一种可解释的机器学习方法用于预测结核性硬化综合体儿童的耐药性
Jie Fu1,2, Genfu Zhang1,2, Zhixian Yang1,2
1Department of Pediatrics, Peking University People's Hospital, Beijing, China.
Frontiers in neurology
|August 20, 2025
概括
这项研究创建了一个机器学习模型,用于预测结核性硬化综合体 (TSC) 的儿童的耐药性 (DRE). 现在可以更准确地对儿科TSC患者的DRE风险进行早期识别.
科学领域:
- 计算神经科学
- 儿童神经病学
- 医学中的机器学习
背景情况:
- 结核性硬化综合体 (TSC) 是一种与相关的遗传性疾病.
- 药物耐药性 (DRE) 在儿童TSC治疗中构成重大挑战.
- 在TSC中预测DRE风险的工具对于及时干预至关重要.
研究的目的:
- 开发和验证可解释的机器学习 (ML) 算法,用于预测TSC儿童的DRE风险.
- 提高儿童TSC中DRE预测模型的透明度和临床实用性.
主要方法:
- 从88名患有TSC相关的儿科患者中回顾性收集临床数据.
- 应用9个ML算法,包括随机森林 (RF),构建预测模型.
- 使用SHapley添加式解释 (SHAP) 进行模型解释性和特征重要性分析.
主要成果:
- 射频模型表现出优异的性能,AUC为0. 862和特异性为0. 930.
- 确定了DRE的主要预测因素:婴儿综合征 (IESS) 病史,多焦点EEG放电,多层皮质结核和多药性 (≥3个ASM).
- 通过十倍交叉验证和决策曲线分析 (DCA) 验证模型性能,证实临床实用性.
结论:
- 开发的基于射频的预测模型有助于早期识别DRE高风险的TSC儿童.
- 通过SHAP改进的模型可解释性为儿科TSC患者提供了个性化的治疗决策.
- 该工具支持临床医生管理TSC的DRE风险,改善患者的护理.
更多相关视频
09:49Using a Bipolar Electrode to Create a Temporal Lobe Epilepsy Mouse Model by Electrical Kindling of the Amygdala
Published on: June 29, 2022
2.6K
08:04Identification and Classification of Position-specific GABAA Receptor Subunit Missense Variants for Their Role In Hippocampal Pyramidal Neurons
Published on: June 6, 2025
489
相关概念视频
Epilepsy and Seizures: Overview
274
Epilepsy is a chronic neurological disease marked by recurrent, unpredictable seizures. These seizures are caused by abnormal electrical discharges in the brain, leading to behavior, sensation, or consciousness alterations. They can also cause transient impairment of awareness, interfering with daily activities.
Various factors can trigger epilepsy, including genetic factors, brain damage, metabolic causes, and unknown etiology. Diagnosis of epilepsy involves electroencephalography (EEG), which...
Various factors can trigger epilepsy, including genetic factors, brain damage, metabolic causes, and unknown etiology. Diagnosis of epilepsy involves electroencephalography (EEG), which...
274
Antiepileptic Drugs: Modulators of Neurotransmitter Release Mediated by SV2A Protein
439
Antiepileptic drugs, such as levetiracetam (Keppra) and brivaracetam (Briviact), have emerged as crucial tools in managing epilepsy. These medications exert their therapeutic effects by targeting the synaptic vesicle protein SV2A, a transmembrane glycoprotein primarily found in the brain.
SV2A is a transmembrane glycoprotein located predominantly in the brain, modulating the release of neurotransmitters for neuronal communication. Both levetiracetam and brivaracetam exhibit a high affinity for...
SV2A is a transmembrane glycoprotein located predominantly in the brain, modulating the release of neurotransmitters for neuronal communication. Both levetiracetam and brivaracetam exhibit a high affinity for...
439
Seizures: Classification
582
Epilepsy is primarily characterized by unpredictable seizures, either provoked by an identifiable factor, such as injury or illness, or unprovoked, occurring spontaneously without apparent cause.
Seizures are typically classified into two main categories: focal and generalized seizures.
Focal Seizures
Focal seizures originate from specific regions of the brain. These seizures are further sub-classified into two types:
Seizures are typically classified into two main categories: focal and generalized seizures.
Focal Seizures
Focal seizures originate from specific regions of the brain. These seizures are further sub-classified into two types:
582
Antiepileptic Drugs: GABAergic Pathway Potentiators
632
γ-aminobutyric acid or GABA, plays a pivotal role as an inhibitory neurotransmitter in the brain. GABA pathway potentiators, also known as GABAergic drugs, are a class of pharmaceutical agents designed to enhance the functioning of the GABAergic system. These medications primarily treat epilepsy, a neurological disorder characterized by recurrent seizures.
The key GABA pathway potentiators used in epilepsy management are as follows.
Benzodiazepines are a well-known class of drugs used for...
The key GABA pathway potentiators used in epilepsy management are as follows.
Benzodiazepines are a well-known class of drugs used for...
632
