow DTs to integrate with healthcare IoT applications and handle requests during testing. 6. **DT to Device Communication**: Establish communication between DTs and their corresponding physical devices to send requests and receive responses. Through this method, MeDeT can effectively generate high - fidelity DTs and adapt them to evolving medical devices, thus providing strong support for system and integration testing of healthcare IoT applications.

MeDeT: Medical Device Digital Twins Creation with Few-shot Meta-learning