Diseño y desarrollo de una herramienta basada en procesamiento de lenguaje natural mediado por Deep Learning para evaluación de registros clínicos electrónicos

Bonilla Arandia, Daniel

Diseño y desarrollo de una herramienta basada en procesamiento de lenguaje natural mediado por Deep Learning para evaluación de registros clínicos electrónicos

dc.contributor.advisor	Delgado Roman, Carlos Ignacio
dc.contributor.author	Bonilla Arandia, Daniel
dc.date.accessioned	2024-07-16T00:52:06Z
dc.date.available	2024-07-16T00:52:06Z
dc.date.issued	2024-06
dc.description.abstract	Las actividades de evaluación de registros clínicos electrónicos (EHR) como la medición de adherencia a Guías de Práctica Clínica (GPC) presentan un reto para los prestadores de servicios de salud, puesto que su adecuada implementación permite reducir la variabilidad de la atención y costos de atención. Sin embargo, es una tarea que requiere una revisión de registros elevado y retroalimentación constante para considerarse exitosa. Esta investigación presenta el diseño y desarrollo de un prototipo basado en Procesamiento de Lenguaje Natural (NLP) mediado por Deep Learning para evaluar la adherencia a Guías de Práctica Clínica (GPC) en registros clínicos electrónicos, haciendo uso de un modelo grande de lenguaje (LLM) preentrenado y la combinación de Prompt Engineering y Retrieval Augmented Generation (RAG) para analizar las historias clínicas y determinar la adherencia.
dc.description.abstractenglish	Evaluation activities for electronic health records (EHR), such as measuring adherence to Clinical Practice Guidelines (CPGs), present a challenge for healthcare providers, as proper implementation helps reduce variability in care and healthcare costs. However, it is a task that requires extensive record review and constant feedback to be considered successful. This research presents the design and development of a prototype based on Natural Language Processing (NLP) mediated by Deep Learning to evaluate adherence to Clinical Practice Guidelines (CPGs) in electronic health records. It utilizes a large pre-trained language model (LLM) and combines Prompt Engineering with Retrieval Augmented Generation (RAG) to analyze medical records and determine adherence.
dc.description.sponsorship	Clinica Universidad de la Sabana
dc.identifier.uri	https://hdl.handle.net/20.500.12495/12672
dc.language.iso	es
dc.relation.references	Abid, A., Abdalla, A., Abid, A., Khan, D., Alfozan, A., & Zou, J. (2019). Gradio: Hassle-Free Sharing and Testing of ML Models in the Wild. arXiv. https://doi.org/10.48550/arXiv.1906.02569
dc.relation.references	Almazrouei, E., Alobeidli, H., Alshamsi, A., Cappelli, A., & Cojocaru, R. (2023). The Falcon Series of Open Language Models. arXiv. https://doi.org/10.48550/arXiv.2311.16867
dc.relation.references	Anyscale. (2024). anyscale. anyscale. https://www.anyscale.com/
dc.relation.references	Baker, R., Camosso-Stefinovic, J., Gillies, C., Shaw, E. J., Cheater, F., Flottorp, S., & Robertson, N. (2010). Tailored interventions to overcome identified barriers to change: effects on professional practice and health care outcomes. In The Cochrane Collaboration (Ed.), Cochrane Database of Systematic Reviews (p. 1). Wiley. https://doi.org/10.1002/14651858.CD005470.pub2
dc.relation.references	Balaguer, A., Benara, V., Cunha, R. L. F., Filho, R. M. E., & Hendry, T. (2024). RAG vs Fine-tuning: Pipelines, Tradeoffs, and a Case Study on Agriculture. arXiv. https://doi.org/10.48550/arXiv.2401.08406
dc.relation.references	Beeching, E., Fourrier, C., Habib, N., & Han, S. (2023). Open LLM Leaderboard. Hugging Face. https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard
dc.relation.references	Belkada, Y., Dettmers, T., Pagnoni, A., Gugger, S., & Mangrulkar, S. (2023, May 24). Making LLMs even more accessible with bitsandbytes, 4-bit quantization and QLoRA. Hugging Face. https://huggingface.co/blog/4bit-transformers-bitsandbytes
dc.relation.references	Burgers, J. S. (2003). Towards evidence-based clinical practice: an international survey of 18 clinical guideline programs. International Journal for Quality in Health Care, 15(1), 31–45. https://doi.org/10.1093/intqhc/15.1.31
dc.relation.references	Cabana, M. D., Rand, C. S., Powe, N. R., Wu, A. W., Wilson, M. H., Abboud, P. A. C., & Rubin, H. R. (1999). Why Don’t Physicians Follow Clinical Practice Guidelines?: A Framework for Improvement. JAMA, 282(15), 1458. https://doi.org/10.1001/jama.282.15.1458
dc.relation.references	Committee on Standards for Developing Trustworthy Clinical Practice Guidelines. (2011). Clinical Practice Guidelines We Can Trust. National Academies Press. https://doi.org/10.17226/13058
dc.relation.references	Congreso de Colombia. (2015). ley 1751 de 2015. Ministerio de Salud y Protección Social. https://www.minsalud.gov.co/normatividad_nuevo/ley%201751%20de%202015.pdf
dc.relation.references	Consejo Asesor Para el Sistema Único de Acreditación en Salud. (2011). Manual de Acreditación en Salud Ambulatorio y Hospitalario de Colombia. Estándares de Acreditación. https://acreditacionensalud.org.co/estandares-de-acreditacion/
dc.relation.references	Cuenta de alto costo. (2023, June 9). Día mundial de la prevención del cáncer de próstata 2023 - Cuenta de Alto Costo. Cuenta de alto Costo. https://cuentadealtocosto.org/cancer/dia-mundial-de-la-prevencion-del-cancer-de-prostata-2023
dc.relation.references	Dettmers, T., Pagnoni, A., Holtzman, A., & Zettlemoyer, L. (2023). QLoRA: Efficient Finetuning of Quantized LLMs. arXiv. https://doi.org/10.48550/arXiv.2305.14314
dc.relation.references	Devlin, J., Chang, M. W., Lee, K., & Toutanova, K. (2019). BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding. arXiv. https://doi.org/10.48550/arXiv.1810.04805
dc.relation.references	Douze, M., Guzhva, A., Deng, C., Johnson, J., & Szilvasy, G. (2024). The Faiss library. arXiv. https://doi.org/10.48550/arXiv.2401.08281
dc.relation.references	Essel, V., Vuuren, U. V., Sa, A. D., Govender, S., Murie, K., Schlemmer, A., Gunst, C., Namane, M., Boulle, A., & Vries, E. D. (2015). Auditing chronic disease care: Does it make a difference? African Journal of Primary Health Care & Family Medicine, 7(1), 7. https://doi.org/10.4102/phcfm.v7i1.753
dc.relation.references	Fischer, F., Lange, K., Klose, K., Greiner, W., & Kraemer, A. (2016). Barriers and Strategies in Guideline Implementation—A Scoping Review. Healthcare, 4(3), 36. https://doi.org/10.3390/healthcare4030036
dc.relation.references	Fleming, S. L., Lozano, A., Haberkorn, W. J., Jindal, J. A., & Reis, E. P. (2023). MedAlign: A Clinician-Generated Dataset for Instruction Following with Electronic Medical Records. arXiv. https://doi.org/10.48550/arXiv.2308.14089
dc.relation.references	Gao, Y., Xiong, Y., Gao, X., Jia, K., & Pan, J. (2023). Retrieval-Augmented Generation for Large Language Models: A Survey. arXiv. https://doi.org/10.48550/arXiv.2312.10997
dc.relation.references	Grol, R., & Grimshaw, J. (2003). From best evidence to best practice: effective implementation of change in patients’ care. The Lancet, 362(9391), 1225–1230. https://doi.org/10.1016/S0140-6736(03)14546-1
dc.relation.references	Grol, R., Wensing, M., Eccles, M., & Davis, D. (2013). Improving Patient Care: The Implementation of Change in Health Care (Vol. 1). Wiley. https://doi.org/10.1002/9781118525975
dc.relation.references	Gugger, S., Debut, L., Wolf, T., Schmid, P., & Mueller, Z. (2022). Accelerate: Training and inference at scale made simple, efficient and adaptable. [Computer software]. Hugging Face. https://github.com/huggingface/accelerate
dc.relation.references	Gunasekar, S., Zhang, Y., Aneja, J., Mendes, C. C. T., & Del Giorno, A. (2023). Textbooks Are All You Need. arXiv. https://doi.org/10.48550/arXiv.2306.11644
dc.relation.references	HuggingFace. (n.d.). Sentence Transformers. HuggingFace. https://huggingface.co/sentence-transformers
dc.relation.references	Hysong, S. J., Best, R. G., & Pugh, J. A. (2006). Audit and feedback and clinical practice guideline adherence: Making feedback actionable. Implementation Science, 1(1), 9. https://doi.org/10.1186/1748-5908-1-9
dc.relation.references	Kong, A., Zhao, S., Chen, H., Li, Q., Qin, Y., Sun, R., & Zhou, X. (2023). Better Zero-Shot Reasoning with Role-Play Prompting. arXiv. https://doi.org/10.48550/arXiv.2308.07702
dc.relation.references	Kwak, G. H., & Hui, P. (2019). DeepHealth: Review and challenges of artificial intelligence in health informatics. arXiv. https://doi.org/10.48550/ARXIV.1909.00384
dc.relation.references	LangChain. (n.d.). LangChain. LangChain. https://github.com/langchain-ai
dc.relation.references	Llanio Navarro, R., Perdomo González, G., & Arús Soler,Enrique. (2007). Propedéutica clínica y semiología médica. Tomo 1. Editorial Ciencias Médicas. Editorial Pueblo y Educación.
dc.relation.references	Lewis, P., Perez, E., Piktus, A., Petroni, F., & Karpukhin, V. (2024). Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks. arXiv. https://doi.org/10.48550/arXiv.2005.11401
dc.relation.references	Li, J., Dada, A., Kleesiek, J., & Egger, J. (2023). ChatGPT in Healthcare: A Taxonomy and Systematic Review. medRxiv. https://doi.org/10.1101/2023.03.30.23287899
dc.relation.references	López-Ramos, H., Gómez Cusnir, P., Moreno, M., Patiño, G., Rasch-Isla, A., Dallos, A., Fernández, N., Jaramillo, A., & Vega, J. (2015). Guía de manejo de la hiperplasia prostática benigna. Sociedad Colombiana de Urología 2014. Urología Colombiana, 24(3), 187.e1–187.e32. https://doi.org/10.1016/j.uroco.2015.04.005
dc.relation.references	Louis, Antoine. (2020a, July 7). A Brief History of Natural Language Processing — Part 1. Medium. https://medium.com/@antoine.louis/a-brief-history-of-natural-language-processing-part-1-ffbcb937ebce
dc.relation.references	Louis, Antoine. (2020b, July 7). A Brief History of Natural Language Processing — Part 2. https://medium.com/@antoine.louis/a-brief-history-of-natural-language-processing-part-2-f5e575e8e37
dc.relation.references	Lugtenberg, M., Zegers-van Schaick, J. M., Westert, G. P., & Burgers, J. S. (2009). Why don’t physicians adhere to guideline recommendations in practice? An analysis of barriers among Dutch general practitioners. Implementation science: IS, 4, 54. https://doi.org/10.1186/1748-5908-4-54
dc.relation.references	Mashette, N. (2023). Building a Document-based Question Answering System with LangChain using Open Source LLM model…. Medium. https://medium.com/@nageshmashette032/building-a-document-based-question-answering-system-with-langchain-using-open-source-llm-model-346efe676be6
dc.relation.references	McGlynn, E. A., Asch, S. M., Adams, J., Keesey, J., Hicks, J., DeCristofaro, A., & Kerr, E. A. (2003). The Quality of Health Care Delivered to Adults in the United States. New England Journal of Medicine, 348(26), 2635–2645. https://doi.org/10.1056/NEJMsa022615
dc.relation.references	Meta Research. (2024a). faiss [Computer software]. Meta Research. https://github.com/facebookresearch/faiss
dc.relation.references	Meta Research. (2024b). Getting to know Llama 2: Everything you need to start building. github. https://github.com/meta-llama/llama-recipes/blob/main/recipes/quickstart/Getting_to_know_Llama.ipynb
dc.relation.references	Mickan, S., Burls, A., & Glasziou, P. (2011). Patterns of ‘leakage’ in the utilisation of clinical guidelines: a systematic review. Postgraduate Medical Journal, 87(1032), 670–679. https://doi.org/10.1136/pgmj.2010.116012
dc.relation.references	Mikolov, T., Chen, K., Corrado, G., & Dean, J. (2013). Efficient Estimation of Word Representations in Vector Space. arXiv. https://doi.org/10.48550/arXiv.1301.3781
dc.relation.references	Ministerio de Salud y Protección Social. (1993). Ley 100 de 1993. Funcion publica. https://www.funcionpublica.gov.co/eva/gestornormativo/norma.php?i=5248
dc.relation.references	Ministerio de Salud y Protección Social. (2010). Seguridad del paciente y la atención segura. Ministerio de Salud y Protección Social. https://www.minsalud.gov.co/sites/rid/Lists/BibliotecaDigital/RIDE/DE/CA/Guia-buenas-practicas-seguridad-paciente.pdf
dc.relation.references	Ministerio de Salud y Protección Social. (2016). Sistema Obligatorio de Garantía de Calidad en Salud (SOGCS). Ministerio de Salud y Protección Social. https://www.minsalud.gov.co/salud/PServicios/Paginas/sistema-obligatorio-garantia-calidad-SOGC.aspx
dc.relation.references	Ministerio de Salud y Protección Social. (2018). Guía práctica del sistema obligatorio de garantía de la calidad en salud SOGCS. Gobernacion de Boyaca. https://boyaca.gov.co/SecSalud/images/Documentos/cartilla_didactica.pdf
dc.relation.references	Moutawwakil, I., & Pierrard, R. (2023). LLM-Perf Leaderboard. Hugging Face. https://huggingface.co/spaces/optimum/llm-perf-leaderboard
dc.relation.references	OpenAI, Achiam, J., Adler, S., Agarwal, S., & Ahmad, L. (2023). GPT-4 Technical Report. arXiv. https://doi.org/10.48550/arXiv.2303.08774
dc.relation.references	Ovadia, O., Brief, M., Mishaeli, M., & Elisha, O. (2024). Fine-Tuning or Retrieval? Comparing Knowledge Injection in LLMs. arXiv. https://doi.org/10.48550/arXiv.2312.05934
dc.relation.references	Rasmussen, J. N., Chong, A., & Alter, D. A. (2007). Relationship Between Adherence to Evidence-Based Pharmacotherapy and Long-term Mortality After Acute Myocardial Infarction. JAMA, 297(2), 177. https://doi.org/10.1001/jama.297.2.177
dc.relation.references	Roehrborn, C. G. (2005). Benign prostatic hyperplasia: an overview. Reviews in Urology, 7(9), 3–14. https://pubmed.ncbi.nlm.nih.gov/16985902/
dc.relation.references	Rudin, R. S., Fischer, S. H., Shi, Y., & Shekelle, P. (2019). Trends in the Use of Clinical Decision Support by Health System–Affiliated Ambulatory Clinics in the United States, 2014-2016. American Journal of Managed Care, 12(19), 4–10. https://researchgate.net/publication/338989890_Trends_in_the_Use_of_Clinical_Decision_Support_by_Health_System-Affiliated_Ambulatory_Clinics_in_the_United_States_2014-2016
dc.relation.references	Sinha, R. K. (2010). Impact of Health Information Technology in Public Health. Sri Lanka Journal of Bio-Medical Informatics, 1(4), 223. https://doi.org/10.4038/sljbmi.v1i4.2239
dc.relation.references	Sutskever, I., Vinyals, O., & Le, Q. V. (2014). Sequence to Sequence Learning with Neural Networks. arXiv. https://doi.org/10.48550/arXiv.1409.3215
dc.relation.references	Tang, R., Han, X., Jiang, X., & Hu, X. (2023). Does synthetic data generation of llms help clinical text mining? arXiv. https://doi.org/10.48550/arXiv.2303.04360
dc.relation.references	Tayefi, M., Ngo, P., Chomutare, T., Dalianis, H., Salvi, E., Budrionis, A., & Godtliebsen, F. (2021). Challenges and opportunities beyond structured data in analysis of electronic health records. WIREs Computational Statistics, 13(6). https://doi.org/10.1002/wics.1549
dc.relation.references	Touvron, H., Martin, L., Stone, K., Albert, P., & Almahairi, A. (2023). Llama 2: Open Foundation and Fine-Tuned Chat Models. arXiv. https://doi.org/10.48550/arXiv.2307.09288
dc.relation.references	Tunstall, L., Werra, L. V., & Wolf, T. (2022). Question Answering. In Natural language processing with transformers: building language applications with Hugging Face (p. 1). O’Reilly Media. https://www.oreilly.com/library/view/natural-language-processing/9781098136789/
dc.relation.references	Unstructured Technologies Inc. (2024). unstructured [Computer software]. Unstructured. https://github.com/Unstructured-IO/unstructured
dc.relation.references	Vajjala, S., Majumder, B., Gupta, A., & Surana, H. (2020). NLP: A Primer. In Practical Natural Language Processing (p. 1). O’Reilly Media, Inc. https://www.oreilly.com/library/view/practical-natural-language/9781492054047/
dc.relation.references	Vasilakes, J., Zhou, S., & Zhang, R. (2021). Natural language processing. In Machine Learning in Cardiovascular Medicine (pp. 123–148). Elsevier. https://doi.org/10.1016/B978-0-12-820273-9.00006-3
dc.relation.references	Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, L., & Polosukhin, I. (2017). Attention is All you Need. arXiv. https://doi.org/10.48550/arXiv.1706.03762
dc.relation.references	Verhamme, K. M. C., Dieleman, J. P., Bleumink, G. S., van der Lei, J., & Sturkenboom, M. C. J. M. (2002). Incidence and Prevalence of Lower Urinary Tract Symptoms Suggestive of Benign Prostatic Hyperplasia in Primary Care—The Triumph Project. European Urology, 42(4), 323–328. https://doi.org/10.1016/S0302-2838(02)00354-8
dc.relation.references	Vuichoud, C., & Loughlin, K. R. (2015). Benign prostatic hyperplasia: epidemiology, economics and evaluation. The Canadian Journal of Urology, 22(1), 1–6. https://pubmed.ncbi.nlm.nih.gov/26497338/
dc.relation.references	Wei, J., Wang, X., Schuurmans, D., Bosma, M., & Ichter, B. (2023). Chain-of-Thought Prompting Elicits Reasoning in Large Language Models. arXiv. https://doi.org/10.48550/arXiv.2201.11903
dc.relation.references	Wolf, T., Debut, L., Sanh, V., Chaumond, J., & Delangue, C. (2020). Transformers: State-of-the-Art Natural Language Processing [Computer software]. Hugging Face. https://huggingface.co/docs/transformers/index
dc.relation.references	Yang, S., Zhu, F., Ling, X., Liu, Q., & Zhao, P. (2021). Intelligent Health Care: Applications of Deep Learning in Computational Medicine. Frontiers in Genetics, 12. https://doi.org/10.3389/fgene.2021.607471
dc.rights	Atribución-NoComercial-CompartirIgual 4.0 Internacional	en
dc.rights.accessrights	http://purl.org/coar/access_right/c_abf2
dc.rights.accessrights	info:eu-repo/semantics/openAccess
dc.rights.local	Acceso abierto
dc.rights.uri	http://creativecommons.org/licenses/by-nc-sa/4.0/
dc.subject	Procesamiento de lenguaje natural
dc.subject	Guías de práctica clínica
dc.subject	Modelo grande de lenguaje
dc.subject	Recuperación generación aumentada
dc.subject	Registros clínicos electrónicos
dc.subject	NLP
dc.subject	GPC
dc.subject	LLM
dc.subject	RAG
dc.subject	EHR
dc.subject.keywords	Natural Language Processing
dc.subject.keywords	Clinical Practice Guidelines
dc.subject.keywords	Large Language Model
dc.subject.keywords	Retrieval Augmented Generation
dc.subject.keywords	Electronic Health Records
dc.subject.keywords	NLP
dc.subject.keywords	CPG
dc.subject.keywords	LLM
dc.subject.keywords	RAG
dc.subject.keywords	EHR
dc.title	Diseño y desarrollo de una herramienta basada en procesamiento de lenguaje natural mediado por Deep Learning para evaluación de registros clínicos electrónicos
dc.title.translated	Design and development of a tool based on Natural Language Processing mediated by Deep Learning for the evaluation of electronic health records