Multimodal imaging of memory in epilepsy from whole brain networks to local neuronal responses: Implications for surgical decision-making

前颞叶切除术（ATL）是消除癫痫发作的一种非常成功的治疗方法， 颞叶癫痫（TLE）。然而，ATL诱导的记忆力下降是频繁的，并且通常是严重的， 对生活质量和功能结果的有害影响。立体定向激光杏仁核切除术 （SLAH）已被引入作为一种微创替代方案，可以最大限度地减少记忆风险 下降然而，目前尚不清楚哪些患者将从SLAH中获益最多，以及SLAH是否 与ATL相比，降低了情景记忆衰退关键方面的风险。在上一次赠款期间， 在资助期间，我们证明了结合结构（sMRI），扩散- 加权（dMRI）和功能（fMRI）成像，以更好地表征神经网络的基础 术前语言和记忆障碍和（重组）组织和TLE。我们建议， 多模态成像（MMI）方法可用于量化术后记忆衰退的风险。在R 01中， 我们扩展了MMI方法，将sMRI/dMRI/fMRI与颅内记录（iEEG）相结合，使我们能够 深入研究TLE中情景记忆网络的空间和时间动态。我们采用多式联运 具有真实世界含义的关联学习任务（即，一张脸配上一个名字）以前没有被 在外科环境中研究。此外，我们还从认知神经科学的角度， 功能可能告知为什么许多患者难以在记忆中进行精细区分（即，受损 模式分离），即使简单的项目记忆似乎完好无损。我们建议，我们的MMI方法将 产生一个更完整的表征情节记忆网络在颞叶癫痫，揭示模式的结构和 个体患者的功能重组，并在以下情况下实现个性化的风险评估方法 考虑手术方案最后，我们将跟踪ATL和SLAH后的认知和成像变化， 促进重组和改善认知结果的患者特异性因素。这个的目标 更新完全符合2020年NINDS癫痫研究基准（第四部分）， 强调迫切需要进行研究，以限制或预防缉获及其 在整个生命周期中进行治疗。我们的更新直接解决了这一要求，努力改善手术 这将对患者护理产生直接和持续的影响。癫痫是一种常见 神经系统疾病每年花费医疗保健系统约155亿美元， 影响生活质量、就业和健康状况。目前的项目对公共卫生有很大的影响 因为它致力于通过使用先进的，非侵入性的， 识别记忆衰退的个体预测因子的技术，可以帮助指导手术决策， 可能降低与切除或消融功能脑区域相关的发病率。

Anterior temporal lobectomy (ATL) is a highly sucessful treatment for eliminating seizures in patients with temporal lobe epilepsy (TLE). However, ATL-induced memory decline is frequent and often severe, having a deleterious impact on quality of life and functional outcomes. Stereotactic laser amygdalohippocampectomy (SLAH) has been introduced as a minimally-invasive alternative that could minimize risk of memory decline. However, it is unclear which patients would benefit the most from SLAH and whether SLAH decreases risk for critical aspects of episodic memory decline compared to ATL. During the previous grant funding period, we demonstrated the clinical value of combining information from structural (sMRI), diffusion- weighted (dMRI), and functional (fMRI) imaging to better characterize the neural networks that underlie preoperative language and memory impairment and (re)organization and in TLE. We propose that the same multimodal imaging (MMI) approach can be used to quantify risk for postoperative memory decline. In this R01, we extend our MMI approach, combining sMRI/dMRI/fMRI with intracranial recordings (iEEG), enabling us to delve deep into the spatial and temporal dynamics of episodic memory networks in TLE. We employ multimodal associative learning tasks with real-world implications (i.e., pairing a face with a name) that have not before been studied in the surgical context. In addition, we draw from cognitive neurosicence models of hippocampal functioning that may inform why many patients struggle to make fine-grain distinctions in memory (i.e., impaired pattern separation), even when simple item memory appears intact. We propose that our MMI approach will yield a more complete characterization of episodic memory networks in TLE, reveal patterns of structural and functional reorganization in individual patients, and enable a personalized approach to risk assessment when considering surgical options. Finally, we will track cognitive and imaging changes post-ATL and SLAH and identify patient-specific factors that promote reorganization and improved cognitive outcomes. The goals of this renewal are perfectly aligned with the 2020 NINDS Benchmarks for Epilepsy Research (Part IV), which stress the critical need for reseach to limit or prevent adverse consequences of seizures and their treatments across the lifespan. Our renewal directly addresses this request, striving to improve surgical decision-making, which will have an immediate and sustained impact on patient care. Epilepsy is a common neurological disease that costs the healthcare system approximately $15.5 billion annually and can negatively impact quality of life, employment, and health status. The current project has strong implications for public health because it strives to improve health outcomes in patients with epilepsy by using advanced, noninvasive technology to identify individual predictors of memory decline that can help to guide surgical decisions and possibly reduce morbidity associated with removal or ablation of eloquent brain regions.