The most important upgrade of version 10.1 is native built-in machine learning segmentation. Researchers can train an AI pixel classifier simply by brushing foreground and background voxels, which enables reliable segmentation of complex biological structures such as cell nuclei, neurons and microglia, much more flexible than simple threshold segmentation. It supports training on adjustable thick oblique slicers and fast preview of prediction results, and the trained AI model can be applied in batch processing for multiple large datasets.

Other key improvements:

1. 2D images will be opened in dedicated 2D working mode, with simplified 2D interactive controls.

2. Multi-thread acceleration for image processing, making full use of multi-core CPUs for fast computation.

3. It retains classic modules: Surfaces, Spots, Filament Tracer, deconvolution, batch processing, statistics export and high-quality volume rendering.

4. Imaris Arena is available for managing massive microscopy datasets up to terabyte scale.

This software supports common confocal, light-sheet and super-resolution microscopy formats, helping scientists visualize, segment, trace and quantify biological samples, and generate publication-ready images, animations and statistical results.