Login
Search
Search
0 Dates
2026
2025
2024
2023
2022
2021
2020
2019
2018
0 Events
CPC 2018
CPC 2019
Curso de Atualização em Medicina Cardiovascular 2019
Reunião Anual Conjunta dos Grupos de Estudo de Cirurgia Cardíaca, Doenças Valvulares e Ecocardiografia da SPC
CPC 2020
CPC 2021
CPC 2022
CPC 2023
CPC 2024
CPC 2025
CPC 2026
0 Topics
A. Basics
B. Imaging
C. Arrhythmias and Device Therapy
D. Heart Failure
E. Coronary Artery Disease, Acute Coronary Syndromes, Acute Cardiac Care
F. Valvular, Myocardial, Pericardial, Pulmonary, Congenital Heart Disease
G. Aortic Disease, Peripheral Vascular Disease, Stroke
H. Interventional Cardiology and Cardiovascular Surgery
I. Hypertension
J. Preventive Cardiology
K. Cardiovascular Disease In Special Populations
L. Cardiovascular Pharmacology
M. Cardiovascular Nursing
N. E-Cardiology / Digital Health, Public Health, Health Economics, Research Methodology
O. Basic Science
P. Other
0 Themes
01. History of Cardiology
02. Clinical Skills
03. Imaging
04. Arrhythmias, General
05. Atrial Fibrillation
06. Supraventricular Tachycardia (non-AF)
07. Syncope and Bradycardia
08. Ventricular Arrhythmias and Sudden Cardiac Death (SCD)
09. Device Therapy
10. Chronic Heart Failure
11. Acute Heart Failure
12. Coronary Artery Disease (Chronic)
13. Acute Coronary Syndromes
14. Acute Cardiac Care
15. Valvular Heart Disease
16. Infective Endocarditis
17. Myocardial Disease
18. Pericardial Disease
19. Tumors of the Heart
20. Congenital Heart Disease and Pediatric Cardiology
21. Pulmonary Circulation, Pulmonary Embolism, Right Heart Failure
22. Aortic Disease
23. Peripheral Vascular and Cerebrovascular Disease
24. Stroke
25. Interventional Cardiology
26. Cardiovascular Surgery
27. Hypertension
28. Risk Factors and Prevention
29. Rehabilitation and Sports Cardiology
30. Cardiovascular Disease in Special Populations
31. Pharmacology and Pharmacotherapy
32. Cardiovascular Nursing
33. e-Cardiology / Digital Health
34. Public Health and Health Economics
35. Research Methodology
36. Basic Science
37. Miscellanea
0 Resources
Abstract
Slides
Vídeo
Report
CLEAR FILTERS
From Free Text to Insight: Performance of LLMs in Structuring Portuguese Angiography Reports and Extracting FFR/iFR Values
Session:
Sessão de Posters 47 - Dos dados às decisões: a revolução da IA em cardiologia
Speaker:
Diogo Rosa Ferreira
Congress:
CPC 2026
Topic:
N. E-Cardiology / Digital Health, Public Health, Health Economics, Research Methodology
Theme:
33. e-Cardiology / Digital Health
Subtheme:
33.4 Digital Health
Session Type:
Posters Eletrónicos
FP Number:
---
Authors:
Diogo Ferreira; João Cravo; Marta Vilela; Daniel Cazeiro; Sofia Morgado; Filipa Valdeira; Cláudia Soares; Fausto Pinto; Miguel Nobre Menezes
Abstract
<p style="text-align:justify"><span style="font-size:11pt"><span style="font-family:Arial,sans-serif"><strong>Introduction:</strong><br /> Coronary angiography and angioplasty reports contain essential anatomical, physiological, and procedural data for research, audits, and quality assessment. However, as unstructured text, extracting variables —such as lesion location, FFR/iFR values, and procedural details—is slow and hard to scale. This limits dataset size, delays research, and hinders quality monitoring. Although Large Language Models (LLMs) show promise in medical text processing, their performance on Portuguese cardiology reports is unknown. This study evaluates whether open-source, locally deployable LLMs can convert free-text angiography reports into structured data for research and quality improvement.</span></span></p> <p style="text-align:justify"><span style="font-size:11pt"><span style="font-family:Arial,sans-serif"><strong>Methods:</strong><br /> A total of 1,314 reports (2012–2023) were analyzed. Two tasks were assessed: extraction of FFR/iFR values per vessel and classification of procedure type. The reference standard was established through manual review by cardiology trainees. Llama3, Mistral and Medllama2 models were tested using zero-shot and one-shot prompting and compared with a REGEX baseline. Performance was measured using global concordance, sensitivity, positive predictive value (PPV) and F1-score, with strict exact-value matching for physiological indices.</span></span></p> <p style="text-align:justify"><span style="font-size:11pt"><span style="font-family:Arial,sans-serif"><strong>Results:</strong><br /> Of the 1,314 reports, 681 described angiography alone and 633 included angioplasty, totaling 591 FFR and 1,190 iFR values. MedLlama2 was excluded for inconsistent structured output. For procedure classification, REGEX achieved the highest global concordance (0.992) and PPV, while Llama3 one-shot showed the best sensitivity (0.998) and an F1-score of 0.984.<br /> For FFR/iFR detection , Llama3 zero-shot demonstrated the strongest performance, with global concordance of 0.948 and sensitivity of 0.872, surpassing both Mistral variants and REGEX.<br /> For exact numerical extraction, Llama3 one-shot reached the highest concordance (0.906). REGEX remained highly specific but less sensitive in detecting index presence, reflecting limitations with varied phrasing.</span></span></p> <p style="text-align:justify"><span style="font-size:11pt"><span style="font-family:Arial,sans-serif"><strong>Conclusion:</strong><br /> Open-source LLMs can accurately structure Portuguese coronary angiography reports, reducing the need for large-scale manual extraction. Llama3 models reliably detect FFR/iFR measurements and show high concordance in exact value extraction, outperforming REGEX and other open-weight models in linguistically complex tasks. This substantially decreases the number of reports requiring human review and enables dataset generation while preserving data privacy. Overall, LLM-based extraction emerges as a practical, deployable solution for research, quality monitoring, and audit, with clear potential to accelerate data-driven cardiovascular care.</span></span></p>
Slides
Our mission: To reduce the burden of cardiovascular disease
Visit our site