Article Text

Download PDFPDF

P282 The design and validation of a novel semiautomatic lung navigation platform
  1. KA Khan1,
  2. P Nardelli2,
  3. J Alex2,
  4. C O’Shea2,
  5. P Cantillon-Murphy2,
  6. MP Kennedy1
  1. 1Department of Respiratory Medicine, Cork University Hospital/University College Cork, Cork, Ireland
  2. 2School of Engineering, University College Cork, Cork, Ireland


Introduction In the era of lung cancer screening, tissue acquisition of peripheral lung lesions remains a challenge. We have developed a 3D electromagnetic navigation platform with airway segmentation and virtual bronchoscopy using a open source 3D slicer environment.

Methods The open source visualisation software (3D Slicer created a detailed airway segmentation and virtual bronchoscopy model from acquired CT images. A magnetic field emitter board provides tracking of a semiautomatic locatable sensor probe (SALP) in the working channel of the bronchoscopewith always-on tip tracked sensor and can be steered both manually and automaticaly with joy stick, for accurate localization of peripheral lung lesion.

An extensive ex-vivo evaluation was performed in a breathing lung model that was developed using inflatable plasticized pig lungs in a negative-pressure Following this, in-vivo real time navigation in a live porcine model using a selection of novel radioopaque fiducials placed endobronchially into distal airways.

Results After completion of a selection of experiments using the breathing pig lung model, fiducials were placed endobronchially in our live porcine model. Thereafter, CT images were used to create a virtual airway 3D segmentation model. After multiplaner re-construction, land mark based registration was performed to align the CT and anaesthetised porcine. Manual and automatic navigation with the bronchoscope containing the SALP was performed. The average navigation distance covered was 85.3 mm. The navigational system accurately determined 84% of the navigation points within the airways.

Conclusion Our navigational platform is inexpensive and open source and is the first to utilise SALP. In our model, there is good agreement between the position of the sensor probe during bronchoscopic navigation and as visualised in virtual bronchoscopy. Further work is being carried out to improve registration and accuracy of the navigational system before a pilot study in patients with peripheral lung nodules.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.