An Ultra-Low Power Reconfigurable Biomedical AI Processor With Adaptive Learning for Versatile Wearable Intelligent Health Monitoring.

Wearable intelligent health monitoring devices with on-device biomedical AI processor can be used to detect the abnormity in users' biomedical signals (e.g., ECG arrythmia classification, EEG-based seizure detection). This requires ultra-low power and reconfigurable biomedical AI processor to support battery-supplied wearable devices and versatile intelligent health monitoring applications while achieving high classification accuracy. However, existing designs have issues in meeting one or more of the above requirements. In this work, a reconfigurable biomedical AI processor (named BioAIP) is proposed, mainly featuring: 1) a reconfigurable biomedical AI processing architecture to support versatile biomedical AI processing. 2) an event-driven biomedical AI processing architecture with approximate data compression to reduce the power consumption. 3) an AI-based adaptive-learning architecture to address patient-to-patient variation and improve the classification accuracy. The design has been implemented and fabricated using a 65nm CMOS process technology. It has been demonstrated with three typical biomedical AI applications, including ECG arrythmia classification, EEG-based seizure detection and EMG-based hand gesture recognition. Compared with the state-of-the-art designs optimized for single biomedical AI tasks, the BioAIP achieves the lowest energy per classification among the designs with similar accuracy, while supporting various biomedical AI tasks.

4D patient dose reconstruction using online measured EPID cine images for lung SBRT treatment validation.

PURPOSE: This study aims to develop an EPID-guided 4D patient dose reconstruction framework and to investigate its feasibility for lung SBRT treatment validation. METHODS: Both the beam apertures and tumor movements were detected based on the continuously acquired EPID images during the treatment. Instead of directly using the transit photon fluence measured by the EPID, this method reconstructed the entrance fluence with the measured beam apertures and the delivered MUs. The entrance fluence distributions were sorted into their corresponding phases based on the detected tumor motion pattern and then accumulated for each phase. Together with the in-room 4DCT taken before every treatment to consider the interfractional-motion, the entrance fluence was then used for the patient dose calculation. Deformable registration was performed to sum up the phase doses for final treatment assessment. The feasibility of using the transit EPID images for entrance fluence reconstruction was evaluated against EPID in-air measurements. The accuracy of 3D- and 4D-dose reconstruction was validated by experiments with a motor-driven cylindrical diode array for six clinical-SBRT plans. RESULTS: The average difference between the measured and reconstructed fluence maps was within 0.16%. The reconstructed 3D-dose showed a less than 1.4% difference for the CAX-dose and at least a 98.3% gamma-passing-rate (2%∕2 mm) for the peripheral dose. Distorted dose distributions were observed in the measurement with the moving phantom. The comparison between the measured and the reconstructed 4D-dose without considering temporal information failed the gamma-evaluation for most cases. In contrast, when temporal information was considered, the dose distortion phenomena were successfully represented in the reconstructed dose (97.6%-99.7% gamma-passing rate). CONCLUSIONS: The proposed method considered uncertainties of the beam delivery system, the interfractional- and intrafractional-motion, and the interplay effect. The experimental validation demonstrates that this method is practical and accurate for online or offline SBRT patient dose verification.

A High Accuracy & Ultra-Low Power ECG-Derived Respiration Estimation Processor for Wearable Respiration Monitoring Sensor.

The respiratory rate is widely used for evaluating a person's health condition. Compared to other invasive and expensive methods, the ECG-derived respiration estimation is a more comfortable and affordable method to obtain the respiration rate. However, the existing ECG-derived respiration estimation methods suffer from low accuracy or high computational complexity. In this work, a high accuracy and ultra-low power ECG-derived respiration estimation processor has been proposed. Several techniques have been proposed to improve the accuracy and reduce the computational complexity (and thus power consumption), including QRS detection using refractory period refreshing and adaptive threshold EDR estimation. Implemented and fabricated using a 55 nm processing technology, the proposed processor achieves a low EDR estimation error of 0.73 on CEBS database and 1.2 on MIT-BIH Polysomnographic Database while demonstrating a record-low power consumption (354 nW) for the respiration monitoring, outperforming the existing designs. The proposed processor can be integrated in a wearable sensor for ultra-low power and high accuracy respiration monitoring.