Icosahedral Diffusion Tensor Encoding Scheme for Real-Time Brain Imaging

The diffusion encoding is segmented into multiple gradient lobes applied in different directions. Instead of using six or more complete diffusion-weighted acquisitions, the method divides the encoding into b1 and b2 lobes with multiple gradient directions (e.g., 6 directions with 3 lobes each), allowing partial information to be collected and combined to reconstruct diffusion parameters more efficiently

The method uses fewer complete diffusion-weighted acquisitions than traditional methods (e.g., 6 directions instead of 6+ directions with full encoding), but compensates by using multiple gradient lobes per direction. This partial action approach reduces scanning time while maintaining measurement precision through the combination of multiple partial encodings

Multiple gradient lobes with different orientations are merged into a single diffusion encoding sequence. The b1 and b2 lobes are combined such that their combined effect provides diffusion weighting information from multiple directions without requiring separate acquisitions for each direction, thus preserving information content while reducing scanning time

The diffusion encoding is performed continuously through multiple gradient lobes within a single acquisition sequence rather than requiring discrete separate acquisitions. This continuous action maintains the information content by collecting diffusion data from multiple orientations in an uninterrupted manner, reducing the time loss associated with repeated positioning and acquisition setup