Juan Guillermo Paniagua Castrillón
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Juan Guillermo Paniagua Castrillón has a total of 23 co-authors. Here are the top co-authors by number of shared publications:
- Andrés Mauricio Muñoz-García
Instituto Tecnológico Metropolitano
2 shared publications - Brayan Alexis Quiceno-Arenas
Instituto Tecnológico Metropolitano
2 shared publications - Luis Fernando Duque Gómez
Centro de Ciencias de la Computación, Facultad de Ciencias Exactas y Aplicadas, Instituto Tecnológico Metropolitano, Calle 73 No 76A-354, Medellín 050034, Colombia
2 shared publications - Moises Oswaldo Bustamante Rúa
Universidad Nacional de Colombia
2 shared publications Olga Lucia Quintero Montoya
Affiliation Computation and Analytics Area, School of Applied Sciences and Analytics, University EAFIT, Medellín, Colombia
2 shared publications- Alejandro Hernández-Arango
Affiliations Department of Internal Medicine, School of Medicine, University of Antioquia, Medellín, Colombia, Hospital Alma Máter de Antioquia, Medellín, Colombia
1 shared publication
Latest Publications
- Open Access
- 30 April 2026
- Published by Public Library of Science (PLoS) in PLOS Digital Health
- Vol. 5 (4) , e0000960
Exploratory association between multimodal AI-derived digital biomarkers and in-hospital mortality in adult patients with pneumonia: A proof-of-concept study
Pneumonia remains a leading cause of in-hospital mortality worldwide. Current prognostic tools such as the IDSA/ATS severity score have meaningful limitations, particularly in capturing dynamic disease progression or integrating heterogeneous biological signals. Artificial intelligence (AI) offers the opportunity to derive complementary prognostic information from routinely collected electronic health record data. This exploratory, proof-of-concept retrospective study enrolled adults (≥18 years) admitted with a primary diagnosis of acute pneumonia at Hospital Alma Máter de Antioquia (Medellín, Colombia) between January 1 and June 30, 2024. After applying pre-defined exclusion criteria, 121 patients (19 non-survivors, 15.7%) comprised the final analytic cohort. Three independent AI modules were applied: (i) a ResNet-18 deep learning model quantified lung consolidation from chest radiographs (CXRs) using Class Activation Mapping; (ii) a Spanish regular-expression natural language processing (NLP) pipeline extracted modified IDSA/ATS severity scores from clinical notes; and (iii) NeuroKit2-based quantitative heart rate variability (HRV) analysis processed electrocardiogram (ECG) signals digitised from PDF archives. Bivariate associations with all-cause in-hospital mortality were examined using logistic regression. Several features exhibited statistically significant associations with mortality under conventional thresholds: AI-quantified total lung compromise ratio (OR 8.32, 95% CI 1.23–56.29), NLP-derived IDSA/ATS severity score (OR 1.78, 95% CI 1.09–2.88), and a coherent ECG/HRV profile characterised by higher heart rate (120.1 vs. 84.4 bpm, p = 0.023), reduced RMSSD (4.1 vs. 23.5 ms, p = 0.041), reduced Poincaré SD1 (3.0 vs. 17.6 ms, p = 0.041), and T-wave amplitude reductions surviving FDR correction in multiple leads. Given the small sample size and low event count (n = 19; events-per-variable ≈ 4), all associations are preliminary and hypothesis-generating only. These proof-of-concept findings suggest that integrated multimodal AI biomarkers automatically derived from low-resource clinical data can capture a cardiopulmonary stress profile associated with pneumonia mortality, and support the design of larger prospective validation studies.
- 30 April 2026
- Published by Public Library of Science (PLoS) in PLOS Digital Health
- Vol. 5 (4) , e0000960
Open AccessExploratory association between multimodal AI-derived digital biomarkers and in-hospital mortality in adult patients with pneumonia: A proof-of-concept study
Pneumonia remains a leading cause of in-hospital mortality worldwide. Current prognostic tools such as the IDSA/ATS severity score have meaningful limitations, particularly in capturing dynamic disease progression or integrating heterogeneous biological signals. Artificial intelligence (AI) offers the opportunity to derive complementary prognostic information from routinely collected electronic health record data. This exploratory, proof-of-concept retrospective study enrolled adults (≥18 years) admitted with a primary diagnosis of acute pneumonia at Hospital Alma Máter de Antioquia (Medellín, Colombia) between January 1 and June 30, 2024. After applying pre-defined exclusion criteria, 121 patients (19 non-survivors, 15.7%) comprised the final analytic cohort. Three independent AI modules were applied: (i) a ResNet-18 deep learning model quantified lung consolidation from chest radiographs (CXRs) using Class Activation Mapping; (ii) a Spanish regular-expression natural language processing (NLP) pipeline extracted modified IDSA/ATS severity scores from clinical notes; and (iii) NeuroKit2-based quantitative heart rate variability (HRV) analysis processed electrocardiogram (ECG) signals digitised from PDF archives. Bivariate associations with all-cause in-hospital mortality were examined using logistic regression. Several features exhibited statistically significant associations with mortality under conventional thresholds: AI-quantified total lung compromise ratio (OR 8.32, 95% CI 1.23–56.29), NLP-derived IDSA/ATS severity score (OR 1.78, 95% CI 1.09–2.88), and a coherent ECG/HRV profile characterised by higher heart rate (120.1 vs. 84.4 bpm, p = 0.023), reduced RMSSD (4.1 vs. 23.5 ms, p = 0.041), reduced Poincaré SD1 (3.0 vs. 17.6 ms, p = 0.041), and T-wave amplitude reductions surviving FDR correction in multiple leads. Given the small sample size and low event count (n = 19; events-per-variable ≈ 4), all associations are preliminary and hypothesis-generating only. These proof-of-concept findings suggest that integrated multimodal AI biomarkers automatically derived from low-resource clinical data can capture a cardiopulmonary stress profile associated with pneumonia mortality, and support the design of larger prospective validation studies.
- 23 April 2026
- Published by Universidad Industrial de Santander in Boletín de Geología
- Vol. 48 (1) , 95
El principio de discrepancia como método de optimización para un esquema de inversión de Gauss-Newton
La inversión geofísica es un proceso matemático utilizado para inferir características del subsuelo a partir de modelos que representan sus propiedades físicas, según datos medidos o parámetros físicos. Uno de los retos de la inversión geofísica es generar modelos que se ajusten a los datos obtenidos. Para resolverlo, se han desarrollado diversas metodologías, como el esquema de Gauss-Newton. Estas metodologías requieren la selección óptima del parámetro de regularización, pero no proporcionan una función o clase predefinida para dicha selección. En consecuencia, este estudio propone un novedoso algoritmo de optimización, basado en el principio de discrepancia, para estimar el parámetro de regularización utilizando un conjunto de datos sintéticos y reales de tomografía de resistividad eléctrica. De acuerdo con los resultados, se obtuvieron valores ideales de tolerancia para cada conjunto de datos, y se demostró la eficacia y las ventajas computacionales del algoritmo propuesto. Además, el algoritmo mostró una mejor definición de las estructuras de interés que otros algoritmos reportados en la literatura científica.
- 23 April 2026
- Published by Universidad Industrial de Santander in Boletín de Geología
- Vol. 48 (1) , 95
El principio de discrepancia como método de optimización para un esquema de inversión de Gauss-Newton
La inversión geofísica es un proceso matemático utilizado para inferir características del subsuelo a partir de modelos que representan sus propiedades físicas, según datos medidos o parámetros físicos. Uno de los retos de la inversión geofísica es generar modelos que se ajusten a los datos obtenidos. Para resolverlo, se han desarrollado diversas metodologías, como el esquema de Gauss-Newton. Estas metodologías requieren la selección óptima del parámetro de regularización, pero no proporcionan una función o clase predefinida para dicha selección. En consecuencia, este estudio propone un novedoso algoritmo de optimización, basado en el principio de discrepancia, para estimar el parámetro de regularización utilizando un conjunto de datos sintéticos y reales de tomografía de resistividad eléctrica. De acuerdo con los resultados, se obtuvieron valores ideales de tolerancia para cada conjunto de datos, y se demostró la eficacia y las ventajas computacionales del algoritmo propuesto. Además, el algoritmo mostró una mejor definición de las estructuras de interés que otros algoritmos reportados en la literatura científica.
- Open Access
- 19 July 2022
- Published by Universidad Nacional de Colombia in DYNA
- Vol. 89 (222) , 48-53
Open-source learning as a skill for geoelectrical data processing: the case of pyGIMLi
It is important that new strategies are developed for the management of geoelectrical data produced from engineering and geoscience processing techniques. For this reason, the article demonstrates how pyGIMLi can be used for this purpose. pyGIMLi is an open-source library for the inversion of apparent resistivity array that are often obtained with different geoelectrical survey equipment. The aim is to be able to use this library unaided for various projects and/or to perform various data operations in which the results obtained are more specific and differentiated than those derived from other processing techniques, taking advantage of the fact that this tool is open-source.
- 19 July 2022
- Published by Universidad Nacional de Colombia in DYNA
- Vol. 89 (222) , 48-53
Open AccessOpen-source learning as a skill for geoelectrical data processing: the case of pyGIMLi
It is important that new strategies are developed for the management of geoelectrical data produced from engineering and geoscience processing techniques. For this reason, the article demonstrates how pyGIMLi can be used for this purpose. pyGIMLi is an open-source library for the inversion of apparent resistivity array that are often obtained with different geoelectrical survey equipment. The aim is to be able to use this library unaided for various projects and/or to perform various data operations in which the results obtained are more specific and differentiated than those derived from other processing techniques, taking advantage of the fact that this tool is open-source.
- Open Access
- 10 May 2019
- Published by Instituto Colombiano del Petroleo in Ct&f - Ciencia, Tecnología Y Futuro
- Vol. 9 (1) , 111-117
Reverse Time Migration in Euclidean and Riemannian coordinates
Reverse time migration in zones with rugged topography is a method that presents some challenging issues. We present an analysis of reverse time migration in transformed domains, in particular for a technique that goes from an Euclidian to a Riemannian scenario, as suggested by some authors in previous literature. Computational results show that there is not significant improvement in the final image when the Riemannian approach is used as compared with images obtained with an Euclidean metric.
- 10 May 2019
- Published by Instituto Colombiano del Petroleo in Ct&f - Ciencia, Tecnología Y Futuro
- Vol. 9 (1) , 111-117
Open AccessReverse Time Migration in Euclidean and Riemannian coordinates
Reverse time migration in zones with rugged topography is a method that presents some challenging issues. We present an analysis of reverse time migration in transformed domains, in particular for a technique that goes from an Euclidian to a Riemannian scenario, as suggested by some authors in previous literature. Computational results show that there is not significant improvement in the final image when the Riemannian approach is used as compared with images obtained with an Euclidean metric.
- 19 December 2018
- Published by Instituto Colombiano del Petroleo in Ct&f - Ciencia, Tecnología Y Futuro
- Vol. 8 (2) , 65-73
Application of the continuous wavelet transform in the extraction of directional data on RTM imaging condition wavefields
Low-frequency artifacts in reverse time migration result from unwanted cross-correlation of the source and receiver wavefields at non-reflecting points along ray-paths. These artifacts can hide important details in migrated models and increase poor interpretation risk. Some methods have been proposed to avoid or reduce the number of these artifacts, preserving reflections, and improving model quality, implementing other strategies such as modification of the wave equation, proposing other imaging conditions, and using image filtering techniques. One of these methods uses wavefield decomposition, correlating components of the wavefields that propagate in opposite directions. We propose a method for extracting directional information from the RTM imaging condition wavefields to obtain characteristics allowing for better, more refined imaging. The method works by separating directional information about the wavefields based on the continuous wavelet transform (CWT), and the analysis of the main changes on the frequency content revealed within the scalogram obtained by a Gaussian wavelet family. Through numerical applications, we demonstrate that this method can effectively remove undesired artifacts in migrated images. In addition, we use the Laguerre-Gauss filtering to improve the results obtained with the proposed method.
- 19 December 2018
- Published by Instituto Colombiano del Petroleo in Ct&f - Ciencia, Tecnología Y Futuro
- Vol. 8 (2) , 65-73
Application of the continuous wavelet transform in the extraction of directional data on RTM imaging condition wavefields
Low-frequency artifacts in reverse time migration result from unwanted cross-correlation of the source and receiver wavefields at non-reflecting points along ray-paths. These artifacts can hide important details in migrated models and increase poor interpretation risk. Some methods have been proposed to avoid or reduce the number of these artifacts, preserving reflections, and improving model quality, implementing other strategies such as modification of the wave equation, proposing other imaging conditions, and using image filtering techniques. One of these methods uses wavefield decomposition, correlating components of the wavefields that propagate in opposite directions. We propose a method for extracting directional information from the RTM imaging condition wavefields to obtain characteristics allowing for better, more refined imaging. The method works by separating directional information about the wavefields based on the continuous wavelet transform (CWT), and the analysis of the main changes on the frequency content revealed within the scalogram obtained by a Gaussian wavelet family. Through numerical applications, we demonstrate that this method can effectively remove undesired artifacts in migrated images. In addition, we use the Laguerre-Gauss filtering to improve the results obtained with the proposed method.
- 21 February 2018
- Published by Sociedade Brasileira de Geofisica in Revista Brasileira de Geofísica
- Vol. 35 (2) , 81-93
LAGUERRE-GAUSS FILTERS IN REVERSE TIME MIGRATION IMAGE RECONSTRUCTION
Reverse time migration (RTM) solves the acoustic or elastic wave equation by means of the extrapolation from source and receiver wavefield in time. A migrated image is obtained by applying a criteria known as imaging condition. The cross-correlation between source and receiver wavefields is the commonly used imaging condition. However, this imaging condition produces spatial low-frequency noise, called artifacts, due to the unwanted correlation of the diving, head and backscattered waves. Several techniques have been proposed to reduce the artifacts occurrence. Derivative operators as Laplacian are the most frequently used. In this work, we propose a technique based on a spiral phase filter ranging from 0 to 2π, and a toroidal amplitude bandpass filter, known as Laguerre-Gauss transform. Through numerical experiments we present the application of this particular filter on three synthetic data sets. In addition, we present a comparative spectral study of images obtained by the zero-lag cross-correlation imaging condition, the Laplacian filtering and the Laguerre-Gauss filtering, showing their frequency features. We also present evidences not only with simulated noisy velocity fields but also by comparison with the model velocity field gradients that this method improves the RTM images by reducing the artifacts and notably enhance the reflective events. Keywords: Laguerre-Gauss transform, zero-lag cross-correlation, seismic migration, imaging condition. RESUMO. A migração reversa no tempo (RTM) resolve a equação de onda acústica ou elástica por meio da extrapolação a partir do campo de onda da fonte e do receptor no tempo. Uma imagem migrada é obtida aplicando um critério conhecido como condição de imagem. A correlação cruzada entre campos de onda de fonte e receptor é a condição de imagem comumente usada. No entanto, esta condição de imagem produz ruído espacial de baixa frequência, chamados artefatos, devido à correlação indesejada das ondas de mergulho, cabeça e retrodifusão. Várias técnicas têm sido propostas para reduzir a ocorrência de artefatos. Operadores derivados como Laplaciano são os mais utilizados. Neste trabalho, propomos uma técnica baseada em um filtro de fase espiral que varia de 0 a 2π, e um filtro passabanda de amplitude toroidal, conhecido como transformada de Laguerre-Gauss. Através de experimentos numéricos, apresentamos a aplicação deste filtro particular em três conjuntos de dados sintéticos. Além disso, apresentamos um estudo comparativo espectral de imagens obtidas pela condição de imagem de correlação cruzada atraso zero, a filtragem de Laplaciano e a filtragem Laguerre-Gauss, mostrando suas características de frequência. Apresentamos evidências não somente com campos simulados de velocidade ruidosa, mas também por comparação com os gradientes de campo de velocidade do modelo que este método melhora as imagens RTM, reduzindo os artefatos e aumentando notavelmente os eventos reflexivos. Palavras-chave: Transformação de Laguerre-Gauss, correlação cruzada atraso zero, migração sísmica, condição de imagem.
- 21 February 2018
- Published by Sociedade Brasileira de Geofisica in Revista Brasileira de Geofísica
- Vol. 35 (2) , 81-93
LAGUERRE-GAUSS FILTERS IN REVERSE TIME MIGRATION IMAGE RECONSTRUCTION
Reverse time migration (RTM) solves the acoustic or elastic wave equation by means of the extrapolation from source and receiver wavefield in time. A migrated image is obtained by applying a criteria known as imaging condition. The cross-correlation between source and receiver wavefields is the commonly used imaging condition. However, this imaging condition produces spatial low-frequency noise, called artifacts, due to the unwanted correlation of the diving, head and backscattered waves. Several techniques have been proposed to reduce the artifacts occurrence. Derivative operators as Laplacian are the most frequently used. In this work, we propose a technique based on a spiral phase filter ranging from 0 to 2π, and a toroidal amplitude bandpass filter, known as Laguerre-Gauss transform. Through numerical experiments we present the application of this particular filter on three synthetic data sets. In addition, we present a comparative spectral study of images obtained by the zero-lag cross-correlation imaging condition, the Laplacian filtering and the Laguerre-Gauss filtering, showing their frequency features. We also present evidences not only with simulated noisy velocity fields but also by comparison with the model velocity field gradients that this method improves the RTM images by reducing the artifacts and notably enhance the reflective events. Keywords: Laguerre-Gauss transform, zero-lag cross-correlation, seismic migration, imaging condition. RESUMO. A migração reversa no tempo (RTM) resolve a equação de onda acústica ou elástica por meio da extrapolação a partir do campo de onda da fonte e do receptor no tempo. Uma imagem migrada é obtida aplicando um critério conhecido como condição de imagem. A correlação cruzada entre campos de onda de fonte e receptor é a condição de imagem comumente usada. No entanto, esta condição de imagem produz ruído espacial de baixa frequência, chamados artefatos, devido à correlação indesejada das ondas de mergulho, cabeça e retrodifusão. Várias técnicas têm sido propostas para reduzir a ocorrência de artefatos. Operadores derivados como Laplaciano são os mais utilizados. Neste trabalho, propomos uma técnica baseada em um filtro de fase espiral que varia de 0 a 2π, e um filtro passabanda de amplitude toroidal, conhecido como transformada de Laguerre-Gauss. Através de experimentos numéricos, apresentamos a aplicação deste filtro particular em três conjuntos de dados sintéticos. Além disso, apresentamos um estudo comparativo espectral de imagens obtidas pela condição de imagem de correlação cruzada atraso zero, a filtragem de Laplaciano e a filtragem Laguerre-Gauss, mostrando suas características de frequência. Apresentamos evidências não somente com campos simulados de velocidade ruidosa, mas também por comparação com os gradientes de campo de velocidade do modelo que este método melhora as imagens RTM, reduzindo os artefatos e aumentando notavelmente os eventos reflexivos. Palavras-chave: Transformação de Laguerre-Gauss, correlação cruzada atraso zero, migração sísmica, condição de imagem.
- 3 August 2017
- Published by Society of Exploration Geophysicists
- Vol. 1, 1284-1289
The use of Laguerre-Gauss transform in 2D reverse time migration imaging
Zero-Lag Cross-Correlation Imaging condition (ZL-CC-IC) is widely used in Reverse Time Migration (RTM) for the recovery of structural images of the subsurface. The presence of spatial low frequency noise so called artifacts, is one of the challenges of post processing in certain complex velocity field models with special features. Paniagua and Sierra-Sosa (2016) proposed a new post processing technique for the edge enhancement and reduction of low frequency artifacts; now our aim is to demonstrate the good performance of this technique and establish quantitative features and relations between the numerical results and the physical phenomenon. We present a comparative spectral study of the zero-lag cross-correlation imaging condition (ZL-CC-IC), zero-lag cross-correlation plus Laplacian filtering (ZL-CC-LP) and zero-lag cross-correlation plus Laguerre-Gauss filtering (ZL-CC-LG) showing their frequency features, also we show some results of the Laguerre-Gauss filtering applied in complex synthetic datasets and finally, using a smoothed velocity model in order to simulate the uncertainties of a real application, we show the good results obtained by using the Laguerre-Gauss Filtering posterior to RTM.
- 3 August 2017
- Published by Society of Exploration Geophysicists
- Vol. 1, 1284-1289
The use of Laguerre-Gauss transform in 2D reverse time migration imaging
Zero-Lag Cross-Correlation Imaging condition (ZL-CC-IC) is widely used in Reverse Time Migration (RTM) for the recovery of structural images of the subsurface. The presence of spatial low frequency noise so called artifacts, is one of the challenges of post processing in certain complex velocity field models with special features. Paniagua and Sierra-Sosa (2016) proposed a new post processing technique for the edge enhancement and reduction of low frequency artifacts; now our aim is to demonstrate the good performance of this technique and establish quantitative features and relations between the numerical results and the physical phenomenon. We present a comparative spectral study of the zero-lag cross-correlation imaging condition (ZL-CC-IC), zero-lag cross-correlation plus Laplacian filtering (ZL-CC-LP) and zero-lag cross-correlation plus Laguerre-Gauss filtering (ZL-CC-LG) showing their frequency features, also we show some results of the Laguerre-Gauss filtering applied in complex synthetic datasets and finally, using a smoothed velocity model in order to simulate the uncertainties of a real application, we show the good results obtained by using the Laguerre-Gauss Filtering posterior to RTM.
- 12 June 2017
- Published by EAGE Publications bv
Attenuation of Reverse Time Migration Artifacts Using Laguerre-Gauss Filtering
- 12 June 2017
- Published by EAGE Publications bv
Attenuation of Reverse Time Migration Artifacts Using Laguerre-Gauss Filtering