JBJuan-Jose Beunza

Juan-Jose Beunza

*This is an algorithmically generated author record based on publicly available information.

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


Juan-Jose Beunza has a total of 37 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Real-World Deployment and Operational Evaluation of an On-Premise Thermal IoMT System in an Intensive Care Unit

    by Eva Avilés, Jose-Luis Lafuente, Samuel González, Juan-Jose Beunza

  • Open Access

    Leveraging Large Language Models for Accurate Retrieval of Patient Information From Medical Reports: Systematic Evaluation Study

    by Angel Manuel Garcia-Carmona, Maria-Lorena Prieto, Enrique Puertas, Juan-Jose Beunza

    Background: The digital transformation of health care has introduced both opportunities and challenges, particularly in managing and analyzing the vast amounts of unstructured medical data generated daily. There is a need to explore the feasibility of generative solutions in extracting data from medical reports, categorized by specific criteria. Objective: This study aimed to investigate the application of large language models (LLMs) for the automated extraction of structured information from unstructured medical reports, using the LangChain framework in Python. Methods: Through a systematic evaluation of leading LLMs—GPT-4o, Llama 3, Llama 3.1, Gemma 2, Qwen 2, and Qwen 2.5—using zero-shot prompting techniques and embedding results into a vector database, this study assessed the performance of LLMs in extracting patient demographics, diagnostic details, and pharmacological data. Results: Evaluation metrics, including accuracy, precision, recall, and F1-score, revealed high efficacy across most categories, with GPT-4o achieving the highest overall performance (91.4% accuracy). Conclusions: The findings highlight notable differences in precision and recall between models, particularly in extracting names and age-related information. There were challenges in processing unstructured medical text, including variability in model performance across data types. Our findings demonstrate the feasibility of integrating LLMs into health care workflows; LLMs offer substantial improvements in data accessibility and support clinical decision-making processes. In addition, the paper describes the role of retrieval-augmented generation techniques in enhancing information retrieval accuracy, addressing issues such as hallucinations and outdated data in LLM outputs. Future work should explore the need for optimization through larger and more diverse training datasets, advanced prompting strategies, and the integration of domain-specific knowledge to improve model generalizability and precision.


  • Open Access

    Leveraging Large Language Models for Accurate Retrieval of Patient Information From Medical Reports: Systematic Evaluation Study (Preprint)

    by Angel Manuel Garcia-Carmona, Maria-Lorena Prieto, Enrique Puertas, Juan-Jose Beunza

    BACKGROUND The digital transformation of health care has introduced both opportunities and challenges, particularly in managing and analyzing the vast amounts of unstructured medical data generated daily. There is a need to explore the feasibility of generative solutions in extracting data from medical reports, categorized by specific criteria. OBJECTIVE This study aimed to investigate the application of large language models (LLMs) for the automated extraction of structured information from unstructured medical reports, using the LangChain framework in Python. METHODS Through a systematic evaluation of leading LLMs—GPT-4o, Llama 3, Llama 3.1, Gemma 2, Qwen 2, and Qwen 2.5—using zero-shot prompting techniques and embedding results into a vector database, this study assessed the performance of LLMs in extracting patient demographics, diagnostic details, and pharmacological data. RESULTS Evaluation metrics, including accuracy, precision, recall, and F1-score, revealed high efficacy across most categories, with GPT-4o achieving the highest overall performance (91.4% accuracy). CONCLUSIONS The findings highlight notable differences in precision and recall between models, particularly in extracting names and age-related information. There were challenges in processing unstructured medical text, including variability in model performance across data types. Our findings demonstrate the feasibility of integrating LLMs into health care workflows; LLMs offer substantial improvements in data accessibility and support clinical decision-making processes. In addition, the paper describes the role of retrieval-augmented generation techniques in enhancing information retrieval accuracy, addressing issues such as hallucinations and outdated data in LLM outputs. Future work should explore the need for optimization through larger and more diverse training datasets, advanced prompting strategies, and the integration of domain-specific knowledge to improve model generalizability and precision.


  • Open Access

    Enhanced Medical Data Extraction: Leveraging LLMs for Accurate Retrieval of Patient Information from Medical Reports

    by Angel M Garcia-Carmona, Maria-Lorena Prieto, Enrique Puertas, Juan-Jose Beunza

    This study presents a strategic approach to developing applications focused on implementing Large Language Models using the Langchain framework in Python. Three language models are highlighted: GPT-3.5 (turbo mode), LLaMA 2, and Vicuna 7B, each with their distinctive features and capabilities. The methodology used is described in detail, including data extraction from medical reports using zero-shot prompting data extraction techniques, interaction with language models, and structured storage of results. The performance of the models in data extraction is evaluated, presenting metrics such as precision, recall, and F1 score. The results demonstrate high model capability in extracting information, although areas for improvement are identified, particularly in data extraction precision. In conclusion, the efficacy of the models in extracting information from medical histories is not considerably acknowledged, with an emphasis on the importance of improving precision and increasing the volume of trained data for future research in healthcare digitalization.


  • Open Access

    Implications of the European Union's New Regulation on Artificial Intelligence (2024) on its Clinical Implementation (Preprint)

    by Nuria Rodríguez Dos Santos, Samuel González, Rubén-David Herrero, Juan-Jose Beunza

    UNSTRUCTURED On March 13, 2024, the Parliament sanctioned the inaugural comprehensive artificial intelligence (AI) statute globally. The European Union deems the governance of this technology as crucial, in light of potential infringements upon fundamental rights and public freedoms, detrimental impacts on the environment, and the integrity of democratic frameworks. Conversely, the undeniable advances in the medical domain necessitate the regulation of such technology to bolster investments and research, thereby providing the requisite legal certainty for users/patients, professionals, enterprises, and investors. The Regulation acknowledges the potential perils certain AI applications pose to fundamental rights, enacting prohibitions on practices such as biometric categorization, emotion recognition, and social scoring, among others. Nevertheless, the Regulation does not neglect the significance of medical research and, recognizing the imperative for support of new technologies and innovation, permits research with fewer constraints compared to the commercial use of AI algorithms in healthcare.

  • Open Access

    Continuous and Non-Invasive Lactate Monitoring Techniques in Critical Care Patients

    by Jose-Luis Lafuente, Samuel González, Clara Aibar, Desirée Rivera, Eva Avilés, Juan-Jose Beunza

    Lactate, once merely regarded as an indicator of tissue hypoxia and muscular fatigue, has now gained prominence as a pivotal biomarker across various medical disciplines. Recent research has unveiled its critical role as a high-value prognostic marker in critical care medicine. The current practice of lactate detection involves periodic blood sampling. This approach is invasive and confined to measurements at six-hour intervals, leading to resource expenditure, time consumption, and patient discomfort. This review addresses non-invasive sensors that enable continuous monitoring of lactate in critical care patients. After the introduction, it discusses the iontophoresis system, followed by a description of the structural materials that are universally employed to create an interface between the integumentary system and the sensor. Subsequently, each method is detailed according to its physical principle, outlining its advantages, limitations, and pertinent aspects. The study concludes with a discussion and conclusions, aiming at the design of an intelligent sensor (Internet of Medical Things or IoMT) to facilitate continuous lactate monitoring and enhance the clinical decision-making support system in critical care medicine.


  • Open Access

    Systematic Review of Automated Diuresis Measurement in Critically Ill Patients

    by Jose-Luis Lafuente, Samuel González, Vicente Gómez-Tello, Enrique Puertas, Eva Avilés, Juan-Jose Beunza

    The measurement of urinary flow is a vital medical indicator for critically ill patients in intensive care units. However, there is a clinical need to automate the real-time measurement of diuresis using Internet of Medical Things devices, allowing continuous monitoring of urine flow. A systematic review of scientific literature, patents, and available commercial products was conducted, leading to the conclusion that there is no suitable device to fulfill this need. We identified six characteristics that such a device should possess: minimizing contact with urine, detecting changes in flow patterns, the ability to record minute-by-minute data, capable of sending early alerts, not relying on exclusive disposable components, and being user-friendly for clinical professionals. Additionally, cost-effectiveness is crucial, encompassing the device, infrastructure, maintenance, and usage. Plain Language Summary: Continuous monitoring of urinary output in intensive care patients provides vital insights, potentially revealing low cardiac output, renal hypoperfusion, or deteriorating renal function. Currently, these measurements are done manually, prompting a need for automated, real-time monitoring using Internet of Medical Things (IoMT) devices. Current automatic urinometers are not widely used due to high costs, maintenance requirements, and logistical complications. However, we propose developing an affordable, updated urinometer that meets clinician needs and can be widely used in hospitals. This innovation would enable early intervention strategies to reverse or improve renal perfusion and prevent potential toxic treatments. The shift from manual to automated monitoring is likely, facilitating real-time data integration such as blood pressure, heart rate, and temperature. This can lead to personalized medical approaches and predictive algorithms for renal function deterioration. As technology, IoMT, big data, and artificial intelligence (AI) evolve, these advancements will be indispensable for hospital medicine, supporting precision medicine and efficient clinical operation. While we cannot predict the definitive method for continuous diuresis measurement, we anticipate the adoption of these technologies in the near future, improving diagnostics and clinical practice. If successful, the application could extend beyond critical care into postoperative or home care, not only for diuresis but for any fluid flow, enabling early pathology identification and potentially revealing diagnostic patterns through AI.