Pub Date : 2026-01-01DOI: 10.1016/j.rceng.2025.502412
L. Ramos-Ruperto , J. Mora-Delgado , A. Rodríguez-González , M.Á. Sicilia , M.J. Pardilla , J.M Sempere , R. Puchades
Machine learning (ML) is a branch of artificial intelligence that is transforming clinical practice by providing tools capable of analyzing large volumes of data, identifying complex patterns, and generating predictions useful for medical decision-making. This article offers a practical and accessible introduction to key ML concepts for internists, addressing its application in tasks such as diagnosis, prognosis, and clinical management. The main types of learning (supervised, unsupervised, and reinforcement learning), the importance of data quality, and the systematic process for developing ML projects in medicine are described. Advanced approaches, such as neural networks and model explainability, are also explored. By integrating these tools, clinicians can improve diagnostic accuracy, personalize treatments, and optimize resources, always with a critical approach that respects medical ethics.
{"title":"Machine learning and deep learning in internal medicine: demystifying concepts","authors":"L. Ramos-Ruperto , J. Mora-Delgado , A. Rodríguez-González , M.Á. Sicilia , M.J. Pardilla , J.M Sempere , R. Puchades","doi":"10.1016/j.rceng.2025.502412","DOIUrl":"10.1016/j.rceng.2025.502412","url":null,"abstract":"<div><div>Machine learning (ML) is a branch of artificial intelligence that is transforming clinical practice by providing tools capable of analyzing large volumes of data, identifying complex patterns, and generating predictions useful for medical decision-making. This article offers a practical and accessible introduction to key ML concepts for internists, addressing its application in tasks such as diagnosis, prognosis, and clinical management. The main types of learning (supervised, unsupervised, and reinforcement learning), the importance of data quality, and the systematic process for developing ML projects in medicine are described. Advanced approaches, such as neural networks and model explainability, are also explored. By integrating these tools, clinicians can improve diagnostic accuracy, personalize treatments, and optimize resources, always with a critical approach that respects medical ethics.</div></div>","PeriodicalId":94354,"journal":{"name":"Revista clinica espanola","volume":"226 1","pages":"Article 502412"},"PeriodicalIF":0.0,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145812659","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-01DOI: 10.1016/j.rceng.2025.502419
G. Martínez de las Cuevas , C. Baldeón Conde , S. Merino Millán , J.M. Olmos Martínez , J.L. Hernández Hernández , D. Nan
Background
We evaluated the impact of a specialized unit on reducing heart failure (HF) readmissions in elderly patients with multiple comorbidities and HFrEF (LVEF < 40%) or mildly reduced EF (LVEF 40–50%), considering different levels of pharmacological optimization.
Methods
Retrospective analysis of a cohort of 135 patients. Readmission rates and their association with optimized treatment were analyzed.
Results
HF admissions decreased by 51% compared to the previous year (p = 0.013). Sixty percent received quadruple therapy, and 62–71% at least three drugs. NT-proBNP levels dropped by 70% (p < 0.001). Quadruple therapy was associated with fewer readmissions at 12 months (p = 0.036), as were ARNI + BB + MRA (p = 0.016) and MRA monotherapy (p = 0.012). The median time to achieve therapeutic optimization was 52 days (27–82 days).
Conclusions
A specialized unit markedly improves therapeutic optimization and reduces readmissions in these patients.
{"title":"Impact of therapeutic optimization in elderly multimorbid patients with heart failure and reduced ejection fraction","authors":"G. Martínez de las Cuevas , C. Baldeón Conde , S. Merino Millán , J.M. Olmos Martínez , J.L. Hernández Hernández , D. Nan","doi":"10.1016/j.rceng.2025.502419","DOIUrl":"10.1016/j.rceng.2025.502419","url":null,"abstract":"<div><h3>Background</h3><div>We evaluated the impact of a specialized unit on reducing heart failure (HF) readmissions in elderly patients with multiple comorbidities and HFrEF (LVEF < 40%) or mildly reduced EF (LVEF 40–50%), considering different levels of pharmacological optimization.</div></div><div><h3>Methods</h3><div>Retrospective analysis of a cohort of 135 patients. Readmission rates and their association with optimized treatment were analyzed.</div></div><div><h3>Results</h3><div>HF admissions decreased by 51% compared to the previous year (p = 0.013). Sixty percent received quadruple therapy, and 62–71% at least three drugs. NT-proBNP levels dropped by 70% (p < 0.001). Quadruple therapy was associated with fewer readmissions at 12 months (p = 0.036), as were ARNI + BB + MRA (p = 0.016) and MRA monotherapy (p = 0.012). The median time to achieve therapeutic optimization was 52 days (27–82 days).</div></div><div><h3>Conclusions</h3><div>A specialized unit markedly improves therapeutic optimization and reduces readmissions in these patients.</div></div>","PeriodicalId":94354,"journal":{"name":"Revista clinica espanola","volume":"226 1","pages":"Article 502419"},"PeriodicalIF":0.0,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145812618","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-01DOI: 10.1016/j.rceng.2025.502414
N. García-Alvarado , M.I. Morales-Casado , P. Beneyto-Martín
Introduction
The prevalence of autoimmune comorbidities in patients with multiple sclerosis (MS) has been the subject of numerous epidemiological studies. Due to the presence of biases and the heterogeneity of the publications, this association has not been firmly demonstrated. The aim of our study is to establish the prevalence of autoimmune diseases in MS patients from our region (Castilla-La Mancha, Spain) and to compare it with the prevalence of autoimmune diseases in a non-MS population, in order to strengthen the evidence for an association between MS and other autoimmune conditions.
Patients and methods
We conducted a retrospective, non-interventional, multicenter study analyzing the electronic medical records of 3,309,298 patients in the Castilla-La Mancha area (Spain) using an artificial intelligence system.
Results
22.5% of MS patients had at least one other autoimmune disease. Hypothyroidism, followed by type 1 diabetes mellitus and psoriasis, were the three most frequent autoimmune diseases in the MS cohort.
Conclusions
In the present study, we observed an association between most of the autoimmune diseases studied and MS when comparing their prevalence in the MS population versus the non-MS population. Confirmation of these findings could lead to changes in preventive strategies, diagnostic protocols, and therapeutic approaches for MS patients. Large-scale data analysis using artificial intelligence may help resolve epidemiological questions that remain unanswered to date.
{"title":"Autoimmune comorbidities in multiple sclerosis. A population-based study using artificial intelligence","authors":"N. García-Alvarado , M.I. Morales-Casado , P. Beneyto-Martín","doi":"10.1016/j.rceng.2025.502414","DOIUrl":"10.1016/j.rceng.2025.502414","url":null,"abstract":"<div><h3>Introduction</h3><div>The prevalence of autoimmune comorbidities in patients with multiple sclerosis (MS) has been the subject of numerous epidemiological studies. Due to the presence of biases and the heterogeneity of the publications, this association has not been firmly demonstrated. The aim of our study is to establish the prevalence of autoimmune diseases in MS patients from our region (Castilla-La Mancha, Spain) and to compare it with the prevalence of autoimmune diseases in a non-MS population, in order to strengthen the evidence for an association between MS and other autoimmune conditions.</div></div><div><h3>Patients and methods</h3><div>We conducted a retrospective, non-interventional, multicenter study analyzing the electronic medical records of 3,309,298 patients in the Castilla-La Mancha area (Spain) using an artificial intelligence system.</div></div><div><h3>Results</h3><div>22.5% of MS patients had at least one other autoimmune disease. Hypothyroidism, followed by type 1 diabetes mellitus and psoriasis, were the three most frequent autoimmune diseases in the MS cohort.</div></div><div><h3>Conclusions</h3><div>In the present study, we observed an association between most of the autoimmune diseases studied and MS when comparing their prevalence in the MS population versus the non-MS population. Confirmation of these findings could lead to changes in preventive strategies, diagnostic protocols, and therapeutic approaches for MS patients. Large-scale data analysis using artificial intelligence may help resolve epidemiological questions that remain unanswered to date.</div></div>","PeriodicalId":94354,"journal":{"name":"Revista clinica espanola","volume":"226 1","pages":"Article 502414"},"PeriodicalIF":0.0,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145812661","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号