arXiv subjects

Adrien Desjardins

Publications and source records attributed to Adrien Desjardins.

4 recordsLinked to original sources

arXiv2026

Dynamic compensation of diffusion-limited oxygen sensing with deep learning

Luminescence-based oxygen sensors suffer from a fundamental trade-off between mechanical robustness and temporal response; polymers that encapsulate the sensing dye also act as diffusion barriers that compromise real-time monitoring. Here, we show that this bottleneck can be computationally mitigated using spatially resolved imaging and deep learning. We develop a Temporal Vision Transformer (TViT) architecture that processes consecutive frames from a low-cost platform consisting of a Raspberry Pi camera, UV LED, and porous PtOEP/polystyrene film. Trained against a high-speed reference sensor, the TViT reduced mean absolute error (MAE) by up to 96% and improved T90 response times by 91%, relative to the classical two-site Stern-Volmer model. We benchmarked seven neural architectures, including physics-informed and curriculum-learning variants. Physical plausibility of predictions was also assessed using a Rauch-Tung-Striebel Kalman smoother. The data-driven TViT achieved the highest performance and strongest physical plausibility, suggesting temporal attention can implicitly capture Fickian diffusion, but physics-informed models minimised temporal lag. Validated under both gaseous and biofouled aqueous conditions, the framework demonstrated robust generalisation across different setups, environments, biofilm states, and dynamic unstructured oxygen conditions. These capabilities are fundamentally beyond classical sensor operation, establishing a new IoT-compatible paradigm for computationally compensated diffusion-limited chemical sensing.

eess.IV↗

arXiv2026

Real-time 3D Ultrasonic Needle Tracking with a Photoacoustic Beacon

Many minimally invasive procedures, such as core needle biopsy of focal liver lesions, nerve blocks, and fetal and vascular interventions, are typically performed under ultrasound guidance, which provides real-time, high-resolution visualisation of tissue anatomy. Accurate and efficient localisation of the needle tip relative to patient anatomy is essential for guiding the needle towards the procedure target, avoiding adverse events and reducing the need for repeat procedures. However, the 3D nature of the procedure and poor image contrast of the needle in heterogeneous tissue or at steep insertion angles often lead to confusion over the true location of the tip within the 2D guidance images, and existing methods to enhance needle visibility largely remain limited to 2D. Here, we present a novel interventional ultrasound system capable of 2D B-mode imaging and 3D needle tracking. The tip location is determined from the time-of-flight of ultrasound generated by a photoacoustic beacon embedded in the needle bevel and received by a sparse receiver array distributed around the imaging system's curvilinear ultrasound probe. The measured tracking accuracy was better than 2 mm for depths up to 140 mm in water, and approximately 2 mm on average in an ex vivo tissue phantom, with referenced positions derived from X-ray computed tomography. In a usability study involving 12 clinicians performing biopsy procedures in a ex vivo tissue phantom, the failure rate was reduced by 35%, from 15.8% to 10.3% after only a few minutes of training. These results demonstrate that the proposed system has strong potential to support a wide range of minimally invasive procedures by enabling clinicians to accurately target small anatomical structures, improving the efficiency and effectiveness of diagnostic sampling and therapeutic delivery or ablation, and reducing the risk of adverse events.

physics.med-ph↗

arXiv2026

Deep Learning-Enabled Dissolved Oxygen Sensing in Biofouling Environments for Ocean Monitoring

The escalating climate crisis and ecosystem degradation demand intelligent, low-cost sensors capable of robust, long-term monitoring in real-world environments. Absolute dissolved oxygen (DO) concentration is a key parameter for predicting climate tipping points. Inexpensive optoelectronic sensors based on microstructured polymer films doped with phosphorescent dyes could be readily deployable; however, signal drift and marine biofouling remain major challenges. Here, we introduce a sensing paradigm that combines camera-based DO sensors with a visual transformer (ViT)-based physics-informed neural network (PINN) for high-fidelity sensing under biofouling conditions. Training and testing data were obtained from an algae-laden water tank over 14 days to capture accelerated biofouling. The ViT-PINN, which embeds the Stern-Volmer (SV) equation into the loss function, reduces mean average error (MAE) by 92% and 89% compared to classical statistical and ML approaches, achieving ~2 umol/L absolute error. A deep ensemble further quantifies predictive uncertainty, enabling self-diagnostic sensing.

eess.IV↗

arXiv2022

Neural Network Kalman filtering for 3D object tracking from linear array ultrasound data

Many interventional surgical procedures rely on medical imaging to visualise and track instruments. Such imaging methods not only need to be real-time capable, but also provide accurate and robust positional information. In ultrasound applications, typically only two-dimensional data from a linear array are available, and as such obtaining accurate positional estimation in three dimensions is non-trivial. In this work, we first train a neural network, using realistic synthetic training data, to estimate the out-of-plane offset of an object with the associated axial aberration in the reconstructed ultrasound image. The obtained estimate is then combined with a Kalman filtering approach that utilises positioning estimates obtained in previous time-frames to improve localisation robustness and reduce the impact of measurement noise. The accuracy of the proposed method is evaluated using simulations, and its practical applicability is demonstrated on experimental data obtained using a novel optical ultrasound imaging setup. Accurate and robust positional information is provided in real-time. Axial and lateral coordinates for out-of-plane objects are estimated with a mean error of 0.1mm for simulated data and a mean error of 0.2mm for experimental data. Three-dimensional localisation is most accurate for elevational distances larger than 1mm, with a maximum distance of 6mm considered for a 25mm aperture.

stat.AP↗