The Chrono-Fourier AI Dataset: A Chronobiological Immune Time-Series Resource for Spectral Modeling of NLR/PLR/SII During Immune Checkpoint Therapy

The Chrono-Fourier AI Dataset is the first open-access chronobiological immune time-series dataset derived from real-world clinical data of patients treated with immune checkpoint inhibitors. It contains raw anonymized longitudinal laboratory measurements—including neutrophils, lymphocytes, platelets, and derived immune indices such as NLR, PLR, and SII—collected across sequential treatment cycles under PD-1, PD-L1, and CTLA-4 blockade. This dataset enables investigation of circadian timing effects, temporal immune dynamics, and immune coherence patterns in relation to therapy administration time (morning vs afternoon), clinical response, toxicity, and survival outcomes. It is specifically structured for Fourier-based spectral modeling and AI-driven time-series analysis, supporting machine learning exploration of rhythm-dependent immunologic behavior. The dataset is provided in CSV format and adheres to FAIR data principles for reproducibility, transparency, and real-world evidence development. It is compatible with Python, R, MATLAB, and modern deep learning frameworks, and is intended to support translational oncology research and future prospective chronotherapy trial design.时序傅里叶人工智能数据集（Chrono-Fourier AI Dataset）是首个开放获取的时间生物学免疫时序数据集，其数据来源于接受免疫检查点抑制剂治疗的患者的真实世界临床数据。该数据集包含经匿名化处理的纵向实验室原始检测数据，涵盖中性粒细胞、淋巴细胞、血小板等血常规指标，以及中性粒细胞与淋巴细胞比值（NLR）、血小板与淋巴细胞比值（PLR）、系统免疫炎症指数（SII）等衍生免疫指标，采集自接受PD-1、PD-L1及CTLA-4阻断治疗的患者的连续治疗周期。 本数据集可支持围绕给药时间（上午与下午）、临床应答、毒性反应及生存结局，开展昼夜节律时间效应、时序免疫动态变化及免疫协同模式相关研究。 该数据集专门针对基于傅里叶变换的光谱建模与人工智能驱动的时序分析进行结构优化，可支持机器学习领域对节律依赖性免疫行为的探索研究。 本数据集以CSV格式提供，遵循FAIR数据原则，以保障研究可重复性、数据透明度及真实世界证据开发。其兼容Python、R、MATLAB及主流深度学习框架，旨在为转化肿瘤学研究及未来前瞻性时序治疗试验设计提供支持。