News & Events

Volume Correlative Light and Electron Microscopy (vCLEM) is a powerful technique for examining the fine structures within cells and tissues, which allows scientist to focus their attention to molecularly defined areas. A significant hurdle in vCLEM has been precisely locating and tracking "Regions of Interest" (ROIs) across different imaging types (multimodal) and magnifications (multiscale).

To address this, Dr Kohki Konishi, a visiting scientist from Nikon Japan, working with Dr Guilherme Neves in the Burrone lab in close collaboration with Dr Matt Russell at the Centre for Ultrastructural Imaging developed and made freely available on GitHub two crucial tools:

1. SegReg: an image registration tool that aligns different image datasets, both 2D and 3D. It works by identifying and segmenting common features in the images and uses a Graphical Processing Unit (GPU) for fast and efficient processing.
2. NavROI: a specialized 3D image viewer designed to visualize how these multimodal images align after registration.

Our integrated workflow, just out in J Microscopy uses SegReg and NavROI and allows precise navigation within large mouse tissue samples where fluorescent signals are preserved. We demonstrate selective targeting for high-resolution Transmission Electron Microscopy (TEM) tomography of specific ROIs, such as inhibitory synapses or the cisternal organelle in a chosen neuron's axon.

This system provides real-time guidance for trimming samples in the X-Y plane, accurately estimates cutting depth (Z-dimension) relative to the ROI, and offers clear visual navigation of complex, multiscale image data. This significantly enhances the efficiency and accessibility of vCLEM analysis, overcoming a major bottleneck in correlating light microscopy data with detailed electron microscopy ultrastructure.

By Guilherme Neves