A5027450937- AI in cancer detection
- Domain Adaptation and Few-Shot Learning
- Anomaly Detection Techniques and Applications
- Topic Modeling
- Hemodynamic Monitoring and Therapy
- Biomedical Text Mining and Ontologies
- Artificial Intelligence in Healthcare and Education
- Heart Rate Variability and Autonomic Control
- Emotion and Mood Recognition
- Radiomics and Machine Learning in Medical Imaging
- Non-Invasive Vital Sign Monitoring
- Image and Signal Denoising Methods
- Control Systems and Identification
- Museums and Cultural Heritage
- Machine Learning and ELM
- Neural Networks and Applications
- Face recognition and analysis
- Machine Learning and Data Classification
- Cultural Heritage Management and Preservation
- Natural Language Processing Techniques
- Cardiac Health and Mental Health
- Building materials and conservation
- Conservation Techniques and Studies
- Speech and Audio Processing
- Artificial Intelligence in Healthcare
University of Valparaíso
2017-2025
Millennium Institute for Integrative Biology
2023-2025
Valparaiso University
2016-2024
Brain tumors are a leading cause of mortality worldwide, with early and accurate diagnosis being essential for effective treatment. Although Deep Learning (DL) models offer strong performance in tumor detection segmentation using MRI, their black-box nature hinders clinical adoption due to lack interpretability. We present hybrid AI framework that integrates 3D U-Net Convolutional Neural Network MRI-based radiomic feature extraction. Dimensionality reduction is performed machine learning, an...
Multimodal emotion recognition implies the use of different resources and techniques for identifying recognizing human emotions. A variety data sources such as faces, speeches, voices, texts others have to be processed simultaneously this task. However, most techniques, which are based mainly on Deep Learning, trained using datasets designed built in controlled conditions, making their applicability real contexts with conditions more difficult. For reason, aim work is assess a set...
Deep learning models are part of the family artificial neural networks and, as such, they suffer catastrophic interference when sequentially. In addition, greater number these have a rigid architecture which prevents incremental new classes. To overcome drawbacks, we propose Self-Improving Generative Artificial Neural Network (SIGANN), an end-to-end deep network system can ease forgetting problem this method, introduce novel detection model that automatically detects samples classes, and...
We propose an efficient computational method to obtain the fractional derivative of a digital signal. The proposal consists new interpretation Grünwald–Letnikov differintegral operator where we have introduced finite Cauchy convolution with dynamic kernel. can be applied any signal without knowing its analytical form. In experiments, compared proposed Riemman–Louville approach for two well-known functions. simulations exhibit similar results both methods; however, outperforms other in...
Cardiovascular diseases represent the leading cause of death worldwide. Thus, cardiovascular rehabilitation programs are crucial to mitigate deaths caused by this condition each year, mainly in patients with coronary artery disease. COVID-19 was not only a challenge area but also an opportunity open remote or hybrid versions these programs, potentially reducing number who leave due geographical/time barriers. This paper presents method for building prediction model using retrospective and...
Detecting and extracting findings in a radiological report is crucial for text mining tasks several applications. In this case, labeled process the image associated with mammography Spanish context computer vision model required. This paper shows methodology generated goal. presents Named Entity Recognition (NER) approach based on transformer deep learning model, using corpus fine-tuning to find three concepts that compose typical finding mammographic report: laterality, location, finding....
Deep learning models are part of the family artificial neural networks and, as such, it suffers catastrophic interference when they learn sequentially. In addition, most these have a rigid architecture which prevents incremental new classes. To overcome drawbacks, in this article we propose Self-Improving Generative Artificial Neural Network (SIGANN), type end-to-end system is able to ease forgetting problem leaning method, introduce novelty detection model automatically detect samples...
<title>Abstract</title> Background: The application of text mining in radiological reports is crucial for many tasks but principally analyzing and projecting trends to enhance diagnostic accuracy. This especially important mammography reports, where early detection crucial. Large language models (LLMs) provide a viable alternative. They can generate accurate results from limited set examples compared with the traditional state-of-the-art models. Methods: work presents framework utilizing...
Mammography is known as one of the best forms to screen possible breast cancer in women, and recently deep learning models have been developed assist radiologist diagnosis. However, their lack interpretability has become a significant drawback extended use clinical practice. This paper introduces novel approach for detecting localising pathological findings mammography exams through EfficientNet-based model. The model trained using cropped segments labelled from Vindr Dataset. Achieving an...
Multimodal emotion recognition involves identifying human emotions in specific situations using artificial intelligence across multiple modalities. MERDWild, a multimodal dataset, addresses the challenge of unifying, cleaning, and transforming three datasets collected uncontrolled environments with aim integrating standardizing database that encompasses modalities: facial images, audio, text. A methodology is presented combines information from these modalities, utilizing �in-the-wild�...