This paper describes the system submitted by our team to the MultiClinAI shared task at the 11th SMM4H-HeaRD Workshop (ACL 2026). The task addresses multilingual clinical Named Entity Recognition (NER) for three entity types (Disease, Procedure, and Symptom) in Spanish clinical texts. Our approach fine-tunes MrBERT-biomed, a domain-adapted ModernBERT model pre-trained on biomedical corpora, using multilingual clinical data from seven European languages. We train independent entity-specific models, each optimized via Bayesian hyperparameter search with Optuna, and apply a deterministic post-processing step that aligns predicted spans to word boundaries. On the official test set, our system achieves overall strict micro-F1 scores of 0.7453, 0.7107, and 0.6603 for Disease, Procedure, and Symptom, respectively.

Discharge summaries are critical documents for continuity of care, yet their manual creation imposes significant burdens on clinical staff. This systematic literature review examines current approaches to automatic generation of discharge summaries using Natural Language Processing (NLP) and Large Language Models (LLMs). Following the Kitchenham guidelines for systematic reviews in software engineering, we searched Scopus and PubMed databases for studies published between 2023 and 2026, identifying 9 primary studies from an initial pool of 102 papers. Our analysis reveals that GPT-4 and its variants dominate current research (appearing in 6 of 9 studies), while open-source alternatives like LLaMA show promise for privacy-preserving deployments. Evaluation primarily relies on automatic metrics (ROUGE, BLEU) combined with human expert assessment. Key challenges include hallucination rates ranging from 33% to 64%, information omission, integration with Electronic Health Record (EHR) systems, and context window limitations. Studies addressing factuality employ human-in-the-loop validation, prompt engineering techniques, and knowledge graph-based correction mechanisms. Despite these challenges, recent implementations demonstrate clinical feasibility, with one study achieving a 94.35% System Usability Score. This review provides a comprehensive synthesis of the state-of-the-art and identifies opportunities for future research in this rapidly evolving field.

Medical discharge reports frequently contain highly technical language that creates significant communication barriers between healthcare professionals and patients, potentially compromising treatment adherence and post-discharge care quality. In this paper, we present SimpliMED, a modular system designed to automatically simplify cardiology discharge reports using Large Language Models (LLMs) and advanced Natural Language Processing techniques (NLP). Our architecture integrates section-based preprocessing with specialized prompts, explicit handling of medical abbreviations, and therapeutic explanations of medications to enhance accessibility. We evaluate our system using a corpus of 307 anonymized cardiology discharge reports from a Spanish medical center. For abbreviation detection, our fine-tuned Small Language Model (SLM) achieves an F1-score of 0.90, significantly outperforming regex-based approaches (F1: 0.67). For medication recognition, we achieve F1-scores of 0.91 for commercial names and 0.70 for active principles. We also contribute a therapeutic dictionary containing 14,611 medications with patient-friendly explanations extracted from the Spanish Agency of Medicines. Expert evaluation by two cardiologists yields an overall quality score of 75%, with highest performance for admission reason (91%) and current illness (75%) sections. While results demonstrate the potential of LLM-based medical text simplification for Spanish clinical language, we identify areas requiring further development before clinical deployment.

An effective disclosure of scientific knowledge and advancements to the general public is often hindered by the complexity of the technical language used in research which often results very difficult, if not impossible, for non-experts to understand. In this paper we present the approach developed by the SINAI team as the result of our participation in BioLaySumm shared task hosted by the BioNLP workshop at ACL 2024. Our approach stems from the experimentation we performed in order to test the ability of state-of-the-art pre-trained large language models, namely GPT 3.5, GPT 4 and Llama-3, to tackle this task in a few-shot manner. In order to improve this baseline, we opted for fine-tuning Llama-3 by applying parameter-efficient methodologies. The best performing system which resulted from applying self-play fine tuning method which allows the model to improve while learning to distinguish between its own generations from the previous step from the gold standard summaries. This approach achieved 0.4205 ROUGE-1 score and 0.8583 BERTScore.

Pre-trained text encoders have rapidly advanced the state-of-the-art on many Natural Language Processing tasks. This paper presents the use of transfer learning methods applied to the automatic detection of codes in radiological reports in Spanish. Assigning codes to a clinical document is a popular task in NLP and in the biomedical domain. These codes can be of two types: standard classifications (e.g. ICD-10) or specific to each clinic or hospital. In this study we show a system using specific radiology clinic codes. The dataset is composed of 208,167 radiology reports labeled with 89 different codes. The corpus has been evaluated with three methods using the BERT model applied to Spanish: Multilingual BERT, BETO and XLM. The results are interesting obtaining 70% of F1-score with a pre-trained multilingual model.

Over time the use of social networks is becoming very popular platforms for sharing health related information. Social Media Mining for Health Applications (SMM4H) provides tasks such as those described in this document to help manage information in the health domain. This document shows the first participation of the SINAI group. We study approaches based on machine learning and deep learning to extract adverse drug reaction mentions from Twitter. The results obtained in the tasks are encouraging, we are close to the average of all participants and even above in some cases.

This paper describes the participation of the SINAI-DL team at RumourEval (Task 7 in SemEval 2019, subtask A: SDQC). SDQC addresses the challenge of rumour stance classification as an indirect way of identifying potential rumours. Given a tweet with several replies, our system classifies each reply into either supporting, denying, questioning or commenting on the underlying rumours. We have applied data augmentation, temporal expressions labelling and transfer learning with a four-layer neural classifier. We achieve an accuracy of 0.715 with the official run over reply tweets.

This paper describes the participation of the SINAI-DL team at Task 5 in SemEval 2019, called HatEval. We have applied some classic neural network layers, like word embeddings and LSTM, to build a neural classifier for both proposed tasks. Due to the small amount of training data provided compared to what is expected for an adequate learning stage in deep architectures, we explore the use of paraphrasing tools as source for data augmentation. Our results show that this method is promising, as some improvement has been found over non-augmented training sets.

Mental health is one of the main concerns of today’s society. Early detection of symptoms can greatly help people with mental disorders. People are using social networks more and more to express emotions, sentiments and mental states. Thus, the treatment of this information using NLP technologies can be applied to the automatic detection of mental problems such as eating disorders. However, the first step to solving the problem should be to provide a corpus in order to evaluate our systems. In this paper, we specifically focus on detecting anorexia messages on Twitter. Firstly, we have generated a new corpus of tweets extracted from different accounts including anorexia and non-anorexia messages in Spanish. The corpus is called SAD: Spanish Anorexia Detection corpus. In order to validate the effectiveness of the SAD corpus, we also propose several machine learning approaches for automatically detecting anorexia symptoms in the corpus. The good results obtained show that the application of textual classification methods is a promising option for developing this kind of system demonstrating that these tools could be used by professionals to help in the early detection of mental problems.

In this paper we describe a new named entity extraction system. Our work proposes a system for the identification and annotation of drug names in Spanish biomedical texts based on machine learning and deep learning models. Subsequently, a standardized code using Snomed is assigned to these drugs, for this purpose, Natural Language Processing tools and techniques have been used, and a dictionary of different sources of information has been built. The results are promising, we obtain 78% in F1 score on the first sub-track and in the second task we map with Snomed correctly 72% of the found entities.