MICLAB22-ADM – MICLab

Paper accepted on SIPAIM 2024!

MICLAB22-ADM — Wed, 27 Nov 2024 19:02:18 +0000

Our lab students Diedre Carmo and Wesna Araujo attended the conference in Antigua, Guatemala, to present our work. The papers will be published in the IEEE Xplore Digital Library.

Our group had three papers accepted and presented at the 2024 edition of the International Symposium on Medical Information Processing and Analysis (SIPAIM), in Antigua, Guatemala:

“Assessing MedSAM’s Generalization to White Matter Hyperintensities in Magnetic Resonance Imaging”. Joany Rodrigues, Diedre Santos, and Leticia Rittner.
“Enhancing fMRS Data Quality and GABA Quantification Using SpectroViT: A Deep Learning Reconstruction Approach”. Wesna Araujo, Ricardo Landim, Thiago Costa, Elvis Silva, Gabriela Castellano and Leticia Rittner.
“Training a ViT model on shape signatures for quality control of MRI corpus callosum segmentations”. Suellen Silva and Leticia Rittner.

Our lab students, Diedre Carmo and Wesna Araujo, had the opportunity to attend the conference and present their work.

Paper accepted on MICCAI 2024!

MICLAB22-ADM — Wed, 27 Nov 2024 17:35:48 +0000

The MICCAI 2024 International Workshop on Computational Diffusion MRI (CDMRI – Marrakesh, Morocco) brought together leading researchers and clinicians to discuss advancements in diffusion MRI. This imaging method can reveal details of brain tissue microstructure and connectivity, with applications in neurodevelopment, aging, and neurosurgery. The event featured keynotes and original research presentations that highlighted the growing potential of diffusion MRI in clinical and research contexts.

Our group presented the paper “Corpus Callosum Parcellation Methods: What Can Tractography Tell Us About Them?”, investigating the consistency of different methods for parcellating the Corpus Callosum. Using advanced diffusion MRI techniques, the study examined the connectivity patterns of this crucial brain structure. The research is the product of an international collaboration between our team in Brazil (Caio Santana, Simone Appenzeller, Leticia Rittner) and colleagues from Chile (Claudio Román, Pamela Guevara), showcasing the global nature of innovation in medical imaging.

Workshop Beyond RGB: Navigating the Multispectral Universe

MICLAB22-ADM — Fri, 12 Jul 2024 13:30:30 +0000

Last week, we hosted the workshop “Beyond RGB: Navigating the Multispectral Universe“. The goal of this meeting was to share the experiences and learning accumulated over the last 12 months by the Visual Computing group, led by DCA reasearchers (Prof. Paula Dornhofer, Prof. Denis Fantinato and Prof. Letícia Rittner) and supported by Eldorado Research Institute.

Paper accepted on ISBI 2024!

MICLAB22-ADM — Tue, 21 May 2024 15:28:02 +0000

Our group had two papers accepted on the 21st IEEE International Symposium on Biomedical Imaging (ISBI):

“FOD-Swin-Net: angular super resolution of fiber orientation distribution using a transformer-based deep model” – by Mateus Oliveira
“Classification of pulmonary embolism on computed tomography angiography using artificial intelligence” – by Beatriz Celante

Members of our lab will be travelling to Athens (Greece) to present our work.

Paper awarded on ICASSP 2024!

MICLAB22-ADM — Fri, 17 May 2024 15:10:57 +0000

Our Rainbow.AI team, led by Letícia Rittner, Paula Paro Costa and Denis Fantinato, has been awarded in the Grand Challenge on Hyperspectral Skin Vision from the largest and most important conference of the IEEE Signal Processing Society: International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2024)!

Our solution, based on a Vision Transformer, was able to reconstruct an hyperspectral image with 61 channels from a simple RGB + NIR image (4 channels). The research was partially funded by the Instituto de Pesquisas Eldorado.

Paper accepted on ICASSP 2024!

MICLAB22-ADM — Mon, 05 Feb 2024 21:01:43 +0000

Our group had a paper accepted to be presented on the International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2024): “Vision Transformer MST++: Efficient Hyperspectral Skin Reconstruction”.

Members of our lab will be traveling to the conference to present our work.

MICLab wins Scientific Merit Award, ranking among the best undergraduate research works at the XXXI Congresso de Iniciação Científica

MICLAB22-ADM — Fri, 01 Dec 2023 16:45:19 +0000

This time the Scientific Merit Award at the XXXI Congresso de Iniciação Científica at Unicamp was given to the work entitled “Development of a low-cost multispectral camera for analyzing skin lesions in lupus patients”. The work developed by Nathan Shen Baldon under the supervision of Prof. Leticia Rittner, discusses that different tissues have different spectral signatures, which can be used to classify different regions of a sample. It was possible to show that the developed prototype has the potential to be applied in medicine, for instance, to analyze and characterize skin lesions. This work had also the contribution of Prof. Simone Appenzeller, from the Medical Science School.

Papers accepted on SIPAIM 2023

MICLAB22-ADM — Fri, 01 Dec 2023 16:39:33 +0000

Our group had two papers accepted and presented on the 19th International Symposium on Medical Information Processing and Analysis (SIPAIM):

“Magnetic Resonance Spectroscopy Fitting through Spectrogram and Vision Transformers”. Marcio Almeida, Gabriel Dias, Mateus Oliveira and Leticia Rittner
“DTI-based Corpus Callosum parcellation using the Tensorial Morphological Gradient and Self-Organizing Maps”. Caio Santana, Thays Santos, Joany Rodrigues, Paulo Julio, Simone Appenzeller and Leticia Rittner

Members of our lab traveled to these conferences to present our work.

First place in MRS Reconstruction Challenge at ISBI 2023!

MICLAB22-ADM — Tue, 25 Apr 2023 12:41:46 +0000

I couldn’t be more proud of my students Gabriel Dias, Mateus Oliveira, that together with Paula Dornhofer Paro Costa and Lucas Hideki Ueda, got the first place on #ISBI2023 challenge on MRS Reconstruction! Our team, the Deep Spectral Divers, built a model able to reconstructs GABA spectra from Edited Magnetic Resonance Spectroscopy (MRS) scans using only 25% of the data required by current methods. The model uses spectograms as input for the reconstruction task. It is nice to see FEEC – Unicamp in the spotlight of such an important conference!

Thank you Hanna B., Rodrigo Pommot Berto, Roberto Souza and Ashley Harris for organizing the challenge!

MICLab wins the Best Paper Award (Master degree student) at the BRAINN Congress for the third year in a row

MICLAB22-ADM — Tue, 25 Apr 2023 12:41:17 +0000

This time the award for the best paper in the Master Degree category at the BRAINN Congress was given to the work entitled “Volumetric Segmentation of the Corpus Callosum: comparing 2D, extended-2D, and 3D approaches to train the model on anisometric diffusion MRI“. This work, is part of the MSc dissertation of Joany Rodrigues, and discusses different CNN models trained to segment the structure. This work also had the contribution of: Paulo Julio, Simone Appenzeller and Letícia Rittner.