A5060062918- Brain Tumor Detection and Classification
- Acute Ischemic Stroke Management
- Medical Image Segmentation Techniques
- AI in cancer detection
- Radiomics and Machine Learning in Medical Imaging
- Cerebrovascular and Carotid Artery Diseases
- Medical Imaging Techniques and Applications
- Advanced Neuroimaging Techniques and Applications
- Advanced MRI Techniques and Applications
- Intracerebral and Subarachnoid Hemorrhage Research
- Ultrasound Imaging and Elastography
- Image and Signal Denoising Methods
- Advanced Neural Network Applications
- Retinal Imaging and Analysis
- Image Processing Techniques and Applications
- Multiple Sclerosis Research Studies
- COVID-19 diagnosis using AI
- Advanced X-ray and CT Imaging
Universitat de Girona
2018-2024
The use of Computed Tomography (CT) imaging for patients with stroke symptoms is an essential step triaging and diagnosis in many hospitals. However, the subtle expression ischemia acute CT images has made it hard automated methods to extract potentially quantifiable information. In this work, we present evaluate deep learning tool lesion core segmentation from perfusion images. For evaluation, Ischemic Stroke Lesion Segmentation (ISLES) 2018 challenge dataset used that includes 94 cases...
Hemorrhagic stroke is the condition involving rupture of a vessel inside brain and characterized by high mortality rates. Even if patient survives, can cause temporary or permanent disability depending on how long blood flow has been interrupted. Therefore, it crucial to act fast prevent irreversible damage. In this work, deep learning-based approach automatically segment hemorrhagic lesions in CT scans proposed. Our based 3D U-Net architecture which incorporates recently proposed...
Automated methods for segmentation-based brain volumetry may be confounded by the presence of white matter (WM) lesions, which introduce abnormal intensities that can alter classification not only neighboring but also distant tissue. These lesions are common in pathologies where is an important prognostic marker, such as multiple sclerosis (MS), and thus reducing their effects critical improving volumetric accuracy reliability. In this work, we analyze effect WM on deep learning based tissue...
Background Manual brain extraction from magnetic resonance (MR) images is time‐consuming and prone to intra‐ inter‐rater variability. Several automated approaches have been developed alleviate these constraints, including deep learning pipelines. However, methods tend reduce their performance in unseen imaging (MRI) scanner vendors different protocols. Purpose To present evaluate for clinical use PARIETAL, a pre‐trained method. We compare its reproducibility scan/rescan analysis robustness...
Diffusion-weighted MRI (DWI) is essential for stroke diagnosis, treatment decisions, and prognosis. However, image disease variability hinder the development of generalizable AI algorithms with clinical value. We address this gap by presenting a novel ensemble algorithm derived from 2022 Ischemic Stroke Lesion Segmentation (ISLES) challenge. ISLES'22 provided 400 patient scans ischemic various medical centers, facilitating wide range cutting-edge segmentation research community. Through...
Segmenting tumors and their subregions is a challenging task as demonstrated by the annual BraTS challenge. Moreover, predicting survival of patient using mainly imaging features, while being desirable outcome to evaluate treatment patient, it also difficult task. In this paper, we present cascaded pipeline segment tumor its then use these results other clinical features together with image coming from pretrained VGG-16 network predict patient. Preliminary training validation dataset show...
The assessment of disease activity using serial brain MRI scans is one the most valuable strategies for monitoring treatment response in patients with multiple sclerosis (MS) receiving disease-modifying treatments. Recently, several deep learning approaches have been proposed to improve this analysis, obtaining a good trade-off between sensitivity and specificity, especially when T1-w T2-FLAIR images as inputs. However, need acquire two different types time-consuming, costly not always...
Acute stroke lesion segmentation tasks are of great clinical interest as they can help doctors make better informed treatment decisions. Magnetic resonance imaging (MRI) is time demanding but provide images that considered gold standard for diagnosis. Automated with an estimate the location and volume lesioned tissue, which in practice to assess evaluate risks each treatment. We propose a deep learning methodology acute sub-acute using multimodal MR imaging. The proposed method evaluated two...
Brain atrophy measurements derived from magnetic resonance imaging (MRI) are a promising marker for the diagnosis and prognosis of neurodegenerative pathologies such as Alzheimer's disease or multiple sclerosis. However, its use in individualized assessments is currently discouraged due to series technical biological issues. In this work, we present deep learning pipeline segmentation-based brain quantification that improves upon automated labels reference method which it learns. This goal...