Modelo basado en LLMs para apoyar la toma de decisiones en el triaje médico
| dc.contributor.advisor | Rodríguez Portela, Arles Ernesto | |
| dc.contributor.author | Martin Salcedo, Alejandro | |
| dc.contributor.cvlac | Rodríguez Portela, Arles Ernesto [0000536636] | |
| dc.contributor.googlescholar | Rodríguez Portela, Arles Ernesto [ck6NRGkAAAAJ&hl] | |
| dc.contributor.orcid | Rodríguez Portela, Arles Ernesto [0000000349106773] | |
| dc.contributor.researchgate | Rodríguez Portela, Arles Ernesto [Arles-Rodriguez] | |
| dc.date.accessioned | 2026-05-19T13:02:26Z | |
| dc.date.available | 2026-05-19T13:02:26Z | |
| dc.date.issued | 2026 | |
| dc.description | ilustraciones a color, diagramas | spa |
| dc.description.abstract | La creciente demanda en los servicios de urgencias y la necesidad de realizar procesos de triaje eficientes, consistentes y alineados con criterios clínicos estructurados representan un desafío para los sistemas de salud, particularmente en la estimación adecuada de los recursos y en el apoyo al razonamiento clínico para la correcta priorización del paciente durante la atención. En este contexto, el presente trabajo tuvo como objetivo desarrollar y evaluar un sistema basado en agentes sustentado en modelos de lenguaje de gran escala (LLMs), integrado con un esquema de recuperación aumentada (RAG), orientado al apoyo en la estimación de recursos y en la estructuración del razonamiento clínico para la priorización del paciente bajo el protocolo de triaje Emergency Severity Index (ESI). Se diseñó una arquitectura híbrida que combina generación de texto y recuperación contextual de información clínica, junto con una estrategia de estructuración progresiva del prompt alineada con el flujo de decisión del triaje. La evaluación se realizó mediante un enfoque técnico-clínico que incluyó métricas computacionales de similitud semántica y fidelidad del sistema, junto con validación por un médico con experiencia en triaje, considerando múltiples experimentos con dos variantes de prompt y tres modelos de lenguaje. Los resultados evidencian mejoras en la coherencia del razonamiento y una concordancia moderada con el criterio clínico, destacando el potencial de estos sistemas como herramientas de apoyo en la toma de decisiones bajo supervisión profesional. (Texto tomado de la fuente) | spa |
| dc.description.abstract | The growing demand on emergency services and the need for efficient, consistent triage processes aligned with structured clinical criteria present a challenge for healthcare systems, particularly in the accurate estimation of resources and in supporting clinical reasoning for proper patient prioritization during care. In this context, this study aimed to develop and evaluate an agent-based system supported by large-scale language models (LLMs), integrated with an augmented retrieval scheme (RAG), designed to support resource estimation and the structuring of clinical reasoning for patient prioritization under the Emergency Severity Index (ESI) triage protocol. A hybrid architecture was designed that combines text generation and contextual retrieval of clinical information, along with a progressive prompt structuring strategy aligned with the triage decision flow. The evaluation was conducted using a technical-clinical approach that included computational metrics of semantic similarity and system fidelity, along with validation by a physician with triage experience, considering multiple experiments with two prompt variants and three language models. The results show improvements in the coherence of reasoning and moderate agreement with clinical criteria, highlighting the potential of these systems as support tools in decision-making under professional supervision. | eng |
| dc.description.degreelevel | Maestría | |
| dc.description.degreename | Magíster en Ingeniería - Ingeniería de Sistemas y Computación | |
| dc.description.researcharea | Machine Learning / Data Science | |
| dc.format.extent | xi, 46 páginas | |
| dc.format.mimetype | application/pdf | |
| dc.identifier.instname | Universidad Nacional de Colombia | spa |
| dc.identifier.reponame | Repositorio Institucional Universidad Nacional de Colombia | spa |
| dc.identifier.repourl | https://repositorio.unal.edu.co/ | spa |
| dc.identifier.uri | https://repositorio.unal.edu.co/handle/unal/89994 | |
| dc.language.iso | spa | |
| dc.publisher | Universidad Nacional de Colombia | |
| dc.publisher.branch | Universidad Nacional de Colombia - Sede Bogotá | |
| dc.publisher.faculty | Facultad de Ingeniería | |
| dc.publisher.place | Bogotá, Colombia | |
| dc.publisher.program | Bogotá - Ingeniería - Maestría en Ingeniería - Ingeniería de Sistemas y Computación | |
| dc.relation.references | W. Soler, M Gómez Muñoz, E. Bragulat, and A. Álvarez. Triage: a key tool in emergency care. Technical report, 2010. | |
| dc.relation.references | Marium M. Raza, Kaushik P. Venkatesh, and Joseph C. Kvedar. Generative ai and large language models in health care: pathways to implementation, 12 2024. | |
| dc.relation.references | Yi Yang, Yitong Ma, Hao Feng, Yiming Cheng, and Zhu Han. Minimizing hallucinations and communication costs: Adversarial debate and voting mechanisms in llm-based multi-agents. Applied Sciences (Switzerland), 15, 4 2025. | |
| dc.relation.references | Ranjan Sapkota, Konstantinos I. Roumeliotis, and Manoj Karkee. Ai agents vs. agentic ai: A conceptual taxonomy, applications and challenges. Information Fusion, 126:103599, February 2025. | |
| dc.relation.references | Himanshu Pandey, Akhil Amod, Risa Labs, and Akhil@risalabs Ai Shivang. Advancing healthcare automation: Multi-agent system for medical necessity justification. Technical report, 2024. | |
| dc.relation.references | Yunfan Gao, Yun Xiong, Xinyu Gao, Kangxiang Jia, Jinliu Pan, Yuxi Bi, Yi Dai, Jiawei Sun, Meng Wang, and Haofen Wang. Retrieval-augmented generation for large language models: A survey. 3 2024. | |
| dc.relation.references | Wenxuan Wang, Zizhan Ma, Zheng Wang, Chenghan Wu, Jiaming Ji, Wenting Chen, Xiang Li, and Yixuan Yuan. A survey of LLM-based agents in medicine: How far are we from baymax? In Wanxiang Che, Joyce Nabende, Ekaterina Shutova, and Mohammad Taher Pilehvar, editors, Findings of the Association for Computational Linguistics: ACL 2025, pages 10345–10359, Vienna, Austria, July 2025. Association for Computational Linguistics. | |
| dc.relation.references | Sagar Goyal, Eti Rastogi, Sree Prasanna Rajagopal, Dong Yuan, Fen Zhao, Jai Chintagunta, Gautam Naik, and Jeff Ward. Healai: A healthcare llm for effective medical documentation. In WSDM 2024 - Proceedings of the 17th ACM International Conference on Web Search and Data Mining, pages 1167–1168. Association for Computing Machinery, Inc, 3 2024. | |
| dc.relation.references | Daniel Asensi Roch. Generación de informes médicos y diagnósticos automáticos, 2024-09-23. | |
| dc.relation.references | Meng Lu, Brandon Ho, Dennis Ren, and Xuan Wang. Triageagent: Towards better multi-agents collaborations for large language model-based clinical triage. Technical report, 11 2024. | |
| dc.relation.references | Roger Arturo Reyes Agamez. Factores que influyen en la clasificación triaje, una revisión de literatura. Technical report, Fundación Universitaria de Ciencias de la Salud, 2023. | |
| dc.relation.references | DeepSeek-AI, Aixin Liu, Bei Feng, Bing Xue, Bingxuan Wang, Bochao Wu, Chengda Lu, Chenggang Zhao, Chengqi Deng, Chenyu Zhang, Chong Ruan, Damai Dai, Daya Guo, Dejian Yang, Deli Chen, Dongjie Ji, Erhang Li, Fangyun Lin, Fucong Dai, Fuli Luo, Guangbo Hao, Guanting Chen, and et. al. Deepseek-v3 technical report, 2025. | |
| dc.relation.references | Aaron Grattafiori, Abhimanyu Dubey, Abhinav Jauhri, Abhinav Pandey, Abhishek Kadian, Ahmad Al-Dahle, Aiesha Letman, Akhil Mathur, Alan Schelten, Alex Vaughan, Amy Yang, Angela Fan, Anirudh Goyal, Anthony Hartshorn, Aobo Yang, Archi Mitra, Archie Sravankumar, Artem Korenev, Arthur Hinsvark, Arun Rao, and et. al. The llama 3 herd of models, 2024. | |
| dc.relation.references | Alberto Tomás Delso Vicente, Marisol Carvajal Camperos, and Daniel Ángel Corral De La Mata. The evolution of natural language processing and its influence on artificial intelligence: A review and future research directions. European Public and Social Innovation Review, 10:1–23, 1 2025. | |
| dc.relation.references | Zichong Wang, Zhibo Chu, Thang Viet Doan, Shiwen Ni, Min Yang, and Wenbin Zhang. History, development, and principles of large language models-an introductory survey. 9 2024. | |
| dc.relation.references | Mengchao Ren. Advancements and applications of large language models in natural language processing: A comprehensive review. Applied and Computational Engineering, 97:55–63, 11 2024. | |
| dc.relation.references | Wazib Ansar, Saptarsi Goswami, and Amlan Chakrabarti. From transformers to llms: A systematic survey of efficiency considerations in nlp. 12 2025. | |
| dc.relation.references | Nihar Ranjan, Kaushal Mundada, Kunal Phaltane, and Saim Ahmad. A survey on techniques in nlp. International Journal of Computer Applications, 134(8):6–9, January 2016. | |
| dc.relation.references | Daniel Jurafsky and James H. Martin. Speech and Language Processing: An Introduction to Natural Language Processing, Computational Linguistics, and Speech Recognition, with Language Models. 3rd edition, 2026. Online manuscript released January 6, 2026. | |
| dc.relation.references | Julia Hirschberg and Christopher D. Manning. Advances in natural language processing. Science, 349(6245):261–266, 2015. | |
| dc.relation.references | Panteleimon Krasadakis, Evangelos Sakkopoulos, and Vassilios S. Verykios. A survey on challenges and advances in natural language processing with a focus on legal informatics and low-resource languages, 2 2024. | |
| dc.relation.references | Tomas Mikolov, Kai Chen, Greg Corrado, and Jeffrey Dean. Efficient estimation of Word representations in vector space. 9 2013. | |
| dc.relation.references | Jeffrey Pennington, Richard Socher, and Christopher D Manning. Glove: Global vectors for Word representation. Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 1532–1543, 2014. | |
| dc.relation.references | Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Lukasz Kaiser, and Illia Polosukhin. Attention is all you need, 2023. | |
| dc.relation.references | Rishi Bommasani, Drew A. Hudson, Ehsan Adeli, Russ Altman, Simran Arora, Sydney von Arx, Michael S. Bernstein, Jeannette Bohg, Antoine Bosselut, Emma Brunskill, Erik Brynjolfsson, S. Buch, Dallas Card, Rodrigo Castellon, Niladri S. Chatterji, Annie S. Chen, Kathleen A. Creel, and et. al. On the opportunities and risks of foundation models. ArXiv, abs/2108.07258, 2021. | |
| dc.relation.references | Jason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma, Brian Ichter, Fei Xia, Ed Chi, Quoc Le, and Denny Zhou. Chain-of-thought prompting elicits reasoning in large language models, 2023. | |
| dc.relation.references | Matthew E. Peters, Mark Neumann, Mohit Iyyer, Matt Gardner, Christopher Clark, Kenton Lee, and Luke Zettlemoyer. Deep contextualized word representations. In Marilyn Walker, Heng Ji, and Amanda Stent, editors, Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long Papers), pages 2227–2237, New Orleans, Louisiana, June 2018. Association for Computational Linguistics. | |
| dc.relation.references | Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. Bert: Pre-training of deep bidirectional transformers for language understanding, 2019. | |
| dc.relation.references | Tom B. Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah, Jared Kaplan, Prafulla Dhariwal, Arvind Neelakantan, Pranav Shyam, Girish Sastry, Amanda Askell, Sandhini Agarwal, Ariel Herbert-Voss, Gretchen Krueger, Tom Henighan, Rewon Child, Aditya Ramesh, Daniel M. Ziegler, Jeffrey Wu, Clemens Winter, Christopher Hesse, Mark Chen, Eric Sigler, Mateusz Litwin, Scott Gray, Benjamin Chess, Jack Clark, Christopher Berner, Sam McCandlish, Alec Radford, Ilya Sutskever, and Dario Amodei. Language models are few-shot learners. 7 2020. | |
| dc.relation.references | Patrick Lewis, Ethan Perez, Aleksandra Piktus, Fabio Petroni, Vladimir Karpukhin, Naman Goyal, Heinrich Küttler, Mike Lewis, Wen tau Yih, Tim Rocktäschel, Sebastian Riedel, and Douwe Kiela. Retrieval-augmented generation for knowledge-intensive nlp tasks. 4 2021. | |
| dc.relation.references | Andrew Brown, Muhammad Roman, and Barry Devereux. A systematic literature review of retrieval-augmented generation: Techniques, metrics, and challenges. 9 2025. | |
| dc.relation.references | Carla González Cubells. El triaje enfermero asistido por inteligencia artificial: Revisión sistemática. Journal Nursing Valencia, 5, 12 2025. | |
| dc.relation.references | Mashale Ali Alshadeedi, Kathiyah Ali Ghazwani, Faisal Ayedh, A Alzahrani, and Jamaan Alamri. Challenges of workforce shortages and workload among hospital healthcare staff a review submitted by the department of physical therapy, radiology, and health services. The Review of Diabetic Studies, pages 1083–1088, 9 2025. | |
| dc.relation.references | Nicole R. Hodgson, Richard Kwun, Chad Gorbatkin, Jeanie Davies, and Jonathan Fisher. Emergency department responses to nursing shortages. International Journal of Emergency Medicine, 17, 12 2024. | |
| dc.relation.references | Hiroyuki Nakao, Isao Ukai, and Joji Kotani. A review of the history of the origin of triage from a disaster medicine perspective. Acute Medicine & Surgery, 4:379–384, 10 2017. | |
| dc.relation.references | Pedro Gargantilla Madera and Berta María Martin Cabrejas. Los orígenes militares del triaje. Triage origins in the military. Emergencias, 31:205–206, 6 2019. | |
| dc.relation.references | Josep Gómez Jiménez and Pilar Ramón-Pardo. MANUAL para la IMPLEMENTACION de un SISTEMA de TRIAJE para los CUARTOS de URGENCIAS. OPS, 2010. | |
| dc.relation.references | Tomás Hernández Ruipérez, César Leal Costa, María de Gracia Adánez Martínez, Bartolomé García Pérez, Daniel Nova López, and José Luis Díaz Agea. Evidencias de validez del sistema de triaje emergency severity index en un servicio de urgencias de un hospital general. Emergencias, 27:301–306, 2015. | |
| dc.relation.references | Karan Singhal, Shekoofeh Azizi, Tao Tu, S. Sara Mahdavi, Jason Wei, Hyung Won Chung, Nathan Scales, Ajay Tanwani, Heather Cole-Lewis, Stephen Pfohl, Perry Payne, Martin Seneviratne, Paul Gamble, Chris Kelly, Nathaneal Scharli, Aakanksha Chowdhery, Philip Mansfield, Blaise Aguera y Arcas, Dale Webster, Greg S. Corrado, Yossi Matias, Katherine Chou, Juraj Gottweis, Nenad Tomasev, Yun Liu, Alvin Rajkomar, Joelle Barral, Christopher Semturs, Alan Karthikesalingam, and Vivek Natarajan. Large language models encode clinical knowledge. 12 2022. | |
| dc.relation.references | Tiffany H. Kung, Morgan Cheatham, Arielle Medenilla, Czarina Sillos, Lorie De Leon, Camille Elepaño, Maria Madriaga, Rimel Aggabao, Giezel Diaz-Candido, James Maningo, and Victor Tseng. Performance of chatgpt on usmle: Potential for ai-assisted medical education using large language models. PLOS Digital Health, 2, 2 2023. | |
| dc.relation.references | Junkai Li, Yunghwei Lai, Weitao Li, Jingyi Ren, Meng Zhang, Xinhui Kang, Siyu Wang, Peng Li, Ya-Qin Zhang, Weizhi Ma, and Yang Liu. Agent hospital: A simulacrum of hospital with evolvable medical agents. 1 2025. | |
| dc.relation.references | Zixiang Wang, Yinghao Zhu, Huiya Zhao, Xiaochen Zheng, Dehao Sui, Tianlong Wang, Wen Tang, Yasha Wang, Ewen Harrison, Chengwei Pan, Junyi Gao, and Liantao Ma. Colacare: Enhancing electronic health record modeling through large language modeldriven multi-agent collaboration. In WWW 2025 - Proceedings of the ACM Web Conference, pages 2255–2261. Association for Computing Machinery, Inc, 4 2025. | |
| dc.relation.references | Yubin Kim, Chanwoo Park, Hyewon Jeong, Yik Siu Chan, Xuhai Xu, Daniel McDuff, Hyeonhoon Lee, Marzyeh Ghassemi, Cynthia Breazeal, and Hae Won Park. Mdagents: An adaptive collaboration of llms for medical decision-making. 10 2024. | |
| dc.relation.references | Shubham Vatsal, Harsh Dubey, and Aditi Singh. Agentic ai in healthcare & medicine: A seven-dimensional taxonomy for empirical evaluation of llm-based agents. 10 2025. | |
| dc.relation.references | Aekkachai Fatai, Chakrit Sattayarom, Wiwat Laochai, and Ekkalak Faksook. Effectiveness of ai assisted esi triage on accuracy and selected outcomes in emergency nursing: A systematic review, 12 2025. | |
| dc.relation.references | Q. Shen and Q. Guo. Mimic-iv-ext triage instruction corpus, 2025. RRID: SCR 007345. | |
| dc.relation.references | Ary L. Goldberger, Luis A. N. Amaral, Leon Glass, Jeffrey M. Hausdorff, Plamen Ch. Ivanov, Roger G. Mark, Joseph E. Mietus, George B. Moody, Chung-Kang Peng, and H. Eugene Stanley. Physiobank, physiotoolkit, and physionet. Circulation, 101(23): e215–e220, 2000. | |
| dc.relation.references | Emergency Nurses Association. Emergency Severity Index (ESI) Handbook. Emergency Nurses Association, Schaumburg, IL, fifth edition, 2023. ESI Version 5. | |
| dc.relation.references | Sean Lee, Aamir Shakir, Darius Koenig, and Julius Lipp. Open source strikes bread - new fluffy embedding model. https://www.mixedbread.com/blog/mxbai-embed-largev1, 2024. Accessed: 2026-01. | |
| dc.relation.references | Chroma Inc. Chroma, ai native open source embedding database. https://www.trychroma.com, 2026. Accessed: 2026-01. | |
| dc.relation.references | Alejandro Martin Salcedo. Repositorio anexo. https://github.com/almartins1/TriajeProyecto, 2026. | |
| dc.relation.references | Tianyi Zhang, Varsha Kishore, Felix Wu, Kilian Q. Weinberger, and Yoav Artzi. Bertscore: Evaluating text generation with bert, 2020. | |
| dc.relation.references | Shahul Es, Jithin James, Luis Espinosa-Anke, and Steven Schockaert. Ragas: Automated evaluation of retrieval augmented generation, 2025. | |
| dc.relation.references | VibrantLabs. Ragas: Supercharge your llm application evaluations. https://github.com/vibrantlabsai/ragas, 2024. | |
| dc.rights.accessrights | info:eu-repo/semantics/openAccess | |
| dc.rights.license | Atribución-NoComercial 4.0 Internacional | |
| dc.rights.uri | http://creativecommons.org/licenses/by-nc/4.0/ | |
| dc.subject.ddc | 000 - Ciencias de la computación, información y obras generales::004 - Procesamiento de datos Ciencia de los computadores | |
| dc.subject.ddc | 000 - Ciencias de la computación, información y obras generales::006 - Métodos especiales de computación | |
| dc.subject.lemb | LINGUISTICA COMPUTACIONAL | spa |
| dc.subject.lemb | Computational linguistics | eng |
| dc.subject.lemb | INDIZACION AUTOMATICA | spa |
| dc.subject.lemb | Automatic indexing | eng |
| dc.subject.lemb | APRENDIZAJE AUTOMATICO (INTELIGENCIA ARTIFICIAL) | spa |
| dc.subject.lemb | Machine learning | eng |
| dc.subject.lemb | APRENDIZAJE BASADO EN EXPLICACIONES | spa |
| dc.subject.lemb | Explanation-based learning | eng |
| dc.subject.lemb | SISTEMAS DE COMUNICACION EN SERVICIOS MEDICOS DE URGENCIAS | spa |
| dc.subject.lemb | Emergency medical services--Communication systems | eng |
| dc.subject.proposal | LLMs | spa |
| dc.subject.proposal | RAG | spa |
| dc.subject.proposal | Triaje | spa |
| dc.subject.proposal | ESI | spa |
| dc.subject.proposal | LLMs | eng |
| dc.subject.proposal | RAG | eng |
| dc.subject.proposal | Triage | eng |
| dc.subject.proposal | ESI | eng |
| dc.title | Modelo basado en LLMs para apoyar la toma de decisiones en el triaje médico | spa |
| dc.title.translated | LLM-based model to support decision-making in medical triage | eng |
| dc.type | Trabajo de grado - Maestría | |
| dc.type.coar | http://purl.org/coar/resource_type/c_bdcc | |
| dc.type.coarversion | http://purl.org/coar/version/c_ab4af688f83e57aa | |
| dc.type.content | Text | |
| dc.type.driver | info:eu-repo/semantics/masterThesis | |
| dc.type.redcol | http://purl.org/redcol/resource_type/TM | |
| dc.type.version | info:eu-repo/semantics/acceptedVersion | |
| dcterms.audience.professionaldevelopment | Estudiantes | |
| dcterms.audience.professionaldevelopment | Investigadores | |
| dcterms.audience.professionaldevelopment | Maestros | |
| dcterms.audience.professionaldevelopment | Público general | |
| oaire.accessrights | http://purl.org/coar/access_right/c_abf2 |
Archivos
Bloque original
1 - 1 de 1
Cargando...
- Nombre:
- TrabajoFinal_AlejandroMartinSalcedo_Corregido.pdf
- Tamaño:
- 968.16 KB
- Formato:
- Adobe Portable Document Format
- Descripción:
- Tesis de Maestría en Ingeniería de Sistemas y Computación

