数据的守护者:以巴西葡萄牙语为有效医疗记录匿名化的NER和LLM
Mauricio Schiezaro1, Guilherme Rosa1, Bruno Augusto Goulart Campos2
1Venturus - Innovation and Technology, Campinas, Brazil.
Frontiers in public health
|January 21, 2026
概括
这项研究介绍了AnonyMed-BR,这是一个用于匿名巴西葡萄牙医疗记录的新型数据集. 提取和生成模型都取得了高性能,证明了在医疗保健中保护隐私的NLP的可行性.
科学领域:
- 自然语言处理 (NLP) 是一种自然语言处理.
- 医疗信息学 医疗信息学
- 数据 隐私 数据 隐私 数据
背景情况:
- 患者隐私对于医学数据研究和NLP应用至关重要.
- 缺乏高质量的,匿名的巴西葡萄牙医疗数据集阻碍了进展.
研究的目的:
- 介绍AnonyMed-BR,这是巴西医疗记录的新数据集.
- 评估提取性和生成性匿名化策略.
- 比较匿名化性能,并评估合成数据的影响.
主要方法:
- 开发了AnonyMed-BR与真实和合成医疗记录,手动注释PII.
- 通过BERT和生成模型 (T5,GPT) 评估命名实体认可 (NER).
- 使用LLM-as-a-Judge框架进行质量,信息丢失和重新识别风险评估.
主要成果:
- 所有模型都实现了高匿名性能 (F1>0.90).
- 提取和生成方法有效地掩盖了敏感数据,同时保留了临床意义.
- 合成数据改善了模型的概括性;特定任务的微调优于域名预训练.
结论:
- AnonyMed-BR是第一个手动注释的数据集,用于巴西葡萄牙医疗文本匿名化.
- 在巴西医疗保健中建立了保护隐私的NLP基金会.
- 证明了可靠的匿名化系统的可行性,使新的下游应用程序成为可能.
相关概念视频
Data Reporting and Recording
5.4K
Reporting and recording are crucial in data documentation. The timely, thorough, and accurate documentation of facts is essential when recording patient data. Failure to record findings during an assessment or interpretation of a problem will result in loss of information and make the patient document unreliable. The reader is left with general impressions if the information is not specific. A recording is documenting data of the individual's health information in a traceable, secure, and...
5.4K
Inhaled Medications
761
Inhaled medications are crucial for managing chronic obstructive pulmonary disease (COPD) and asthma. They are essential for effective treatment and control, ensuring optimal respiratory health and well-being. Inhaled medication delivers drugs directly to the lungs, providing a rapid onset of action and reducing systemic side effects compared to oral or injectable medications. Three primary types of inhalation devices are used to administer these medications: nebulizers, metered-dose inhalers...
761
Heart Failure V: Medical Management
227
Medical Management of Acute Decompensated Heart Failure (ADHF)The primary goals of therapy for patients hospitalized with acute decompensated heart failure (ADHF) include:Relieving symptomsOptimizing volume statusSupporting oxygenation and ventilationMaintaining cardiac output (CO) and end-organ perfusionIdentifying and addressing the cause of ADHFPreventing complicationsProviding patient education on factors precipitating HF exacerbationPlanning for dischargeOngoing monitoring and assessment...
227
Endocarditis III: Medical Management
217
Infective endocarditis management involves a multifaceted approach encompassing infection prevention, lifestyle modifications, pharmacological therapy, and surgical management.Infection Prevention:Hand Hygiene: Thorough handwashing is crucial to prevent the spread of infection. Hand hygiene should be performed regularly, especially before and after using the restroom.Oral Hygiene: Good oral hygiene is essential. It includes brushing teeth immediately after waking up and before bed, flossing...
217
Myocarditis III: Medical Management
182
Myocarditis: Comprehensive Medical ManagementMyocarditis, the heart muscle inflammation, requires a comprehensive medical management strategy that addresses the underlying cause, provides supportive care, manages symptoms, and reduces cardiac workload.Infections and Autoimmune CausesAdminister appropriate antimicrobial therapy when an infectious agent causes myocarditis. For instance, penicillin treats infections caused by Group A Streptococcus. In cases where autoimmune processes are...
182
Pericarditis III: Medical Management
321
The primary objectives of managing pericarditis are to determine the underlying cause, provide effective therapy for treatment and symptom relief, and promptly detect signs and symptoms of cardiac tamponade. The following outlines the essential aspects of medical management for pericarditis:ObjectivesDetermine the Cause: Identifying the underlying cause of pericarditis is crucial for targeted treatment. Causes include viral infections, autoimmune diseases, post-cardiac injury syndrome, and...
321


