Fraxinus

An open source, software only, navigation system for bronchoscopy.

Bronchoscopy is the endobronchial procedure for inspection and diagnostic sampling in the airways, e.g. to diagnose lung cancer. Navigating the flexible bronchoscope through the airways to the target for the diagnostic sampling, like biopsy or fine needle aspiration, is challenging due to the numerous divisions of the airways tree structure and the lack of direct visibility of smaller peripheral lesions. Even with fluoroscopy guidance, the diagnostic success rate for bronchoscopically non-visible tumors is as low as 15% compared to 80% for visible tumors, dependent on tumor size, the doctors experience and the method used for sampling[1-3]. Using navigated bronchoscopy is a possibility to reduce these challenges and thus increase the biopsy success rate. 

Download Fraxinus

DISCLAIMER: Fraxinus is a research tool: It is not intended for normal clinical use, and is not FDA nor CE approved.

 

Fraxinus is free.

Fraxinus is technically a front end, containing all open source code and abilities of CustusX, but with a simplified and streamlined interface adapted to bronchoscopy procedures, planning and guidance.

Fraxinus has a user interface customized to the different phases of bronchoscopy planning: Import of CT Thorax images in DICOM format, target pinpointing, automatic anatomical segmentation of airways, and navigation. The navigation screen contains virtual bronchoscopy, 3D models of airways, and cross-sectional slices of the chest CT. For the target navigation, the system generates a route to the target, and the user navigates along this route using a mouse or keyboard. Fraxinus contains no real-time position tracking of bronchoscope or tools. 

 

Testing

Fraxinus is tested and run on several different platforms and machine configurations. 

The algorithm for airway segmentation has been tested and the results have been published [4], the data used to test the algorithm is available to be downloaded here.

To import Mhd datasets:

  1. Start Fraxinus
  2. Create new patient
  3. Select File -> Import data (instead of importing DICOM)
  4. <follow normal usage>

Phantom data

CT Dicom data for EBUS Phantom - Ultrasonic Bronchoscopy Simulator LM-099 307 (KOKEN CO., LTD, Tokyo, Japan) can be downloaded here.

 

1. W. A. Baaklini, M. A. Reinoso, A. B. Gorin, A. Sharafkaneh, and P. Manian. Diagnosticyield of fiberoptic bronchoscopy in evaluating solitary pulmonary nodules.CHESTJournal, 117(4):1049–1054, 2000

2. G. Schreiber and D. C. McCrory. Performance characteristics of different modalities fordiagnosis of suspected lung cancer: summary of published evidence.Chest journal, 123(1suppl):115S–128S, 2003

3. R. Yung. Tissue diagnosis of suspected lung cancer: selecting between bronchoscopy,transthoracic needle aspiration, and resectional biopsy.Respiratory care clinics of NorthAmerica, 9(1):51–76, 2003.

4. Reynisson PJ, Scali M, Smistad E, Hofstad EF, Leira HO, et al. (2015) Airway Segmentation and Centerline Extraction from Thoracic CT – Comparison of a New Method to State of the Art Commercialized Methods. PLoS ONE 10(12): e0144282. doi: 10.1371/journal.pone.0144282