积极的深度学习促进住院患者的青素过敏延迟:一项实施研究
Melinda Jiang1,2, Brandon Stretton1,2, Joshua Kovoor2,3
1Royal Adelaide Hospital, Adelaide, South Australia, Australia.
International archives of allergy and immunology
|January 19, 2025
概括
深度学习显著改善了住院患者的青素过敏取消标签率,证明了安全和成本效益. 这种人工智能驱动的方法成功地识别了患者进行过敏测试,减少了错误的标签和节省医疗保健成本.
科学领域:
- 临床信息学 临床信息学
- 人工智能在医学中的应用
- 药物监督 药物监督 药物监督
背景情况:
- 错误的青素过敏标签会导致大量的医疗费用和不良事件.
- 准确的青素过敏评估对于适当的抗生素选择至关重要.
- 需要采取积极的策略来解决错误的过敏指标的流行问题.
研究的目的:
- 评估深度学习引导的积极主动咨询服务的有效性,用于住院患者的青素过敏脱标.
- 将通过人工智能驱动干预实现的删除标签率与历史数据进行比较.
- 评估实施的取消标签协议的安全和经济影响.
主要方法:
- 一项涉及深度学习算法的前性实施研究,用于识别潜在的青素过敏患者.
- 使用基于已建立的取消标签协议的积极咨询服务.
- 在一个干预组中的取消标签率与14周的历史对照组之间的比较.
主要成果:
- 深度学习算法确定了439名患者中的121名患者适合过敏询问.
- 在住院期间,成功的脱标发生在16.5%的确诊患者中,其中9.9%的患者被转诊进行门诊测试.
- 与历史对照组 (0%) 相比,干预组的取消标签率在统计学上显著增加,没有报告任何不良事件. 预计每年节省的资金超过117万澳元.
结论:
- 深度学习促进的主动性住院患者的青素过敏脱标是一种有效和安全的策略.
- 由人工智能驱动的方法通过减少错误的过敏标签,证明了显著的经济效益.
- 需要在不同的临床环境中进行进一步的研究,以验证这种创新的取消标签方法.
相关概念视频
02:06Allergic Reactions
26.4K
Overview
26.4K
01:30Allergic Reactions: Anaphylaxis
280
Anaphylaxis is a severe, life-threatening hypersensitivity reaction mediated by Immunoglobulin E (IgE) antibodies. When IgE binds to allergens, it triggers the release of mediators– histamine, leukotrienes, and prostaglandins from mast cells and basophils. These mediators cause vasodilation, edema, and inflammation, leading to various symptoms.The primary allergens causing anaphylaxis include food items (e.g., peanuts, shellfish), drugs (e.g., penicillin, asparaginase, corticotropin,...
280
01:27Production of Antibiotics
265
Penicillin, one of the earliest and most widely used antibiotics, is produced industrially by the filamentous fungus Penicillium chrysogenum. Large stirred-tank bioreactors ranging from tens to hundreds of thousands of liters maintain tightly controlled temperature, pH, and dissolved oxygen conditions to support fungal metabolism and maximize antibiotic yield. Penicillin is a secondary metabolite, synthesized primarily during the stationary growth phase, which requires a carefully managed...
265