机器学习预测使用Candida评分的儿科真菌病死亡率
Khouloud Abdulrahman Al-Sofyani1,2,3, Ibrahim Hussain Ali Muzaffar1,2, Abdulrahman Mohammedsaeed Baqasi1,2
1Department of Pediatrics, Faculty of Medicine, King Abdulaziz University, Jeddah, Saudi Arabia.
Scientific reports
|November 13, 2025
概括
儿童真菌病死亡率可以使用Candida评分和临床因素来预测. 机器学习模型,特别是随机森林,在这个PICU研究中显示出有前途的歧视.
科学领域:
- 关键护理医学 关键护理医学
- 传染性疾病 传染性疾病
- 儿科 儿科 儿科
背景情况:
- 儿科真菌病是一种儿童血液中的真菌感染,与高死亡率有关,特别是在儿科重症监护病房 (PICU).
- 准确预测死亡风险对于及时有效的临床治疗儿科真菌病至关重要.
研究的目的:
- 评估Candida评分与儿科真菌病死亡率的临床变量结合的预测性表现.
- 将多变量逻辑回归模型与机器学习算法 (随机森林,梯度增强机器) 的区分进行比较.
主要方法:
- 从单个 PICU (2016-2020) 中对85例儿科真菌病例进行了回顾性分析.
- 一个预规定的多变量逻辑回归模型被用作主要分析.
- 随机森林和梯度增强机被用作探索性比较模型.
- 模型歧视是使用持有测试集进行评估的,并通过十倍交叉验证和引导重新抽样来验证.
主要成果:
- 该研究包括85名儿科患者,平均年龄为6个月;45.9%的患者死亡.
- 在测试组中,逻辑回归实现了0.800.00的AUC.
- 随机森林显示了最高的歧视,AUC为0.861,其次是梯度增强 (AUC为0.847).
结论:
- 将Candida评分与临床预测指标相结合,显示了在儿科真菌病中分层死亡风险的潜力.
- 在这种情况下,机器学习模型,特别是随机森林模型,与传统的物流回归相比,可能会提供更好的歧视.
- 这些发现是探索性的,在临床实施之前需要在更大的多中心队列中进行外部验证.
相关概念视频
01:24Automated Microbial Diagnostics
Automated diagnostic analyzers have transformed clinical microbiology by providing rapid and reliable methods for pathogen identification and antibiotic susceptibility testing. Among these systems, the Vitek 2 is widely used because it automates the traditionally labor-intensive processes of microbial identification (ID) and antibiotic susceptibility testing (AST), delivering standardized and timely results that are essential for effective patient care.Microbial Identification with ID CardsThe...
01:20Candidiasis
Candidiasis is a fungal infection caused by opportunistic species of Candida. It can affect various anatomical sites, including the skin, oral cavity, nails, and genitourinary tract. Among its forms, vaginal candidiasis is the most common type of mucosal infection. It typically results from the overgrowth of Candida albicans in the vaginal mucosa. Under normal conditions, C. albicans exists as a commensal organism within the vaginal microbiota, regulated by the dominance of lactobacilli, which...