Knife-Edge Scanning Microscopy: High-Throughput Imaging and Analysis of Massive Volumes of Biological Microstructures - High-Throughput Image Reconstruction and Analysis - page 31

Figure 2.16 Interactive visualization of fibrous and thread-like data. (a) Self-orienting surfaces im-
plemented on the GPU give high-quality renderings of neuron data viewed close up, and (b) allow
large volumes to be viewed at interactive rates. (c) Orientation filtering allows us to interactively ex-
plore orientation of fibers, allowing us (d) to discover fiber bundle directions across larger datasets.
(Adapted from [33].)
2.6 Discussion
The main contribution of our work is three-fold: (1) high-throughput serial-
sectioning imaging of brain microstructures, (2) fast and accurate algorithms for
2D and 3D reconstruction of neural and vascular morphology, and (3) interac-
tive real-time visualization of the reconstructed structures. These techniques are
expected to provide critical neuroanatomical data at a submicron-level for whole
small animal brains. There are several open issues that need to be addressed. In
the following sections, we will discuss validation and editing, as well as how to
exploit parallelism.