Benchmarking Anisotropic vEM Evaluation for Axial Super-Resolution(BRAVE-ASR)

This dataset provides a benchmark for evaluating z-axis super-resolution methods in volume electron microscopy (Benchmarking Anisotropic vEM Evaluation for Axial Super-Resolution, BRAVE-ASR). It contains high-resolution and multi-thickness ground truth volumes of mouse brain tissue, enabling a direct comparison between simulated low-resolution data and physically acquired data. The data was acquired using a Thermo Fisher Plasma-FIB (pFIB) with a 5 nm XY pixel size and a 1 µs dwell time. The dataset includes four co-aligned primary volumes with physical slice thicknesses of 5 nm, 10 nm, 20 nm, and 40 nm, which have undergone SIFT-based affine registration and histogram normalization. A larger registered but without additional histogram normalization 5 nm raw volume is also included.本数据集为体积电子显微镜（volume electron microscopy, vEM）轴向超分辨率方法的性能评估提供了标准基准，其命名为BRAVE-ASR（全称为Benchmarking Anisotropic vEM Evaluation for Axial Super-Resolution）。数据集包含小鼠脑组织的高分辨率多厚度真值（ground truth）体数据，可实现模拟低分辨率数据与实际采集数据的直接对比。 该数据通过赛默飞世尔（Thermo Fisher）Plasma-FIB（pFIB）设备采集，其XY方向像素尺寸为5 nm，电子束驻留时间为1 µs。数据集包含四组共对准的主体积数据集，物理切片厚度分别为5 nm、10 nm、20 nm与40 nm，所有数据均经过基于尺度不变特征变换（Scale-Invariant Feature Transform，SIFT）的仿射配准与直方图归一化处理。此外，数据集还附带一组尺寸更大的已配准5 nm原始体数据，但未进行额外的直方图归一化处理。