1 | José Ignacio Orlando

Semi-supervised learning with Noisy Students improves domain generalization in optic disc and cup segmentation in uncropped fundus images

The paper evaluates domain generalization strategies for optic disc and cup segmentation in fundus images, highlighting issues with existing methods when applied to uncropped images, and proposes a semi-supervised learning approach based on the Noisy Student framework to improve performance across diverse datasets.

Eugenia Moris, Ignacio Larrabide, José Ignacio Orlando

Learning normal asymmetry representations for homologous brain structures

The paper presents a novel method for learning normal asymmetry patterns in brain structures. It accurately characterizes normal asymmetries and detects pathological alterations without relying on diseased cases for training. The approach shows promise in improving the identification of neurodegenerative conditions..

Duilio Deangeli, Emmanuel Iarussi, Juan Pablo Princich, Mariana Bendersky, Ignacio Larrabide, José Ignacio Orlando

A ResNet is All You Need? Modeling A Strong Baseline for Detecting Referable Diabetic Retinopathy in Fundus Images

With no other methodological innovation than a carefully designed training, our ResNet model achieved an AUC = 0.955 (0.953 - 0.956) on a combined test set of 61007 test images from different public datasets, which is in line or even better than what other more complex deep learning models reported in the literature.

Tomás Castilla, Marcela S. Martínez, Mercedes Leguía, Ignacio Larrabide, José Ignacio Orlando

A deep learning model for brain vessel segmentation\ in 3DRA with arteriovenous malformations

We train the first deep learning model for segmenting brain arteries from 3D rotational angiographies in cases with brain arterio-venous malformations.

Camila García, Yibin Fang, Jianmin Liu, Ana Paula Narata, José Ignacio Orlando, Ignacio Larrabide

$A deep learning model for brain vessel segmentation\ in 3DRA with arteriovenous malformations$

Assessing Coarse-to-Fine Deep Learning Models for Optic Disc and Cup Segmentation in Fundus Images

We experimentally validate whether using coarse-to-fine models instead of one-stage models is appropriate or not for segmenting the optic disc and the optic cup in color fundus images. We observed that models trained with the right amount of data can perform much better than coarse-to-find approaches.

Eugenia Moris, Nicolás Dazeo, María Paula Albina De Rueda, Francisco Filizzola, Nicolás Iannuzzo, Danila Nejamkin, Kevin Wignall, Mercedes Leguía, Ignacio Larrabide, José Ignacio Orlando

Assessing Coarse-to-Fine Deep Learning Models for Optic Disc and Cup Segmentation in Fundus Images

Multiclass Segmentation as Multitask Learning for Drusen Segmentation in Retinal Optical Coherence Tomography

We posed a multiclass segmentation task as a single multitask model with binary segmentation targets. Our results indicate that this approach might be useful to deal with “sandwiched” structures.

Rhona Asgari, José Ignacio Orlando, Sebastian M. Waldstein, Ferdinand Schlanitz, Magdalena Baratsits, Ursula Schmidt-Erfurth, Hrvoje Bogunović

An Amplified-Target Loss Approach for Photoreceptor Layer Segmentation in Pathological OCT Scans

We introduce an augmented target loss function framework for photoreceptor layer segmentation that penalizes errors in the central area of each B-scan. It allows to significantly improve performance with respect to the standard loss functions.

José Ignacio Orlando, Anna Breger, Hrvoje Bogunović, Sophie Riedl, Bianca S. Gerendas, Martin Ehler, Ursula Schmidt-Erfurth

Foveal Avascular Zone Segmentation in Clinical Routine Fluorescein Angiographies Using Multitask Learning

We introduced a multitask learning y-shaped neural network that simultaneously segment the FAZ in FA images and predict a distance map. This extra branch aids to improve results in clinical routine images.

Dominik Hofer, José Ignacio Orlando, Philipp Seeböck, Georgios Mylonas, Felix Goldbach, Amir Sadeghipour, Bianca S. Gerendas, Ursula Schmidt-Erfurth

U2-Net: A Bayesian U-Net model with epistemic uncertainty feedback for photoreceptor layer segmentation in pathological OCT scans

We developed a Bayesian U-Net model for photoreceptor layer segmentation in OCT that predicts epistemic uncertainty maps highlighting potential areas of error in the segmentation.

José Ignacio Orlando, Philipp Seeböck, Hrvoje Bogunović, Sophie Klimscha, Christoph Grechenig, Sebastian M. Waldstein, Bianca S. Gerendas, Ursula Schmidt-Erfurth

Using CycleGANs for effectively reducing image variability across OCT devices and improving retinal fluid segmentation

We used CycleGANs to translate OCT images from one vendor to another. This approach allows us to increase the performance of fluid segmentation models trained on one vendor and evaluated on another.

Philipp Seeböck, David Romo-Bucheli, Sebastian Waldstein, Hrvoje Bogunović, José Ignacio Orlando, Bianca S. Gerendas, Georg Langs, Ursula Schmidt-Erfurth