The primary aim is to develop a software algorithm that has the capacity to detect the normal 18 anatomical structures of the lung by using the position of the scope during the bronchoscopy procedure and using existing bronchoscopy technology.
2.000 subjects, preferable 150-200 number of patients per site. The 2.000 videos will be divided into an 80%/20% split (training and testing). It is a wish to obtain a spread of videos coming from across Europe, preferable from Germany, France and Denmark, with up to 800 videos coming from Denmark The patient population indicated are patients indicated for full airway bronchoscopy. Of this group it is the aim to enrol the following: 1. Patients with obstructing tumors (when tumor blocks the view of the camera) - no more than 5% (120 patients) of studies 2. Bronchoscopies on indication of haemoptysis (active bleeding) - no more than 5% (120 patients) of studies 3. Patients with stents or valves - no more than 5% (120 patients) of studies bronchoscopies. 4. Patients with former lung operations, partial or full resections - no more than 2% (48 patients).
Study Type
OBSERVATIONAL
Enrollment
833
All subjects who are scheduled for a bronchoscopy may be eligible.
Rigshospitalet
Copenhagen, Denmark
Number of anatomical segments accessed during bronchoscopy
accurate photo documentation of all 18 anatomical segments +/- abnormalities
Time frame: 1 day
Number of anatomical segments identified
Identified anatomical segments by the core lab/ Identified anatomical segments by the Machine Learning (after training, using the test videos).
Time frame: 1 day
Rate of total number of lesions detected
total lesions detected by the core lab/ total lesions detected by Machine Learning (after training, using the test videos).
Time frame: 1 day
Number of sections and lesions detected by core lab and machine learning
Anatomical sections + lesions detected by core lab versus anatomical sections + lesions detected by machine learning
Time frame: 1 day
This platform is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare professional.