no code implementations • 4 Sep 2023 • Luca Sestini, Benoit Rosa, Elena De Momi, Giancarlo Ferrigno, Nicolas Padoy
In this work, we develop a framework for instance segmentation not relying on spatial annotations for training.
no code implementations • 21 Dec 2022 • Jorge F. Lazo, Benoit Rosa, Michele Catellani, Matteo Fontana, Francesco A. Mistretta, Gennaro Musi, Ottavio De Cobelli, Michel de Mathelin, Elena De Momi
We address the challenge of tissue classification when annotations are available only in one domain, in our case WLI, and the endoscopic images correspond to an unpaired dataset, i. e. there is no exact equivalent for every image in both NBI and WLI domains.
no code implementations • 1 Jul 2022 • Jorge F. Lazo, Chun-Feng Lai, Sara Moccia, Benoit Rosa, Michele Catellani, Michel de Mathelin, Giancarlo Ferrigno, Paul Breedveld, Jenny Dankelman, Elena De Momi
Navigation inside luminal organs is an arduous task that requires non-intuitive coordination between the movement of the operator's hand and the information obtained from the endoscopic video.
no code implementations • 16 Feb 2022 • Luca Sestini, Benoit Rosa, Elena De Momi, Giancarlo Ferrigno, Nicolas Padoy
We then use the obtained instrument masks as pseudo-labels in order to train a per-frame segmentation model; to this aim, we develop a learning-from-noisy-labels architecture, designed to extract a clean supervision signal from these pseudo-labels, leveraging their peculiar noise properties.
no code implementations • 2 Dec 2021 • Guiqiu Liao, Oscar Caravaca-Mora, Benoit Rosa, Philippe Zanne, Alexandre Asch, Diego Dall Alba, Paolo Fiorini, Michel de Mathelin, Florent Nageotte, Michalina J. Gora
The non-constant rotation speed of optical components in the OCT catheter tip causes rotational distortion in OCT volumetric scanning.
no code implementations • 8 Apr 2021 • Jorge F. Lazo, Sara Moccia, Aldo Marzullo, Michele Catellani, Ottavio De Cobelli, Benoit Rosa, Michel de Mathelin, Elena De Momi
In this work we study the implementation of 3 different Convolutional Neural Networks (CNNs), using a 2-steps training strategy, to classify images from the urinary tract with and without lesions.
no code implementations • 5 Apr 2021 • Jorge F. Lazo, Aldo Marzullo, Sara Moccia, Michele Catellani, Benoit Rosa, Michel de Mathelin, Elena De Momi
Of these, two architectures are taken as core-models, namely U-Net based in residual blocks($m_1$) and Mask-RCNN($m_2$), which are fed with single still-frames $I(t)$.
no code implementations • 28 Feb 2021 • Luca Sestini, Benoit Rosa, Elena De Momi, Giancarlo Ferrigno, Nicolas Padoy
3-D pose estimation of instruments is a crucial step towards automatic scene understanding in robotic minimally invasive surgery.
no code implementations • 13 Jan 2021 • Jorge F. Lazo, Aldo Marzullo, Sara Moccia, Michele Catellani, Benoit Rosa, Michel de Mathelin, Elena De Momi
For the training of these networks, we analyze the use of two different color spaces: gray-scale and RGB data images.