Attention-based CNN-BiLSTM for sleep state classification of spatiotemporal wide-field calcium imaging data
Attention-based CNN-BiLSTM for sleep state classification of spatiotemporal wide-field calcium imaging data
复制标题
DOI：
10.1016/j.jneumeth.2024.110250
复制
发表时间：
2024-08-16
影响因子：
3
通讯作者：
Anastasio，Mark A.
中科院分区：
文献类型：
--
作者：
Zhang，Xiaohui;Landsness，Eric C.;Anastasio，Mark A.
背景具有基因编码钙指示剂的宽场钙成像（WFCI）可以对小鼠神经元活动进行时空记录。当应用于睡眠研究时，WFCI 数据通过使用辅助脑电图和肌电图记录手动评分为清醒、非快速眼动 (NREM) 和快速眼动睡眠状态。然而，这个过程非常耗时、具有侵入性，并且常常受到评估者间和评估者内部可靠性较低的影响。因此，需要一种针对时空 WFCI 数据进行自动睡眠状态分类的方法。新方法提出了一种由卷积神经网络（CNN）提取图像帧的空间特征和具有注意机制的双向长短期记忆网络（BiLSTM）组成的混合网络架构，用于识别不同时间点之间的时间依赖关系，将 WFCI 数据分类为清醒、NREM 和 REM 睡眠状态。结果睡眠状态分类的准确度为 84%，Cohen’sκ 为0.64。梯度加权类别激活图显示，在将 WFCI 数据分类为 NREM 睡眠时，皮质的额叶区域更为重要，而后部区域对清醒的识别贡献最大。注意力分数表明，所提出的网络以特定于状态的方式关注短期和长期时间依赖性。与现有方法相比，在持续重复的 3 小时 WFCI 记录上，CNN-BiLSTM 达到了 0.67 的 aκ，与对应于人类 EEG/EMG 评分的 0.65 的 aκ 相当。结论 CNN-BiLSTM 有效地从时空 WFCI 数据中对睡眠状态进行分类，并将实现更广泛的应用WFCI 睡眠研究。
BackgroundWide-field calcium imaging (WFCI) with genetically encoded calcium indicators allows for spatiotemporal recordings of neuronal activity in mice. When applied to the study of sleep, WFCI data are manually scored into the sleep states of wakefulness, non-REM (NREM) and REM by use of adjunct EEG and EMG recordings. However, this process is time-consuming, invasive and often suffers from low inter- and intra-rater reliability. Therefore, an automated sleep state classification method that operates on spatiotemporal WFCI data is desired.New methodA hybrid network architecture consisting of a convolutional neural network (CNN) to extract spatial features of image frames and a bidirectional long short-term memory network (BiLSTM) with attention mechanism to identify temporal dependencies among different time points was proposed to classify WFCI data into states of wakefulness, NREM and REM sleep.ResultsSleep states were classified with an accuracy of 84 % and Cohen’sκof 0.64. Gradient-weighted class activation maps revealed that the frontal region of the cortex carries more importance when classifying WFCI data into NREM sleep while posterior area contributes most to the identification of wakefulness. The attention scores indicated that the proposed network focuses on short- and long-range temporal dependency in a state-specific manner.Comparison with existing methodOn a held out, repeated 3-hour WFCI recording, the CNN-BiLSTM achieved aκof 0.67, comparable to aκof 0.65 corresponding to the human EEG/EMG-based scoring.ConclusionsThe CNN-BiLSTM effectively classifies sleep states from spatiotemporal WFCI data and will enable broader application of WFCI in sleep research.