Logo do repositório

Percorrer por autor "Vieira, Ana"

A mostrar 1 - 2 de 2
  • Abdominal MRI synthesis using styleGAN2-ADA
    Publication . Gonçalves, Bernardo; Vieira, Pedro; Vieira, Ana
    The lack of labeled medical data still poses one of the biggest issues when creating Deep Learning models in the medical field. Modern data augmentation techniques like the generation of synthetic images have gained a special interest. In recent years there has been a significant improvement in GANs. StyleGAN2 achieves impressive results in the generation of natural images. StyleGAN2-ADA was created to respond to the lack of training data when training an image synthesis model, which is very frequent in the medical field. Some works used styleGAN to generate melanomas, breast cancer histological images, and MR and CT images. In this work, we apply, for the first time, a styleGAN2-ADA to a small dataset of abdominal MRI with 1.3k images. From the augmentation pipeline created by the authors of styleGAN2-ADA, we removed all augmentations except the geometric transformations and pixel blitting operations. We trained our network for 70 hours. Our generated dataset has a precision score of 59,33 % and a FID score of 18,14. We conclude that the styleGAN2-ADA is a viable solution to generate MRI using a small dataset.
    2023-07Artigo científico Acesso aberto Ver mais
  • PORTRESS – the PORTuguese Reuma.pt registry for Sjögren’S disease
    Publication . Bandeira, Matilde; Silvério-António, Manuel; Costa, Roberto Pereira da; Lopes, Ana Rita; Santos, Filipe Cunha; Pereira, Paulo J.; Raimundo, Diana Belchior; Cunha, Anita; Oliveira, Cláudia Pinto; Duarte, Ana Catarina; Dias, João; Santos, Mariana Emília; Gonçalves, Maria João; Moniz, Ana Catarina; Maduro, Ana Isabel; Luís, Mariana; Valido, Ana; Oliveira, Margarida; Brites, Luísa; Tenazinha, Catarina; Vieira, Ana; Khmelinskii, Nikita; Barcelos, Filipe; Fonseca, João Eurico; Romão, Vasco C.
    Aims: Sjögren's disease (SjD) is a complex disease with a wide variety of manifestations and outcomes. We recently created PORTRESS, the Portuguese SjD registry within Reuma.pt. We aim to describe this registry and characterize our national cohort. Methods: We included patients with a clinical diagnosis of SjD, registered in PORTRESS up to November 2023. Demographic, clinical, treatment, and patient-reported outcomes (PROs) data were collected. Variables were compared according to parametric or non-parametric tests, as applicable. Results: A total of 1375 patients were included. Patients fulfilled AECG 2002 or ACR/EULAR 2016 classification criteria in 62% and 57% of cases, respectively, although more than half didn't have a complete assessment of all items. Of note, the vast majority (93%) had both SjD manifestations and a positive anti-Ro and/or minor salivary gland biopsy. Most patients (88%) exhibited at least one active ESSDAI domain during the course of their disease. Hydroxychloroquine and corticosteroids were used in 52% and 30% of patients, while other immunosuppressants and pilocarpine were used in 12% and 18% of cases, respectively. The mean ESSDAI at inclusion was 3.0±4.4 (range 0-42), and, at the last follow-up, 2.1±3.7 (0-31), corresponding to a significant decrease. Dryness, pain, and fatigue PROs were scored high, with a significant increase from baseline to follow-up. Conclusion: PORTRESS is a web-based SjD registry facilitating efficient nationwide data storage. It enables research, trial recruitment, and a comprehensive longitudinal view of patients' evolution. Although systemic activity improved over follow-up, symptom burden worsened when compared to baseline, underlining a major unmet need in SjD.
    2025-09Artigo científico Acesso aberto Ver mais