arXiv subjects

Sergio González

Publications and source records attributed to Sergio González.

6 recordsLinked to original sources

arXiv2024

Multi-modal Heart Failure Risk Estimation based on Short ECG and Sampled Long-Term HRV

Cardiovascular diseases, including Heart Failure (HF), remain a leading global cause of mortality, often evading early detection. In this context, accessible and effective risk assessment is indispensable. Traditional approaches rely on resource-intensive diagnostic tests, typically administered after the onset of symptoms. The widespread availability of electrocardiogram (ECG) technology and the power of Machine Learning are emerging as viable alternatives within smart healthcare. In this paper, we propose several multi-modal approaches that combine 30-second ECG recordings and approximate long-term Heart Rate Variability (HRV) data to estimate the risk of HF hospitalization. We introduce two survival models: an XGBoost model with Accelerated Failure Time (AFT) incorporating comprehensive ECG features and a ResNet model that learns from the raw ECG. We extend these with our novel long-term HRVs extracted from the combination of ultra-short-term beat-to-beat measurements taken over the day. To capture their temporal dynamics, we propose a survival model comprising ResNet and Transformer architectures (TFM-ResNet). Our experiments demonstrate high model performance for HF risk assessment with a concordance index of 0.8537 compared to 14 survival models and competitive discrimination power on various external ECG datasets. After transferability tests with Apple Watch data, our approach implemented in the myHeartScore App offers cost-effective and highly accessible HF risk assessment, contributing to its prevention and management.

eess.SP↗

arXiv2023

Streamlined Framework for Agile Forecasting Model Development towards Efficient Inventory Management

This paper proposes a framework for developing forecasting models by streamlining the connections between core components of the developmental process. The proposed framework enables swift and robust integration of new datasets, experimentation on different algorithms, and selection of the best models. We start with the datasets of different issues and apply pre-processing steps to clean and engineer meaningful representations of time-series data. To identify robust training configurations, we introduce a novel mechanism of multiple cross-validation strategies. We apply different evaluation metrics to find the best-suited models for varying applications. One of the referent applications is our participation in the intelligent forecasting competition held by the United States Agency of International Development (USAID). Finally, we leverage the flexibility of the framework by applying different evaluation metrics to assess the performance of the models in inventory management settings.

cs.LG↗

arXiv2022

Interpretable estimation of the risk of heart failure hospitalization from a 30-second electrocardiogram

Survival modeling in healthcare relies on explainable statistical models; yet, their underlying assumptions are often simplistic and, thus, unrealistic. Machine learning models can estimate more complex relationships and lead to more accurate predictions, but are non-interpretable. This study shows it is possible to estimate hospitalization for congestive heart failure by a 30 seconds single-lead electrocardiogram signal. Using a machine learning approach not only results in greater predictive power but also provides clinically meaningful interpretations. We train an eXtreme Gradient Boosting accelerated failure time model and exploit SHapley Additive exPlanations values to explain the effect of each feature on predictions. Our model achieved a concordance index of 0.828 and an area under the curve of 0.853 at one year and 0.858 at two years on a held-out test set of 6,573 patients. These results show that a rapid test based on an electrocardiogram could be crucial in targeting and treating high-risk individuals.

cs.LG↗

arXiv2020

Fuzzy k-Nearest Neighbors with monotonicity constraints: Moving towards the robustness of monotonic noise

This paper proposes a new model based on Fuzzy k-Nearest Neighbors for classification with monotonic constraints, Monotonic Fuzzy k-NN (MonFkNN). Real-life data-sets often do not comply with monotonic constraints due to class noise. MonFkNN incorporates a new calculation of fuzzy memberships, which increases robustness against monotonic noise without the need for relabeling. Our proposal has been designed to be adaptable to the different needs of the problem being tackled. In several experimental studies, we show significant improvements in accuracy while matching the best degree of monotonicity obtained by comparable methods. We also show that MonFkNN empirically achieves improved performance compared with Monotonic k-NN in the presence of large amounts of class noise.

cs.LG↗

arXiv2014

Persistent C II Absorption in the Normal Type Ia Supernova 2002fk

We present well-sampled $UBVRIJHK$ photometry of SN 2002fk starting 12 days before maximum light through 122 days after peak brightness, along with a series of 15 optical spectra from -4 to +95 days since maximum. Our observations show the presence of C II lines in the early-time spectra of SN 2002fk, expanding at ~11,000 km s$^{-1}~$ and persisting until ~8 days past maximum light with a velocity of $\sim$9,000 km s$^{-1}~$. SN 2002fk is characterized by a small velocity gradient of $\dot v_{Si~II}=26$ km s$^{-1}$ day$^{-1}$, possibly caused by an off-center explosion with the ignition region oriented towards the observer. The connection between viewing angle of an off-center explosion and the presence of C II in the early time spectrum suggests that the observation of C II could be also due to a viewing angle effect. Adopting the Cepheid distance to NGC 1309 we provide the first $H_{0}$ value based on near-IR measurements of a Type Ia supernova between 63.0$\pm$ 0.8 ($\pm$ 2.8 systematic) and 66.7$\pm$1.0 ($\pm$ 3.5 systematic) km/s/Mpc, depending on the absolute magnitude/decline rate relationship adopted. It appears that the near-IR yields somewhat lower (6-9 %) $H_0$ values than the optical. It is essential to further examine this issue by (1) expanding the sample of high-quality near-IR light curves of SNe in the Hubble flow, and (2) increasing the number of nearby SNe with near-IR SN light curves and precise Cepheid distances, which affords the promise to deliver a more precise determination of $H_0$.

astro-ph.HE↗

arXiv2010

SN 2009bb: a Peculiar Broad-Lined Type Ic Supernova

Ultraviolet, optical, and near-infrared photometry and optical spectroscopy of the broad-lined Type Ic supernova (SN) 2009bb are presented, following the flux evolution from -10 to +285 days past B-band maximum. Thanks to the very early discovery, it is possible to place tight constraints on the SN explosion epoch. The expansion velocities measured from near maximum spectra are found to be only slightly smaller than those measured from spectra of the prototype broad-lined SN 1998bw associated with GRB 980425. Fitting an analytical model to the pseudo-bolometric light curve of SN 2009bb suggests that 4.1+-1.9 Msun of material was ejected with 0.22 +-0.06 Msun of it being 56Ni. The resulting kinetic energy is 1.8+-0.7x10^52 erg. This, together with an absolute peak magnitude of MB=-18.36+-0.44, places SN 2009bb on the energetic and luminous end of the broad-lined Type Ic (SN Ic) sequence. Detection of helium in the early time optical spectra accompanied with strong radio emission, and high metallicity of its environment makes SN 2009bb a peculiar object. Similar to the case for GRBs, we find that the bulk explosion parameters of SN 2009bb cannot account for the copious energy coupled to relativistic ejecta, and conclude that another energy reservoir (a central engine) is required to power the radio emission. Nevertheless, the analysis of the SN 2009bb nebular spectrum suggests that the failed GRB detection is not imputable to a large angle between the line-of-sight and the GRB beamed radiation. Therefore, if a GRB was produced during the SN 2009bb explosion, it was below the threshold of the current generation of gamma-ray instruments.

astro-ph.SR↗