在脊椎外科手术中评估多模态自然语言处理以确定术后安全指标
Kyle A Mani1, Anthony P Terraciano, Samuel N Goldman
1From the Department of Orthopaedic Surgery, Albert Einstein College of Medicine, Bronx, NY (Mani, Terraciano, Goldman, Bhatta, and Shankar), and the Department of Neurological Surgery (De La Garza Ramos) and the Department of Orthopaedic Surgery, Montefiore Medical Center, Bronx, NY (Fourman, Eleswarapu).
The Journal of the American Academy of Orthopaedic Surgeons
|September 2, 2025
概括
这项研究开发了一种多式机器学习 (ML) 模型,用于脊椎手术,结合结构化数据和外科医生笔记. 与仅使用结构化数据或文本的模型相比,组合模型显著改善了术后结果的预测.
科学领域:
- 脊椎手术
- 机器学习
- 数据科学
- 医疗信息学
背景情况:
- 标准的脊椎手术机器学习模型往往忽略了非结构化的自由文本临床叙述中的有价值信息,例如手术前的笔记.
- 这种限制阻碍了ML在预测患者结果方面的全部潜力.
研究的目的:
- 开发和评估一个多模式的ML模型,将结构化电子健康记录 (EHR) 数据与手术前笔记中的非结构化临床叙述集成在一起.
- 将这种多式模式的预测性能与仅使用结构化电子健康记录数据或仅使用NLP处理的文本的模型进行比较.
主要方法:
- 在评估了多个ML算法后,选择了XGBoost算法用于模型开发.
- 开发了三种模型:基于电子病历的结构化模型,基于NLP (手术前笔记) 的模型和综合多式模型.
- 自然语言处理 (NLP) 技术包括令牌化,源码化和词包向量化用于文本数据. 使用网格搜索和十倍交叉验证优化了超参数.
主要成果:
- 多模式模型在预测延长停留时间 (ROC-AUC: 0.908) 和非家庭出院 (ROC-AUC: 0.920) 方面表现出色.
- 只有NLP模型也显示出强大的预测能力,在两种结果上都超过了只有结构化的电子健康记录模型.
- 可解释的人工智能确定了关键的预测特征,包括体重指数,年龄,保险状况,并发症指数,种族,手术史和手术水平.
结论:
- 整合非结构化的外科医生笔记显著提高了脊椎手术中的ML模型的预测准确性.
- 自由文本临床叙述比传统的结构化EHR变量对外科术后的预测效益更大.
相关概念视频
01:28Cardiomyopathy VII: Pre and Post Operative Nursing Management
27
Patients with hypertrophic cardiomyopathy (HCM) and left ventricular outflow tract (LVOT) obstruction who remain symptomatic despite optimal medical therapy may undergo a septal myectomy (Morrow procedure). This procedure involves excising a portion of the hypertrophied septum below the aortic valve using a heart-lung machine to improve blood flow through the LVOT. Effective preoperative and postoperative nursing management ensures successful patient outcomes, minimizes complications, and...
27
01:23Peripheral Artery Disease V: Postoperative Nursing Management
24
During the postoperative period, it is crucial to focus on maintaining circulation, identifying and managing potential complications, and planning for discharge.Nursing AssessmentVital signs monitoring: Regularly monitor vital signs, including blood pressure, heart rate, respiratory rate, and temperature, to detect early signs of complications such as bleeding and infection.Circulation assessment: Monitor pulses, perform Doppler assessments, and check capillary refill, color, temperature, and...
24
01:22Aneurysm IV: Nursing Management
20
Vigilant monitoring for aneurysm rupture is essential for patients undergoing aortic surgery.Preoperative Nursing ManagementContinuously monitor the patient for manifestations of aneurysm rupture, such as pallor, weakness, tachycardia, hypotension, abdominal, back, groin, or periumbilical pain, changes in consciousness, and a pulsating abdominal mass. Regularly assess the patient's peripheral pulses.Instruct the patient to consume a clear liquid diet the day before surgery and administer...
20
01:11Local Anesthetics: Clinical Application as Spinal Anesthesia
772
Spinal anesthetics are given during lower abdomen and limb surgeries to block sensory and motor neurons. They are administered in the mid to low lumbar regions, primarily acting on the cauda equina's nerve roots. The blockade level depends on the local anesthetic (LA) concentration. Usually, low LA concentrations are sufficient to block sensory fibers, while only high LA concentrations block motor fibers. Other factors like injection volume and speed, the patient's posture, and the drug...
772