More often than not, DICOM files are loaded from a location on disk due to the size and complexity of the files. However, no environment and scenario is the same and we periodically get requests regarding how to load DICOM files from memory, more specifically a byte[] rather than a Stream. Due to the way that .NET garbage collection works we must use a .NET GCHandle object when passing an IntPtr to the overloaded DicomDataSet.Load function.

private byte[] loadedFile; // byte array for DICOM file from database, server, etc.
private DicomDataSet dicomDS;
private DicomElement element;

// Pin object so it is not garbage collected before work is done.
GCHandle gch = GCHandle.Alloc(loadedFile);

// Load DicomDS from byte[] using IntPtr of array.
dicomDS.Load(Marshal.UnsafeAddrOfPinnedArrayElement(loadedFile, 0),
   loadedFile.Length, DicomDataSetFlags.None);

// Unpin object.
gch.Free();

// Get image from DicomDS.
element = dicomDS.FindFirstElement(null, DicomTag.PixelData, true);
rasterImageViewer1.Image = dicomDS.GetImage(element, 0, 0,
   Leadtools.RasterByteOrder.Gray, DicomGetImageFlags.None);

To download the entire C# example project for versions 17.5 and 18, check out this forum post.

This week I'd like to highlight another example project one of our support agents developed in response to a customer request. Group annotations are a necessary feature in any application utilizing annotations and markup. In fact, the LEADTOOLS HTML5 SDK includes support for group annotations out of the box by simply clicking and dragging with the select tool. However, this customer had a requirement to create and edit group annotations programmatically.

Here is how you create the annotations and then place them in groups. After that you can use the mouse to select, move and resize each group as if they were a single entity.

Most image viewing applications and controls are designed for viewing a single image at a time. However some images and documents, such as TIFF and PDF, have multiple pages. Since these documents are typically full of text, it is very natural to keep reading on to the next page without the need to press a button to load the next page. We have received several requests for this feature and one of our developer support agents has created an demo showing how to implement continuous multipage scrolling with our HTML5 viewer and RESTful Web Services.

The basic idea works similarly to an image slider or carousel you might see on a homepage. There is an outer container which, when scrolled to a specified location, will asynchronously create an additional HTML5 Viewer for the next page and allow you to keep scrolling until you have reached the end of the document. Below is the JavaScript snippet which modifies the DOM with jQuery and calls the LEADTOOLS Raster Web service to retrieve the image.