• Loading very large 4d real value volumes:  was able to parallelize conversion of integer to pseudo-real value using DICOM tags Slope\Intersect.
•  Added appropriate progress message to this conversion.
• 4D PET volume (3.6 GB) is loaded in 67 sec on the 4 core laptop.

PET 4D: During the loading of the large 4D datasets, a substantial portion of the loading time is spent on calculating histogram for all timepoints and finding optimal settings for all timepoints.  Trying to find an acceptable solution to avoid these heavy calculations, which are already fully parallelized and not much chance of further speed-up.

PET loading:  Final item is obtaining correct timepoint informattion.  All ususal timing tags such as ACQUSITION_TIME are not useful.  So need to implement the logic for using the FRAMEE_REFERENCE_TIME (ms) tag

Cerebral hematoma (VA) grant:   Had a meeeting with PI Uzma Saddamani and Dr. Rusinek to discuss initial steps in implementing the algorithms for this grant

Automatic Build Test revealed a major memory leak when loading and displaying volume data, so this is in the process of being fixed.

Fixed the error of the uninitialized bounding rectangle in the Histogram visualization

Working on the problem displaying histogram on real-valued volumes such as Parametric maps and PET volumes.

Resolved the "artifical background" problem for determining Display Center\Window.
Solution was to see if "0" entries constitute an artificially  high proportion of all voxels. In that case they are excluded from the calculation.