Last week, we had the pleasure of attending the 20th edition of the IEEE MeMeA conference, held in the beautiful setting of Chania, Crete. We presented three research contributions, each exploring different ways advanced sensing, AI, and real-time processing can reshape the future of medical technology.

📌 1. Deep Edge-AI for Prosthetic Control: Feasibility of ISPU-Based Solutions for a Robotic Extra Limb
In collaboration with the University of Siena, our first work explored the potential and limitations of real-time, low-power control using Deep Edge AI. We deployed a neural network directly on a sensor (IMU with integrated ISPU) to control a robotic sixth finger. Results were promising: inference increased power consumption by only 12.5 % over the baseline signal acquisition. A major step toward intelligent, ultra-low-power wearables.

📌 2. A Feasibility Assessment on the Use of FBG Sensors for Pulse Wave Detection
Together with the University of Parthenope, we conducted a feasibility assessment of plethysmographic signals from an FBG sensor, comparing it to a commercial PPG system. Key features, such as heart rate tracking and overall cardiac activity, were evaluated, demonstrating that FBG signals can reliably follow cardiac dynamics within a negligible error margin. We also compared two filtering techniques, addressing a gap in the literature for FBG-based bio-signals.

📌 3. Fiber Optic-Based Method for Sensing Blood Perfusion in Intracranial Tissues
Our third work proposes a fiber-optic sensing method to detect blood perfusion in intracranial tissues, even during minimally invasive biopsies. The system uses a lock-in approach to enhance signal accuracy and was tested on a custom phantom with simulated blood flow. Results confirm its ability to detect blood movement with high sensitivity, supporting compatibility with biopsy needles.