View in #ohif on Slack
@Salim_Kanoun: hi there I’m experiencing different loading strategies either for one of the displayset (if only one study loading) or one of two studies loading, see videos
the progressive loading introduced https://github.com/OHIF/Viewers/pull/4276 seems to not apply for all loadings
#4276 feat: Add interleaved HTJ2K and volume progressive loading
What is the difference between volume progressive loading and nth ImageLoadStrategy defined in the hanging protocol ? are they 2 different concepts ?
Additional info, if I remove the ImageLoadStrategy from the HP, then i have the new progressive loading everywhere. This new loading shoudn’t be a new imageLoadStrategy among others ? (or an enhancement of the nth strategy)
@Bill_Wallace: The new progressive loading is built into CS3D and works cooperatively with the viewport in ways that can’t be used with imageLoadStrategy because much of the behaviour is below the level of a single image load.
Testing TMTV mode with a single study seems to load both series progressively, albeit the PT one loads far faster since it is much smaller.
The progressive loading should work with multiple studies across several MPR views, although it doesnt directly interleave between studies, but rather loads by priority in a way that is supposed to interleave correctly across studies. Still, with one study having a CT and the second a PT, the difference in study sizes will tend to load most of the PT study early on.
@Salim_Kanoun: I will check that, I think for this progressive loading you are going to need a “fully loaded” signal for the physician
the problem is see is that the images are displayed the physician doesn’t know if he is working on fully loaded image or partial rendering
that could be problematic in a clinical use
when we had an ordered loading you had either full image resolution or nothing
here i see the risk that you believe you can review the image but in fact you only have a partial rendering
when you say the progressive image loading is bellow the level of a single image load, is this apply for a non HTJ2K TS ?
For an HTKJ2K I undestand that you can load a slice and then increase resolution when data reading progress
but for other TS you need the full payload to render it so it there another difference with image load strategy in this case ?