Gadwal Aishwarya, Cerdas Maria Gabriela, Khan Areej, Khan Zainab, Martina Di Carluccio, Nimmagadda Sai Priya, Ali Rizvi Syed Mustafa, Aljallawi Sara Mahmood, Jaryal Shelly, Tasgaonkar Nuren, Ahmed Adnan, Busireddy Sowmika Reddy
Radiology, Kasturba Medical College, Mangalore, IND.
General Medicine, Universidad de Ciencias Médicas (UCIMED), San José, CRI.
Cureus. 2025 Feb 21;17(2):e79409. doi: 10.7759/cureus.79409. eCollection 2025 Feb.
Non-convulsive seizures (NCS) are often underdiagnosed due to their subtle presentation including changes in behavior and mental status. Although electroencephalography (EEG) remains the gold standard for detection, challenges, such as subjective interpretation, individual observer variability, and limited availability, often prolong diagnosis. This can lead to severe complications, including cognitive decline and higher mortality rates. Recent developments in artificial intelligence (AI) are revolutionizing epilepsy care by providing enhanced accuracy and efficiency for diagnosing and managing NCS. Machine learning models, including convolutional neural networks (CNN), recurrent neural networks (RNN), and support vector machines (SVM) have demonstrated high precision in analyzing EEG data and predicting seizures. Innovations such as Ceribell Clarity algorithm (Ceribell, Sunnyvale, CA) allow fast, real-time seizure detection, reducing diagnostic delays in emergency and critical care. Wearable AI-driven technologies like wearable monitoring devices, predictive analytics, and explainable AI enhance personalized care and support better clinician decision-making. This review underlines AI's potential in neurology and neurosurgery, highlighting its role in enhancing diagnostic precision, accelerating interventions, and supporting surgical and treatment planning. By incorporating AI into clinical practice, healthcare systems can overcome diagnostic challenges and deliver patient-centered care. AI is becoming a key element in the future of medicine, driving advances in precision neurology and improving patient outcomes worldwide.
非惊厥性癫痫发作（NCS）因其临床表现隐匿，包括行为和精神状态的改变，常常未得到充分诊断。尽管脑电图（EEG）仍是检测的金标准，但诸如主观解读、个体观察者差异以及可用性有限等挑战，常常会延长诊断时间。这可能导致严重并发症，包括认知能力下降和更高的死亡率。人工智能（AI）的最新发展正在彻底改变癫痫护理，为非惊厥性癫痫发作的诊断和管理提供更高的准确性和效率。机器学习模型，包括卷积神经网络（CNN）、循环神经网络（RNN）和支持向量机（SVM），在分析脑电图数据和预测癫痫发作方面已显示出高精度。诸如Ceribell Clarity算法（Ceribell，加利福尼亚州桑尼维尔）等创新技术可实现快速、实时的癫痫发作检测，减少急诊和重症监护中的诊断延迟。可穿戴式人工智能驱动技术，如可穿戴监测设备、预测分析和可解释人工智能，可增强个性化护理并支持临床医生做出更好的决策。本综述强调了人工智能在神经病学和神经外科中的潜力，突出了其在提高诊断精度、加速干预以及支持手术和治疗规划方面的作用。通过将人工智能纳入临床实践，医疗保健系统可以克服诊断挑战并提供以患者为中心的护理。人工智能正成为未来医学的关键要素，推动精准神经病学的发展并改善全球患者的治疗效果。
