5 Tips to Save Time with DICOM Segmentation in Simpleware Software

1) Ensure Patient Data Confidentiality

Working with DICOM series usually involves processing patient image data, which is subject to confidentiality regulations such as GDPR and HIPAA. Personally identifiable data is stored in the DICOM tags and often needs to be anonymized before any segmentation can be carried out. By default, Simpleware software anonymizes DICOM tags as part of the import process. However, it also provides options to anonymize only selected DICOM tags if some imaging data needs to be retained for the workflow

2) Use Quick and Interactive Tools for Different Tasks

Simpleware software provides quick and interactive tools for working with medical scans to extract regions of interest. The image volume greyscale data (e.g. from CT or MRI) is segmented to create one or more volumes called masks, which are then used to identify particular regions within the background image.

The most common options for obtaining a rapid segmentation include applying a Threshold tool to select a range of greyscales that contain the regions of interest, such as the heart or bone. Following this step with a Mask Flood Fill removes any disconnected regions, while Split Regions enables additional areas that are connected to be removed or separated. For more advanced segmentation where fine details are needed, 3D Editing and manual Painting tools are available.

3) Take Advantage of the GUI and Scripting

Everyday use of the software involving large amounts of DICOM import and segmentation can be made more straightforward by collecting commonly used segmentation tools in the My Tools ribbon. Scripting options in the Simpleware Scripting API are available to speed up common workflows using Python and C#. Less experienced users can also record steps with a Macro option, while those who want to add their own filters or tools can import them as Plug-Ins. Running the software from the Command Line similarly makes it easier to create a workflow with other programs that involve repeatable segmentation steps. More experienced Simpleware software users have many options to customize the GUI with small scripts, which work particularly well when processing a large number of similar datasets.

4) Automate DICOM Segmentation

The Simpleware AI-based machine learning algorithms are making it easier to significantly speed up common DICOM segmentation workflows for a variety of anatomies such as hips, knees, and hearts. Trained algorithms can automatically segment and add landmarks to DICOM data, reducing the amount of manual intervention needed by users, who then only need to review and make any remaining edits. The key advantage of this approach to DICOM segmentation is that it frees up time for users during the initial image processing stage, allowing them to move faster to the next phases of the workflow, such as design, 3D printing, and simulation.

5) Scale Up Automation

It is becoming increasingly important for those segmenting DICOM data to scale up their workflows to account for large studies. For example, processing hundreds of scans per month for surgical planning or in silico modeling requires a lot of datasets to be processed. For these situations, the Simpleware team provides fully customized single-check solutions using AI/ML techniques for specific series of steps and requirements. With this approach, common DICOM segmentation tasks can be integrated into a repeatable workflow able to scale to hundreds or thousands of cases every month, enabling patient-specific designs and other decisions to be made without undue delay.