Tyndall National Institute - Conference Items
Permanent URI for this collection
Browse
Recent Submissions
Item Machine learning approaches for EM signature analysis in chipless RFID technology(Institute of Electrical and Electronics Engineers (IEEE), 2024) Rather, Nadeem; Simorangkir, Roy B. V. B.; Buckley, John L.; O’Flynn, Brendan; Tedesco, Salvatore; Science Foundation Ireland; Enterprise Ireland; European Regional Development FundIn this paper, for the first time, we provide a comprehensive review of Machine Learning (ML) approaches in Chipless Radio Frequency Identification (CRFID) technology, which is a fast-developing sector with applications in inventory management, anti-counterfeiting, health monitoring, and environmental monitoring, to name a few. ML techniques are rapidly being integrated to improve CRFID systems’ capabilities for robust detection of information. The combination of ML with CRFID technology is presented, examining various ML approaches, applications, challenges, and future perspectives. It is observed that ML has been successfully deployed in CRFID with high accuracy in the detection of information from CRFID tags. Challenges, such as data quality, security, and scalability are identified. Moreover, the literature currently struggles in the application of ML models on high-capacity tags, and lacks standardized data collection and sharing methodologies. We suggest the development of common data collection protocols, data sharing initiatives, and collaboration to establish a cohesive framework for CRFID data-driven research.Item Adaptive neural network equalisation using skip connections for future 100 Gbit/s/λ passive optical networks(Institution of Engineering and Technology, 2023) Murphy, Stephen L.; Jamali, Fariba; Townsend, Paul D.; Antony, Cleitus; Science Foundation IrelandWe present a novel equaliser based on a neural network with skip connections for 100 Gbit/s PAM4 SOA-preamplfied PONs which can converge within 2000 symbols on a burst-by-burst basis, and effectively compensate SOA patterning and 81.6 ps/nm fiber dispersion, enabling 21 dB system dynamic range.Item Improved high permeability CoZrTaB laminated thin films with novel CMOS compatible dielectric material(Institute of Electrical and Electronics Engineers (IEEE), 2023-09-04) Wei, Guannan; Das, Rajasree; Lordan, Daniel; Sai, Ranajit; Hayes, Mike; Lorenc, Marek; Clarke, Barry; Hurley, David; McCloskey, Paul; Enterprise IrelandThis paper present an optimized CoZrTaB-based laminated thin films with a novel wet etch-able oxide dielectric material. Wet etching capability was studied on the stack material exhibiting a narrow and clean undercut. Good uniaxial anisotropy with low coercivity was achieved via in-situ magnetic alignment during magnetron sputtering. Permeability of 432 and Q-factor of 23.4 at 100 MHz were observed in high frequency permeameter measurement. Finally thermal annealing was carried out at various temperatures. Uniaxial anisotropy was maintained up to 300 °C, while an enhancement of permeability (by 25%) was observed.Item Development of a PPG-based hardware and software system deployable on elbow and thumb for real-time estimation of pulse transit time(IEEE, 2023-07-24) Valerio, Andrea; Hajzeraj, Adhurim; Talebi, Omid Varnosfaderani; Belcastro, Marco; Tedesco, Salvatore; Demarchi, Danilo; O'Flynn, Brendan; Enterprise Ireland; Science Foundation Ireland; European Regional Development FundBlood pressure (BP) is a vital parameter used by clinicians to diagnose issues in the human cardiovascular system. Cuff-based BP devices are currently the standard method for on-the-spot and ambulatory BP measurements. However, cuff-based devices are not comfortable and are not suitable for long-term BP monitoring. Many studies have reported a significant correlation between pulse transit time (PTT) with blood pressure. However, this relation is impacted by many internal and external factors which might lower the accuracy of the PTT method. In this paper, we present a novel hardware system consisting of two custom photoplethysmography (PPG) sensors designed particularly for the estimation of PTT. In addition, a software interface and algorithms have been implemented to perform a real-time assessment of the PTT and other features of interest from signals gathered between the brachial artery and the thumb. A preclinical study has been conducted to validate the system. Five healthy volunteer subjects were tested and the results were then compared with those gathered using a reference device. The analysis reports a mean difference among subjects equal to -3.75±7.28 ms. Moreover, the standard deviation values obtained for each individual showed comparable results with the reference device, proving to be a valuable tool to investigate the factors impacting the BP-PTT relationship.Clinical Relevance— The proposed system proved to be a feasible solution to detect blood volume changes providing good quality signals to be used in the study of BP-PTT relationship.Item Acoustic emissions and age-related changes of the knee(IEEE, 2023-07) Khokhlova, Liudmila; Dimitrios-Sokratis, Komaris; O'Flynn, Brendan; Tedesco, Salvatore; Science Foundation Ireland; European Regional Development FundAcoustic emission (AE) monitoring is currently being widely investigated as a diagnostic tool in orthopedics, in particular for osteoarthritis (OA) diagnostics. Considering that age is one of the main risk factors for OA, investigating age-related changes in joint AEs might provide an additional incentive for further studies and consequent translation to clinical practice. The aim of this study is to investigate age-related changes in knee AE and determine AE hit definition modes as well as AE hit parameters that allow for improved age group differentiation. Knee AEs were recorded from 51 participants in two age groups (18-35 and 50-75 years old) whilst cycling with 30 and 60 rpm cadence. Two AE sensors with 15-40 kHz and 100-450 kHz frequency ranges were used, and three AE event detection modes investigated. Additionally, participants’ Knee Osteoarthritis Outcome Scores (KOOS) were recorded. Low frequency sensors (15-40kHz) and hit modes with shortened hit and peak definition times showed the potential to distinguish between age groups. Moreover, a weak correlation was found between only three parameters (AE event median duration, rise time, and signal strength) and age, indicating that changes in joint AE are most likely associated with pathological changes rather than physiological ageing within the healthy norm.Clinical Relevance— the use of AE monitoring was examined in the context of age-related changes in knee health. The study indicates the potential for knee AE monitoring to be used as a quantitative measure of pathological changes in the knee status.