The purpose of the study was to build a predictive model for estimating the risk of ICU admission or mortality among patients hospitalized with COVID-19 and provide a user-friendly tool to assist clinicians in the decision-making process. The study cohort comprised 3623 patients with confirmed COVID-19 who were hospitalized in the SALUD hospital network of Aragon (Spain), which includes 23 hospitals, between February 2020 and January 2021, a period that includes several pandemic waves. Up to 165 variables were analysed, including demographics, comorbidity, chronic drugs, vital signs, and laboratory data. To build the predictive models, different techniques and machine learning (ML) algorithms were explored: multilayer perceptron, random forest, and extreme gradient boosting (XGBoost). A reduction dimensionality procedure was used to minimize the features to 20, ensuring feasible use of the tool in practice. Our model was validated both internally and externally. We also assessed its calibration and provide an analysis of the optimal cut-off points depending on the metric to be optimized. The best performing algorithm was XGBoost. The final model achieved good discrimination for the external validation set (AUC = 0.821, 95% CI 0.787–0.854) and accurate calibration (slope = 1, intercept = −0.12). A cut-off of 0.4 provides a sensitivity and specificity of 0.71 and 0.78, respectively. In conclusion, we built a risk prediction model from a large amount of data from several pandemic waves, which had good calibration and discrimination ability. We also created a user-friendly web application that can aid rapid decision-making in clinical practice.

The increase in the proportion of elderly in Europe brings with it certain challenges that society needs to address, such as custodial care. We propose a scalable, easily modulated and live assistive technology system, based on a comfortable smart footwear capable of detecting walking behaviour, in order to prevent possible health problems in the elderly, facilitating their urban life as independently and safety as possible. This brings with it the challenge of handling the large amounts of data generated, transmitting and pre-processing that information and analysing it with the aim of obtaining useful information in real/near-real time. This is the basis of information theory. This work presents a complete system aiming at elderly people that can detect different user behaviours/events (sitting, standing without imbalance, standing with imbalance, walking, running, tripping) through information acquired from 20 types of sensor measurements (16 piezoelectric pressure sensors, one accelerometer returning reading for the 3 axis and one temperature sensor) and warn the relatives about possible risks in near-real time. For the detection of these events, a hierarchical structure of cascading binary models is designed and applied using artificial neural network (ANN) algorithms and deep learning techniques. The best models are achieved with convolutional layered ANN and multilayer perceptrons. The overall event detection performance achieves an average accuracy and area under the ROC curve of 0.84 and 0.96, respectively.

Small-for-gestational-age (SGA) infants have been associated with increased risk of adverse perinatal outcomes (APOs). In this work, we assess the predictive ability of the ultrasound-estimated percentile weight (EPW) at 35 weeks of gestational age to predict late-onset SGA and APOs, according to six growth standards, and whether the ultrasound–delivery interval influences the detection rate. To this purpose, we analyze a retrospective cohort study of 9585 singleton pregnancies. EPWs at 35 weeks were calculated to the customized Miguel Servet University Hospital (MSUH) and Figueras standards and the non-customized MSUH, Fetal Medicine Foundation (FMF), INTERGROWTH-21st, and WHO standards. As results of our analysis, for a 10% false positive rate, the detection rates for SGA ranged between 48.9% with the customized Figueras standard (AUC 0.82) and 60.8% with the non-customized FMF standard (AUC 0.87). Detection rates to predict SGA by ultrasound–delivery interval (1–6 weeks) show higher detection rates as intervals decrease. APOs detection rates ranged from 27.0% with FMF to 7.9% with the Figueras standard. In conclusion, the ability of EPW to predict SGA at 35 weeks is good for all standards, and slightly better for non-customized standards. The APO detection rate is significantly greater for non-customized standards.

Background/Objectives: Sarcopenia and cognitive decline (CD) are prevalent in aging populations, impacting functionality and quality of life. The early detection of these diseases is challenging, often relying on in-person screening, which is difficult to implement regularly. This study aims to develop artificial intelligence algorithms based on gait analysis, integrating sensor and computer vision (CV) data, to detect sarcopenia and CD. Methods: A cross-sectional case-control study was conducted involving 42 individuals aged 60 years or older. Participants were classified as having sarcopenia if they met the criteria established by the European Working Group on Sarcopenia in Older People and as having CD if their score in the Mini-Mental State Examination was ≤24 points. Gait patterns were assessed at usual walking speeds using sensors attached to the feet and lumbar region, and CV data were captured using a camera. Several key variables related to gait dynamics were extracted. Finally, machine learning models were developed using these variables to predict sarcopenia and CD. Results: Models based on sensor data, CV data, and a combination of both technologies achieved high predictive accuracy, particularly for CD. The best model for CD achieved an F1-score of 0.914, with a 95% sensitivity and 92% specificity. The combined technologies model for sarcopenia also demonstrated high performance, yielding an F1-score of 0.748 with a 100% sensitivity and 83% specificity. Conclusions: The study demonstrates that gait analysis through sensor and CV fusion can effectively screen for sarcopenia and CD. The multimodal approach enhances model accuracy, potentially supporting early disease detection and intervention in home settings.

Background/Objective: Gastric cancer (GC) is a complex disease representing a significant global health concern. Advanced tools for the early diagnosis and prediction of adverse outcomes are crucial. In this context, artificial intelligence (AI) plays a fundamental role. The aim of this work was to develop a diagnostic and prognostic tool for GC, providing support to clinicians in critical decision-making and enabling personalised strategies. Methods: Different machine learning and deep learning techniques were explored to build diagnostic and prognostic models, ensuring model interpretability and transparency through explainable AI methods. These models were developed and cross-validated using data from 590 Spanish Caucasian patients with primary GC and 633 cancer-free individuals. Up to 261 variables were analysed, including demographic, environmental, clinical, tumoral, and genetic data. Variables such as Helicobacter pylori infection, tobacco use, family history of GC, TNM staging, metastasis, tumour location, treatment received, gender, age, and genetic factors (single nucleotide polymorphisms) were selected as inputs due to their association with the risk and progression of the disease. Results: The XGBoost algorithm (version 1.7.4) achieved the best performance for diagnosis, with an AUC value of 0.68 using 5-fold cross-validation. As for prognosis, the Random Survival Forest algorithm achieved a C-index of 0.77. Of interest, the incorporation of genetic data into the clinical–demographics models significantly increased discriminatory ability in both diagnostic and prognostic models. Conclusions: This article presents GastricAITool, a simple and intuitive decision support tool for the diagnosis and prognosis of GC.

Comparing pandemic waves could aid in understanding the evolution of COVID-19. The objective of the present study was to compare the characteristics and outcomes of patients hospitalized for COVID-19 in different pandemic waves in terms of severity and mortality. We performed an observational retrospective cohort study of 5,220 patients hospitalized with SARS-CoV-2 infection from February to September 2020 in Aragon, Spain. We compared ICU admissions and 30-day mortality, clinical characteristics, and risk factors of the first and second waves of COVID-19. The SARS-CoV-2 genome was also analyzed in 236 samples. Patients in the first wave (n = 2,547) were older (median age 74 years [IQR 60–86] vs. 70 years [53–85]; p < 0.001) and had worse clinical and analytical parameters related to severe COVID-19 than patients in the second wave (n = 2,673). The probability of ICU admission at 30 days was 16% and 10% (p < 0.001) and the cumulative 30-day mortality rates 38% and 32% in the first and second wave, respectively (p = 0.007). Survival differences were observed among patients aged 60 to 80 years. We also found some variability among death risk factors and the viral genome between waves. Therefore, the two analyzed COVID-19 pandemic waves were different in terms of disease severity and mortality.