Monday, April 30, 2012

DICOM Modality Worklist

Modality worklist (MWL) is one of DICOM’s workflow services that really make a difference. It’s the difference between grocery store workflow with notes on little pieces of paper and a true modern accountable workflow.

Technically speaking, DICOM Modality Worklist is a task manager just like a piece of paper with short text and a check box or the tasks application on your iPhone (or Android). But for the imaging center or RAD department the advantages are enormous. The most obvious benefit is that there’s no need to reconcile all kind of misspelled names in the PACS because the patient name is no longer keyed in on the modality workstation but received electronically via the MWL query. The fact that the requested procedure is also received electronically reduces the chance for doing the wrong procedure to the wrong patient. Combined with Modality Performed Procedure step (MPPS), that allows the modality to report the task status, take ownership over the task and checkmark it as done when completed, the up side is obvious. No wonder then, that many HMO’s require Modality Worklist as a mandatory feature for every imaging device they purchase. 

The most basic abstraction of a task is a short description of what should be done and a checkbox. That’s all it takes. The MWL data model is a bit more complicated and has two levels.
The top, parent, level is called “Requested Procedure” (RP) and holds the information about the patient (name, id), the study (accession number, study instance UID) and the procedure. The procedure can be described as text using attribute (0032,1060) – “Requested Procedure Description” or in a more sophisticated manner using the (0032,1064) – “Requested Procedure Code Sequence” where static tables of codes and meanings can be used to configure and maintain procedures in the RIS or HIS.
The child level is called “Scheduled Procedure Step” (SPS) and holds attributes relevant to the modality and the actual procedure to be made. A single requested procedure may hold more than one SPS if the request is for a multi-modality study, for example a chest X-Ray and a CT or whatever combination, or if for example two protocols should be applied (e.g. Chest and Abdomen). As a modality, we will use the data in the RP to identify the patient and eliminate re-typing of the name and ID and the SPS to determine what exactly to do.

Wednesday, April 18, 2012

HL7Kit Users Forum

Just wanted to inform everyone that there's a new hangout -
So, if you're already using HL7Kit or just evaluating it or even just downloaded it or never even heard about it, that's an opportunity here!
[Updated 18 May 2016]
The forum for HL7Kit Users will reopen soon. Stay tuned.
Go to, register and get more from your software.

Tuesday, April 10, 2012

Low (Resolution) and Order (of applying transformations)

I'm doing a project that involves taking huge images of 400,000,000 pixels, 20000 X 20000, 700 MB on disk, scaling them down and cutting them into reasonably sized tiles (512x512 pixels).

In this post I want to present how simple code re factoring involving merely changing the order of applying the same transformations rewarded an enormous X20 performance gain.

As a first step, I had to order extra 8GB RAM for my workstation. Once thay have arrived (eBay, 100$, 1 week) I could finally click the input file and display it on windows image preview. Before that, my workstation, initially having only 2GB, hang for 20 minutes every time I accidentally hovered over the file.

The next step was to realize that .NET 2.0 bitmap classes won't do the trick. But than I was surprised to discover that WPF BitmapSource and its sub-classes (System.Windows.Media.Imaging) are capable of handling these very large images so I built the code around them avoiding a native C++ bite-cruncher completely. This was a very good start. It made me very happy.

The third step was to wrap the cropped images in DICOM. This is a subject for a dedicated, yet to come, post about the new multi-frame objects and concatenations. Then came the last two steps that proved out to be challenging.

My initial code looked like this:

  1. Use ScaleTransform to down-scale the original image to the required resolution
  2. Loop over the down-scaled image on X and Y to do the tiling (The Cut function)
  3. Use CroppedBitmap to do the cropping (in the loop)

Here it is: