[Bug] LUT/Image display issues

[rikous91]

Describe the Bug

There seems to be an issue with the LUT or the decoder for the attached study. The first images do not load correctly, but the problem seems to resolve itself when scrolling, and the first images then load correctly.

Steps to Reproduce

1. Download study
2. Load it here
3. Look at the first instances of the series.
4. Scroll through the series.
5. Look at the first instances again.

The current behavior

The expected behavior

OS

Windows 10

Node version

18.16.0

Browser

Chrome 131.0.6778.109

[sedghi]

I think our rendering is fine, at least consistent with other viewers

weasis

volview

Notice that if you scroll the lut is fine

[rikous91]

Sorry, try it here https://viewer-dev.ohif.org/local

At this point, the correct behavior can be seen here: https://v3-demo.ohif.org/local.

In the more recent beta versions, however, the images load correctly, but after scrolling, if you go back to the first ones, it seems to automatically apply a correction.

[sedghi]

The previous (v3-demo) did not handle voi correctly, voi should be applied per image and not global. The fact that volview and weasis shows the same rendering make me think it is your data isssue

[rikous91]

So, is it normal that at first the initial instances appear white, but then when you scroll and return to those same images, they’re no longer "white" and the LUT readjusts itself? I’m attaching a video to explain it better.

OHIF.Viewer.-.Google.Chrome.2025-01-23.09-12-36.online-video-cutter.com.mp4

[sedghi]

what i mean is this data is not normal, as other viewers also have trouble rendering it

[haldawe2]

AnonimizedXA.zip

This is a dicom from an XA that shows the exact same behaviour as descrived above, excepct that when scrolling to other frames the screen keeps white (the test case uploaded is only one frame, to simplify our troubleshooting).

This file loads fine in Weasis, MicroDicom and in https://v3-demo.ohif.org/local as well, while in 3.9.3 shows white as well.

Is voi handled by Cornerstone or OHIF in this case? Can't seem to find where it is set for each image.

[haldawe2]

Found out when the Window Width is 1 (which should be working according to the standard I think), it inverts the colors of the image, in our test case changing the value to 2 solves the issue.