Cross-Modality Coherent Registration of Whole Mouse Brains

Abstract

Recent whole brain mapping projects are collecting large-scale 3D images using powerful and informative modalities, such as STPT, fMOST, VISoR, or MRI. Registration of these multi-dimensional whole-brain images onto a standard atlas is essential for characterizing neuron types and constructing brain wiring diagrams. However, cross-modality image registration is challenging due to intrinsic variations of brain anatomy and artifacts resulted from different sample preparation methods and imaging modalities. We introduced a cross-modality registration method, called mBrainAligner, which uses coherent landmark mapping as well as deep neural networks to align whole mouse brain images to the standard Allen Common Coordinate Framework atlas. We also built a single cell resolution atlas using the fMOST modality, and used our method to generate whole brain map of 3D full single neuron morphology and neuron cell types.

D-LMBmap: A fully automated deep learning pipeline for whole-brain profiling of neural circuitry

“…We evaluated both pipelines, Atlas-trained and Sample-trained. In comparison to three recent mouse brain region segmentation methods, SeBRe 19 , BIRDS 17 , and mBrainAligner 44 , we found that both D-LMBmap pipelines deliver superior results in all six brain regions (Fig. 3E and F).…”

“…Inaccurate registration can cause signals in these brain regions to be wrongly assigned to the ventricular system, leading to misinterpretations. To validate how well D-LMBmap addresses the challenge of registering brain regions adjacent to the ventricular system, we counted the falsely detected number of neurons in the lateral ventricles (LV), the third ventricle (3 rd V), and the cerebral aqueduct (AQ) after whole-brain registration by ClearMap 45 , mBrainAligner 44 and D-LMBmap (Fig. 4E).…”