Description

A tracheostomy is a surgical procedure that creates an opening in the neck to access the trachea, allowing patients to breathe when their airway is blocked or compromised. While this procedure saves lives, complications often arise if the tracheostomy tube doesn’t fit properly—leading to discomfort, infections, or even airway damage. Customized tubes could address these issues, but designing them requires precise, individualized measurements of the trachea, which have been difficult to obtain efficiently.

As part of a collaboration between the Industrial Design and Ambient Intelligence groups at Saxion, I worked on developing an automated workflow to extract detailed tracheal measurements from CT scans. Using 3D Slicer and a deep learning segmentation model (VISTA-3D), I created an automatic Python based workflowto automate the measurement process.

By automating this process, it is possibel to analyse large amounts of CT scand in an automatic objective way. This information could then be used to design better fitting tubes, which leads to fewer complications, and improved patient comfort. While this project is just one piece of the puzzle, it’s a meaningful step toward more precise, data-driven medical solutions.

References

1. [Preprint] B. Ton, S. Farooq, J. Veenstra, Automated Tracheal Morphometrics Using Deep Learning: Toward Custom Tracheostomy Tubes, medRxiv, 2025