[IGSTK-Users] Problem reading DICOM CT images

[Torleif Sandnes]

Hi Luis.
>
>      Why is Osirix tagging the modality as "SC"  ?
>
>   Is it because the gray levels have been modified ?
No, Osirix sets the modality to SC even if only the patient data has  
been anonymized, but in my case it was because i scaled down the series.

>
>   If the intensity range has been modified, then the "SC"
>   modality is actually warning you that you cannot use
>   this image as a CT. For one thing, you will not be able
>   rely on the range of Hounsfield units for the purpose of
>   applying window levels, or for the purpose of initializing
>   segmentations. It will even compromise the potential of
>   performing registration against other real CT images, if
>   the metrics rely on the pixels intensities to be in the
>   same range.
Ok, that seems reasonable. This is definitely a good reason to reject  
SC modalities.

> As we discussed several days ago, we are going to add a
> "modality-oblivious" layer to IGSTK so that you can use
> those readers and images for less-safe, more-generic
> applications, that are willing to accept "any-modality"
> as a valid one. It is then up to the application developer
> to choose between the classes that provide flexibility at
> the price of lower safety, and the classes that improve
> safety by restricting functionalities.

That is great news for me!

Thanks,
Torleif