A5072005501- AI in cancer detection
- Radiomics and Machine Learning in Medical Imaging
- Digital Imaging for Blood Diseases
- Image Retrieval and Classification Techniques
- Cell Image Analysis Techniques
- Advanced Image and Video Retrieval Techniques
- Colorectal Cancer Screening and Detection
- Medical Imaging and Analysis
- Cardiovascular Function and Risk Factors
- Medical Image Segmentation Techniques
- Multimodal Machine Learning Applications
- Cutaneous Melanoma Detection and Management
- Video Analysis and Summarization
- Image Processing Techniques and Applications
- ECG Monitoring and Analysis
- Cardiac Imaging and Diagnostics
- Brain Tumor Detection and Classification
- Retinal Imaging and Analysis
- Data Visualization and Analytics
- Genomics and Phylogenetic Studies
- Water Resource Management and Quality
- Knowledge Societies in the 21st Century
- IoT and GPS-based Vehicle Safety Systems
- Gene expression and cancer classification
- Bat Biology and Ecology Studies
University of the Llanos
2015-2024
Universidad Nacional de Colombia
2009-2023
This paper presents a deep learning approach for automatic detection and visual analysis of invasive ductal carcinoma (IDC) tissue regions in whole slide images (WSI) breast cancer (BCa). Deep approaches are learn-from-data methods involving computational modeling the process. is similar to how human brain works using different interpretation levels or layers most representative useful features resulting into hierarchical learned representation. These have been shown outpace traditional...
With the increasing ability to routinely and rapidly digitize whole slide images with scanners, there has been interest in developing computerized image analysis algorithms for automated detection of disease extent from digital pathology images. The manual identification presence breast cancer by a pathologist is critical patient management tumor staging assessing treatment response. However, this process tedious subject inter- intra-reader variability. For methods be useful as decision...
Breast cancer (BCa) grading plays an important role in predicting disease aggressiveness and patient outcome. A key component of BCa grade is the mitotic count, which involves quantifying number cells process dividing (i.e., undergoing mitosis) at a specific point time. Currently, mitosis counting done manually by pathologist looking multiple high power fields (HPFs) on glass slide under microscope, extremely laborious time consuming process. The development computerized systems for...
Precise detection of invasive cancer on whole-slide images (WSI) is a critical first step in digital pathology tasks diagnosis and grading. Convolutional neural network (CNN) the most popular representation learning method for computer vision tasks, which have been successfully applied pathology, including tumor mitosis detection. However, CNNs are typically only tenable with relatively small image sizes (200 × 200 pixels). Only recently, Fully convolutional networks (FCN) able to deal...
Breast cancer (BCa) grading plays an important role in predicting disease aggressiveness and patient outcome. A key component of BCa grade is mitotic count, which involves quantifying the number cells process dividing (i.e. undergoing mitosis) at a specific point time. Currently mitosis counting done manually by pathologist looking multiple high power fields on glass slide under microscope, extremely laborious time consuming process. The development computerized systems for automated...
Histopathological images are an important resource for clinical diagnosis and biomedical research. From image understanding point of view, the automatic annotation these is a challenging problem. This paper presents new method histopathological based on three complementary strategies, first, part-based representation, called bag features, which takes advantage natural redundancy capturing fundamental patterns biological structures, second, latent topic model, non-negative matrix...
<p>This paper presents a review of the state-of-the-art in histopathology image representation used automatic analysis tasks. Automatic images is important for building computer-assisted diagnosis tools, enhancing systems and virtual microscopy systems, among other applications. Histopathology have rich mix visual patterns with particularities that make them difficult to analyze. The discusses these particularities, acquisition process challenges found when doing analysis. Second an...
Convolutional neural networks (CNN) have been very successful at addressing different computer vision tasks thanks to their ability learn image representations directly from large amounts of labeled data. Features learned a dataset can be used represent images via an approach called transfer learning. In this paper we apply learning the challenging task medulloblastoma tumor differentiation. We compare two CNN models which were previously trained in domains (natural and histopathology...
This paper presents a novel method for basal-cell carcinoma detection, which combines state-of-the-art methods unsupervised feature learning (UFL) and bag of features (BOF) representation. BOF, is form representation learning, has shown good performance in automatic histopathology image classi cation. In patches are usually represented using descriptors such as SIFT DCT. We propose to use UFL learn the patch itself. accomplished by applying topographic (T-RICA), automatically learns visual...
Learning data representations directly from the itself is an approach that has shown great success in different pattern recognition problems, outperforming state-of-the-art feature extraction schemes for tasks computer vision, speech and natural language processing. Representation learning applies unsupervised supervised machine methods to large amounts of find building-blocks better represent information it. Digitized histopathology images represents a very good testbed representation since...
This paper presents BIGS the Big Image Data Analysis Toolkit, a software framework for large scale image processing and analysis over heterogeneous computing resources, such as those available in clouds, grids, computer clusters or throughout scattered resources (desktops, labs) an opportunistic manner. Through BIGS, eScience is conceived to exploit coarse grained parallelism based on data partitioning parameter sweeps, avoiding need of inter-process communication and, therefore, enabling...
This paper describes an extension of the Bag-of-Visual-Words (BoVW) representation for image categorization (IC) histophatology images. is one most used approaches in several high-level computer vision tasks. However, BoVW has important limitation: disregarding spatial information among visual words. may be useful to capture discriminative visual-patterns specific In order overcome this problem we propose use n-grams. N-grams based-representations are very popular field natural language...
This work addresses the problem of lung sound classification, in particular, distinguishing between wheeze and normal sounds. Wheezing detection is an important step to associate sounds with abnormal state respiratory system, usually associated tuberculosis or another chronic obstructive pulmonary diseases (COPD). The paper presents approach for automatic which uses different state-of-the-art features combination a C-weighted support vector machine (SVM) classifier that works better...
La clasificación de cobertura del suelo es importante para estudios cambio climático y monitoreo servicios ecosistémicos. Los métodos convencionales coberturas se realizan mediante la interpretación visual imágenes satelitales, lo cual costoso, dispendioso e impreciso. Implementar computacionales permite generar en satelitales manera automática, rápida, precisa económica. Particularmente, los aprendizaje automático son técnicas promisorias estimación cambios suelo. En este trabajo presenta...
Left ventricular ejection fraction (LVEF) is the most useful cardiac index to assess function of patients from echocardiograms (ECHO). Cardiologists manually delineate left (LV) contour at end diastolic (EDV) and systolic (ESV) times calculate LVEF. This paper presents a novel end-to-end deep-learning model directly estimate LVEF only pairwise sample (EDV ESV) ECHO images, which consists two stages. Firstly, simplified U-Net constructed by reducing number convolutional filters per layer...