arXiv subjects

Julien Bec

Publications and source records attributed to Julien Bec.

3 recordsLinked to original sources

arXiv2025

Pixel Super-Resolved Fluorescence Lifetime Imaging Using Deep Learning

Fluorescence lifetime imaging microscopy (FLIM) is a powerful quantitative technique that provides metabolic and molecular contrast, offering strong translational potential for label-free, real-time diagnostics. However, its clinical adoption remains limited by long pixel dwell times and low signal-to-noise ratio (SNR), which impose a stricter resolution-speed trade-off than conventional optical imaging approaches. Here, we introduce FLIM_PSR_k, a deep learning-based multi-channel pixel super-resolution (PSR) framework that reconstructs high-resolution FLIM images from data acquired with up to a 5-fold increased pixel size. The model is trained using the conditional generative adversarial network (cGAN) framework, which, compared to diffusion model-based alternatives, delivers a more robust PSR reconstruction with substantially shorter inference times, a crucial advantage for practical deployment. FLIM_PSR_k not only enables faster image acquisition but can also alleviate SNR limitations in autofluorescence-based FLIM. Blind testing on held-out patient-derived tumor tissue samples demonstrates that FLIM_PSR_k reliably achieves a super-resolution factor of k = 5, resulting in a 25-fold increase in the space-bandwidth product of the output images and revealing fine architectural features lost in lower-resolution inputs, with statistically significant improvements across various image quality metrics. By increasing FLIM's effective spatial resolution, FLIM_PSR_k advances lifetime imaging toward faster, higher-resolution, and hardware-flexible implementations compatible with low-numerical-aperture and miniaturized platforms, better positioning FLIM for translational applications.

cs.CV↗

arXiv2025

Autonomous labeling of surgical resection margins using a foundation model

Assessing resection margins is central to pathological specimen evaluation and has profound implications for patient outcomes. Current practice employs physical inking, which is applied variably, and cautery artifacts can obscure the true margin on histological sections. We present a virtual inking network (VIN) that autonomously localizes the surgical cut surface on whole-slide images, reducing reliance on inks and standardizing margin-focused review. VIN uses a frozen foundation model as the feature extractor and a compact two-layer multilayer perceptron trained for patch-level classification of cautery-consistent features. The dataset comprised 120 hematoxylin and eosin (H&E) stained slides from 12 human tonsil tissue blocks, resulting in ~2 TB of uncompressed raw image data, where a board-certified pathologist provided boundary annotations. In blind testing with 20 slides from previously unseen blocks, VIN produced coherent margin overlays that qualitatively aligned with expert annotations across serial sections. Quantitatively, region-level accuracy was ~73.3% across the test set, with errors largely confined to limited areas that did not disrupt continuity of the whole-slide margin map. These results indicate that VIN captures cautery-related histomorphology and can provide a reproducible, ink-free margin delineation suitable for integration into routine digital pathology workflows and for downstream measurement of margin distances.

cs.CV↗

arXiv2022

Optimizing the light penetration depth in APDs and SPADs for high gain-bandwidth and ultra-wide spectral response

Controlling light penetration depth in Avalanche Photodiodes (APDs) and Single Photon Avalanche Diodes (SPADs) play a major role in achieving high multiplication gain by delivering light near the multiplication region where the electric field is the strongest. Such control in the penetration depth for a particular wavelength of light has been previously demonstrated using integrated photon-trapping nanostructures. In this paper, we show that an optimized periodic nanostructure design can control the penetration depth for a wide range of visible and near-infrared wavelengths simultaneously. A conventional silicon APD structure suffers from high photocarrier loss due to recombination for shorter wavelengths as they are absorbed near the surface region, while silicon has low absorption efficiency for longer wavelengths. This optimized nanostructure design allows shorter wavelengths of light to penetrate deeper into the device, circumventing recombination sites while trapping the longer wavelengths in the thin silicon device by bending the vertically propagating light into horizontal modes. This manipulation of penetration depth improves the absorption in the device, increasing light sensitivity while nanostructures reduce the reflectance from the top surface. While delivery of light near the multiplication region reduces the photogenerated carrier loss and shortens transit time, leading to high multiplication gain in APDs and SPADs over a wide spectral range. These high gain APDs and SPADs will find their potential applications in Time-Of-Flight Positron Emission Tomography (TOF-PET), Fluorescence Lifetime Imaging Microscopy (FLIM), and pulse oximetry where high detection efficiency and high gain-bandwidth is required over a multitude of wavelengths.

physics.optics↗