pixel normalization

[Max-Kharitonov]

Context

Describe the Bug

In case if we are using CPU rendering it works fine, but in case GPU this image inverted.
Please see attached anonimized file
DICOM sample Issue with Invertion.zip

Steps to Reproduce

run GPU rendering

run CPU rendering

The current behavior

Image inverted and not rendered in proper way

The expected behavior

Image should be rendered in the correct way and the result should not depend on the rendering type.

OS

Windows 10, Mac os

Node version

v18.14.1

Browser

Chrome 120.0.6099.131

Changes & Results

Such images have a pixel range of [negative, negative]. Therefore, we are modifying the pixels per LUT to normal values [0, 255]. After these manipulations, everything renders fine.

Testing

Try to open stackViewport with this series using GPU rendering:
DICOM.sample.Issue.with.Invertion (1).zip

[netlify]

✅ Deploy Preview for cornerstone-3d-docs ready!

Name	Link
🔨 Latest commit	9a9c5ba
🔍 Latest deploy log	https://app.netlify.com/sites/cornerstone-3d-docs/deploys/65e194fd5cb3e00008e6e9ed
😎 Deploy Preview	https://deploy-preview-1135--cornerstone-3d-docs.netlify.app
📱 Preview on mobile	Toggle QR Code... Use your smartphone camera to open QR code link.

---

To edit notification comments on pull requests, go to your Netlify site configuration.

[sedghi]

can you please provide context, what was the issue and how are you solving it with your PR? basically please fill in the template for the PR and provide test steps

[Max-Kharitonov]

@sedghi
this PR is related to #985. Such images have a pixel range of [negative, negative]. Therefore, we are modifying the pixels per LUT to normal values [0, 255]. After these manipulations, everything renders fine.

You could find test series in original report: #985

[sedghi]

Thanks for your pull request. I apologize for the delay; I was away, but I'm now back to contributing to open source more actively.

I don't think your approach is correct. The minimum value can be negative, and there's no issue with that. The problem lies in the upper and lower voi ranges being the same, preventing setVOI from grabbing the correct range from the data

I'm willing to merge your pull request if you implement it like this.

[Max-Kharitonov]

Thanks for your pull request. I apologize for the delay; I was away, but I'm now back to contributing to open source more actively.

I don't think your approach is correct. The minimum value can be negative, and there's no issue with that. The problem lies in the upper and lower voi ranges being the same, preventing setVOI from grabbing the correct range from the data

I'm willing to merge your pull request if you implement it like this.

In case if min and max values negative? I mean all pixels are negative

[s2ramana]

@sedghi any update on this?

[sedghi]

I'm not sure if this is the right direction