# How to create SimpleITK Image from a dicom dataset
**URL:** https://discourse.itk.org/t/how-to-create-simpleitk-image-from-a-dicom-dataset/4692
**Category:** Beginner Questions
**Created:** [December 21, 2021, 3:30pm UTC](https://discourse.itk.org/t/how-to-create-simpleitk-image-from-a-dicom-dataset/4692 "2021-12-21T15:30:47Z")
**Posts on this page:** 5
**Page:** 1
### Author:  [@Ervis](https://discourse.itk.org/u/Ervis)
#### Post date: [December 21, 2021, 3:30pm UTC](https://discourse.itk.org/t/how-to-create-simpleitk-image-from-a-dicom-dataset/4692/1 "2021-12-21T15:30:47Z")
Hi
Is there a way to create/form/load a SimpleITK image / volume from a dicom dataset ?
Essentially I need to load the dicom data received from a dicom node post QueryRetrieve. But instead of storing the data on disk on temp files then reading them using ImageSeriesReader to create the same Image object from dicom dataset directly ? I also need to stay in c# / .NET framwork.
Thank you in advance for any advise
Ervis
---
### Author:  [@dchen](https://discourse.itk.org/u/dchen)
#### Post date: [December 21, 2021, 3:56pm UTC](https://discourse.itk.org/t/how-to-create-simpleitk-image-from-a-dicom-dataset/4692/2 "2021-12-21T15:56:22Z")
The following example shows how to create a SimpleITK Image from a buffer in C#. Assuming you have the DICOM data in memory, something like the example ought to do the trick.
[https://simpleitk.readthedocs.io/en/master/link\_CSharp\_docs.html](https://simpleitk.readthedocs.io/en/master/link_CSharp_docs.html)
---
### Author:  [@Ervis](https://discourse.itk.org/u/Ervis)
#### Post date: [December 21, 2021, 4:05pm UTC](https://discourse.itk.org/t/how-to-create-simpleitk-image-from-a-dicom-dataset/4692/3 "2021-12-21T16:05:38Z")
Thank you Dave
I am aware of that buffer load method.
The issue I have is that my memory is an actual dicom dataset ( essentially an entire dicom file with it’s metadata etc) also I have a list of those datasets (the equivalent of multiple dicom files/slices for the scan)
Right now I am writing those files on disk and then using ImageSeriesReader to construct the volume in the correct geometry metadata etc
I am trying to eliminate that step in the least error prone way if that makes any sense
---
### Author:  [@dchen](https://discourse.itk.org/u/dchen)
#### Post date: [December 21, 2021, 4:46pm UTC](https://discourse.itk.org/t/how-to-create-simpleitk-image-from-a-dicom-dataset/4692/4 "2021-12-21T16:46:09Z")
What library are you using to load the DICOM data set into memory? There ought to be a way to get the pixel data.
---
### Author:  [@Ervis](https://discourse.itk.org/u/Ervis)
#### Post date: [December 21, 2021, 5:06pm UTC](https://discourse.itk.org/t/how-to-create-simpleitk-image-from-a-dicom-dataset/4692/5 "2021-12-21T17:06:54Z")
I am using fo-dicom FellowOakDicom
simple example query the dicom server / association and write the images to disk
I want to eliminate the writing to disk part and pass the multiple dicom datasets to ITK and to go on my merry way
```auto
// now get all the images of a serie with cGet in the same association
client = DicomClientFactory.Create(_qrServerHost, _qrServerPort, false, _aet, _qrServerAET);
var cGetRequest = CreateCGetBySeriesUID(studyUID, serieUids.First());
client.OnCStoreRequest += (DicomCStoreRequest req) =>
{
Console.WriteLine(DateTime.Now.ToString() + " recived");
SaveImage(req.Dataset);
return Task.FromResult(new DicomCStoreResponse(req, DicomStatus.Success));
};
public static void SaveImage(DicomDataset dataset)
{
var studyUid = dataset.GetSingleValue(DicomTag.StudyInstanceUID).Trim();
var instUid = dataset.GetSingleValue(DicomTag.SOPInstanceUID).Trim();
var path = Path.GetFullPath(_storagePath);
path = Path.Combine(path, studyUid);
if (!Directory.Exists(path))
{
Directory.CreateDirectory(path);
}
path = Path.Combine(path, instUid) + ".dcm";
new DicomFile(dataset).Save(path);
}
```
In theory I can query each individual dataset for the tag containing the pixel data and use the buffer import to create individual ITK Images , but I believe that would be error prone and cumbersome adding all the metadata manually for each image.
I was hoping there was some class or interface similar to ImageSeriesReader that takes in a list of file names or directory but instead of flat files to use the dataset directly