Cookies Policy
The website need some cookies and similar means to function. If you permit us, we will use those means to collect data on your visits for aggregated statistics to improve our service. Find out More
Accept Reject
  • Menu
About

About

Sofia Cardoso Pereira was born in Lisbon, Portugal. She completed her BSc in Biology, specializing in cell biology and biotechnology, at the University of Lisbon in 2018. In 2020, she earned her MSc degree in Data Science from the University of Porto. Presently, Sofia is pursuing her Ph.D. at the University of Porto while working on her project titled "Explainable AI-based Decision Support Models for COVID-19 Detection." at the Centre for Biomedical Engineering Research (C-BER) within INESC TEC.

Interest
Topics
Details

Details

  • Name

    Sofia Cardoso Pereira
  • Role

    Research Assistant
  • Since

    01st September 2020
  • Nationality

    Portugal
  • Contacts

    +351222094106
    sofia.c.pereira@inesctec.pt
Publications

2024

STERN: Attention-driven Spatial Transformer Network for abnormality detection in chest X-ray images

Authors
Rocha, J; Pereira, SC; Pedrosa, J; Campilho, A; Mendonça, AM;

Publication
ARTIFICIAL INTELLIGENCE IN MEDICINE

Abstract
Chest X-ray scans are frequently requested to detect the presence of abnormalities, due to their low-cost and non-invasive nature. The interpretation of these images can be automated to prioritize more urgent exams through deep learning models, but the presence of image artifacts, e.g. lettering, often generates a harmful bias in the classifiers and an increase of false positive results. Consequently, healthcare would benefit from a system that selects the thoracic region of interest prior to deciding whether an image is possibly pathologic. The current work tackles this binary classification exercise, in which an image is either normal or abnormal, using an attention-driven and spatially unsupervised Spatial Transformer Network (STERN), that takes advantage of a novel domain-specific loss to better frame the region of interest. Unlike the state of the art, in which this type of networks is usually employed for image alignment, this work proposes a spatial transformer module that is used specifically for attention, as an alternative to the standard object detection models that typically precede the classifier to crop out the region of interest. In sum, the proposed end-to-end architecture dynamically scales and aligns the input images to maximize the classifier's performance, by selecting the thorax with translation and non-isotropic scaling transformations, and thus eliminating artifacts. Additionally, this paper provides an extensive and objective analysis of the selected regions of interest, by proposing a set of mathematical evaluation metrics. The results indicate that the STERN achieves similar results to using YOLO-cropped images, with reduced computational cost and without the need for localization labels. More specifically, the system is able to distinguish abnormal frontal images from the CheXpert dataset, with a mean AUC of 85.67% -a 2.55% improvement vs. the 0.98% improvement achieved by the YOLO-based counterpart in comparison to a standard baseline classifier. At the same time, the STERN approach requires less than 2/3 of the training parameters, while increasing the inference time per batch in less than 2 ms. Code available via GitHub.

2023

Lightweight multi-scale classification of chest radiographs via size-specific batch normalization

Authors
Pereira, SC; Rocha, J; Campilho, A; Sousa, P; Mendonca, AM;

Publication
COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE

Abstract
Background and Objective: Convolutional neural networks are widely used to detect radiological findings in chest radiographs. Standard architectures are optimized for images of relatively small size (for exam-ple, 224 x 224 pixels), which suffices for most application domains. However, in medical imaging, larger inputs are often necessary to analyze disease patterns. A single scan can display multiple types of radi-ological findings varying greatly in size, and most models do not explicitly account for this. For a given network, whose layers have fixed-size receptive fields, smaller input images result in coarser features, which better characterize larger objects in an image. In contrast, larger inputs result in finer grained features, beneficial for the analysis of smaller objects. By compromising to a single resolution, existing frameworks fail to acknowledge that the ideal input size will not necessarily be the same for classifying every pathology of a scan. The goal of our work is to address this shortcoming by proposing a lightweight framework for multi-scale classification of chest radiographs, where finer and coarser features are com-bined in a parameter-efficient fashion. Methods: We experiment on CheXpert, a large chest X-ray database. A lightweight multi-resolution (224 x 224, 4 48 x 4 48 and 896 x 896 pixels) network is developed based on a Densenet-121 model where batch normalization layers are replaced with the proposed size-specific batch normalization. Each input size undergoes batch normalization with dedicated scale and shift parameters, while the remaining parameters are shared across sizes. Additional external validation of the proposed approach is performed on the VinDr-CXR data set. Results: The proposed approach (AUC 83 . 27 +/- 0 . 17 , 7.1M parameters) outperforms standard single-scale models (AUC 81 . 76 +/- 0 . 18 , 82 . 62 +/- 0 . 11 and 82 . 39 +/- 0 . 13 for input sizes 224 x 224, 4 48 x 4 48 and 896 x 896, respectively, 6.9M parameters). It also achieves a performance similar to an ensemble of one individual model per scale (AUC 83 . 27 +/- 0 . 11 , 20.9M parameters), while relying on significantly fewer parameters. The model leverages features of different granularities, resulting in a more accurate classifi-cation of all findings, regardless of their size, highlighting the advantages of this approach. Conclusions: Different chest X-ray findings are better classified at different scales. Our study shows that multi-scale features can be obtained with nearly no additional parameters, boosting performance. (c) 2023 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license ( http://creativecommons.org/licenses/by-nc-nd/4.0/ )

2022

Mapping Cashew Orchards in Cantanhez National Park (Guinea-Bissau)

Authors
Pereira, SC; Lopes, C; Pedroso, JP;

Publication
REMOTE SENSING APPLICATIONS-SOCIETY AND ENVIRONMENT

Abstract
The forests and woodlands of Guinea-Bissau are a biodiversity hotspot under threat, which are progressively being replaced by cashew tree orchards. While the exports of cashew nuts significantly contribute to the gross domestic product and support local livelihoods, the country's natural capital is under significant pressure due to unsustainable land use. In this context, official entities strive to counter deforestation, but the problem persists, and there are currently no systematic or automated means for objectively monitoring and reporting the situation. Furthermore, previous remote sensing approaches failed to distinguish cashew orchards from forests and woodlands due to the significant spectral overlap between the land cover types and the highly intertwined structure of the cashew tree patches. This work contributes to overcoming such difficulty. It develops an affordable, reliable, and easy-to-use procedure based on machine learning models and Sentinel-2 images, automatically detecting cashew orchards with a dice coefficient of 82.54%. The results of this case study designed for the Cantanhez National Park are proof of concept and demonstrate the viability of mapping cashew orchards. Therefore, the work is a stepping stone towards wall-to-wall operational monitoring in the region.

2022

Attention-driven Spatial Transformer Network for Abnormality Detection in Chest X-Ray Images

Authors
Rocha, J; Pereira, SC; Pedrosa, J; Campilho, A; Mendonca, AM;

Publication
2022 IEEE 35TH INTERNATIONAL SYMPOSIUM ON COMPUTER-BASED MEDICAL SYSTEMS (CBMS)

Abstract
Backed by more powerful computational resources and optimized training routines, deep learning models have attained unprecedented performance in extracting information from chest X-ray data. Preceding other tasks, an automated abnormality detection stage can be useful to prioritize certain exams and enable a more efficient clinical workflow. However, the presence of image artifacts such as lettering often generates a harmful bias in the classifier, leading to an increase of false positive results. Consequently, healthcare would benefit from a system that selects the thoracic region of interest prior to deciding whether an image is possibly pathologic. The current work tackles this binary classification exercise using an attention-driven and spatially unsupervised Spatial Transformer Network (STN). The results indicate that the STN achieves similar results to using YOLO-cropped images, with fewer computational expenses and without the need for localization labels. More specifically, the system is able to distinguish between normal and abnormal CheXpert images with a mean AUC of 84.22%.

2021

Segmentation of COVID-19 Lesions in CT Images

Authors
Rocha, J; Pereira, S; Campilho, A; Mendonça, AM;

Publication
IEEE EMBS International Conference on Biomedical and Health Informatics, BHI 2021, Athens, Greece, July 27-30, 2021

Abstract
The worldwide pandemic caused by the new coronavirus (COVID-19) has encouraged the development of multiple computer-aided diagnosis systems to automate daily clinical tasks, such as abnormality detection and classification. Among these tasks, the segmentation of COVID lesions is of high interest to the scientific community, enabling further lesion characterization. Automating the segmentation process can be a useful strategy to provide a fast and accurate second opinion to the physicians, and thus increase the reliability of the diagnosis and disease stratification. The current work explores a CNN-based approach to segment multiple COVID lesions. It includes the implementation of a U-Net structure with a ResNet34 encoder able to deal with the highly imbalanced nature of the problem, as well as the great variability of the COVID lesions, namely in terms of size, shape, and quantity. This approach yields a Dice score of 64.1%, when evaluated on the publicly available COVID-19-20 Lung CT Lesion Segmentation GrandChallenge data set. © 2021 IEEE