Trial document
This trial has been registered retrospectively.
DRKS00014337
Trial Description
Title
Clinical evaluation of semi-automatic open-source algorithmic software segmentation of the mandibular bone
Trial Acronym
enFaced
URL of the Trial
https://www.tugraz.at/index.php?id=26448
Brief Summary in Lay Language
Computer-assisted technologies such as image-based segmentation play an important in the diagnosis and treatment support in oral- and maxillofacial surgery. However, although many image-based segmentation approaches exist, their clinical in-house use is often strongly limited due to technical, human or financial resources. Especially in open-source based segmentation, systematic evaluations of segmentation approaches are lacking. Therefore, the aim of this study is to assess the real segmentation quality of a commonly available and license-free segmentation method in the mandible. The assessment is done in a comparison between segmented CT-image data of the mandible. Segmented patient-specific clinical image data performed by an automatic segmentation algorithm are compared with the manual ground truth segmentations performed by clinical experts. Assessment parameters are amongst others the Dice Score Coefficient (DSC, %) and the Hausdorff Distance (HD, voxel).
This study is a systematic comparison that evaluates a license-free, open-source segmentation approach in the mandible based on clinical CT-data for the improvement of segmentation algorithms and a potential clinical use in patient-individualized medicine in the field of oral and maxillofacial surgery. Further, the results presented are reproducible by others and can be used for both clinical and research purposes.
Brief Summary in Scientific Language
Computer-assisted technologies such as image-based segmentation play an important in the diagnosis and treatment support in oral- and maxillofacial surgery. However, although many image-based segmentation approaches exist, their clinical in-house use is often strongly limited due to technical, human or financial resources. Especially in open-source based segmentation, systematic evaluations of segmentation approaches are lacking. Therefore, the aim of this study is to assess the real segmentation quality of a commonly available and license-free segmentation method in the mandible. The assessment is done in a comparison between segmented CT-image data of the mandible. Segmented patient-specific clinical image data performed by an automatic segmentation algorithm are compared with the manual ground truth segmentations performed by clinical experts. Assessment parameters are amongst others the Dice Score Coefficient (DSC, %) and the Hausdorff Distance (HD, voxel).
This study is a systematic comparison that evaluates a license-free, open-source segmentation approach in the mandible based on clinical CT-data for the improvement of segmentation algorithms and a potential clinical use in patient-individualized medicine in the field of oral and maxillofacial surgery. Further, the results presented are reproducible by others and can be used for both clinical and research purposes.
Do you plan to share individual participant data with other researchers?
[---]*
Description IPD sharing plan:
[---]*
Organizational Data
- DRKS00014337
- 2018/05/04
- [---]*
- yes
- Approved
- EK-29-143 ex 16/17, Ethikkommission der Medizinischen Universität Graz
Secondary IDs
- [---]*
Health Condition or Problem studied
-
bone fracture, bone defect
- S02 - Fracture of skull and facial bones
- K10 - Other diseases of jaws
- K07 - Dentofacial anomalies [including malocclusion]
Interventions/Observational Groups
- Establishment of a CT database of the mandible with datasets from the clinical routine for the evaluation of an image-based computer program.
- Automatic lower jawbone segmentation with an open-source algorithm on CT data basis: 1 time of measurement (T) with Hausdorff Distance (HD, voxel)/ Dice Similarity Score (DSC, %)/ Volume (V, mm3)/ Voxel (Vx, Number)/ Time (t, min., sec.)
- Manual lower jawbone segmentation by clinical experts (Ground Truth) on CT data basis: 1 time of measurement (T) with Hausdorff Distance (HD, voxel)/ Dice Similarity Score (DSC, %)/ Volume (V, mm3)/ Voxel (Vx, Number)/ Time (t, min., sec.)
Characteristics
- Non-interventional
- Observational study
- Non-randomized controlled trial
- Open (masking not used)
- [---]*
- Other
- Basic research/physiological study
- Parallel
- N/A
- N/A
Primary Outcome
Comparison image based segmentation of lower jaw CT-data between automatic (algorithm) vs. manual (clinical experts) at measurement point T by Hausdorff Distance (HD, voxel) and Dice Similarity Score (DSC, %).
Secondary Outcome
Comparison image based segmentation of lower jaw CT-data between automatic (algorithm) vs. manual (clinical experts) at measurement point T by Volume (V, mm3), Voxel (Vx, Number), time (t, min., sec.).
Countries of Recruitment
- Austria
Locations of Recruitment
- Medical Center
Recruitment
- Actual
- 2017/04/01
- 10
- Monocenter trial
- National
Inclusion Criteria
- Both, male and female
- 18 Years
- 99 Years
Additional Inclusion Criteria
• Physiologically functional, complete lower jawbones
• Age ≥18 Years
• Age <99 Years
• Data due to the medical indication in the clinical diagnostic at the department of Oral and maxillofacial surgery
Exclusion Criteria
• Implants or osteosynthesis material
• Age <18 Years
• Lower jawbone necrosis or pathologic-cystic changes of the bones
Addresses
-
start of 1:1-Block address primary-sponsor
- Medizinische Universität Graz
- Mr. DDDr. Jürgen Wallner
- Auenbrugger Platz 1
- 8036 Graz
- Austria
end of 1:1-Block address primary-sponsorstart of 1:1-Block address contact primary-sponsor- 004331638530193
- [---]*
- j.wallner at medunigraz.at
- http://www.lkh-graz.at; www.medunigraz.at
end of 1:1-Block address contact primary-sponsor -
start of 1:1-Block address other
- TU Graz
- Mr. Dr. Dr. habil. Jan Egger
- Inffeldgasse 16c/2
- 8010 Graz
- Austria
end of 1:1-Block address otherstart of 1:1-Block address contact other- +43 316 873-5076
- +43 316 873 5050
- egger at tugraz.at
- https://www.tugraz.at/home/
end of 1:1-Block address contact other -
start of 1:1-Block address scientific-contact
- Medizinische Universität Graz
- Mr. DDDr. Jürgen Wallner
- Auenbruggerplatz 5/1
- 8036 Graz
- Austria
end of 1:1-Block address scientific-contactstart of 1:1-Block address contact scientific-contact- +43/316/385-12428
- [---]*
- j.wallner at medunigraz.at
- https://www.medunigraz.at/en/
end of 1:1-Block address contact scientific-contact -
start of 1:1-Block address public-contact
- Medizinische Universität Graz
- Mr. DDDr. Jürgen Wallner
- Auenbruggerplatz 5/1
- 8036 Graz
- Austria
end of 1:1-Block address public-contactstart of 1:1-Block address contact public-contact- +43/316/385-12428
- [---]*
- j.wallner at medunigraz.at
- https://www.medunigraz.at/en/
end of 1:1-Block address contact public-contact
Sources of Monetary or Material Support
-
start of 1:1-Block address materialSupport
- Fonds zur Förderung der wissenschaftlichen Forschung (FWF)
- Sensengasse 1
- 1090 Wien
- Austria
end of 1:1-Block address materialSupportstart of 1:1-Block address contact materialSupport- [---]*
- [---]*
- [---]*
- https://www.fwf.ac.at/
end of 1:1-Block address contact materialSupport -
start of 1:1-Block address otherSupport
- Austrian Science Fund (FWF) KLI 678-B31: “enFaced: Virtual and Augmented Reality Training and Navigation Module for 3D-Printed Facial Defect Reconstructions” (PIs: Jürgen Wallner and Jan Egger)
- 1090 Wien
- Austria
end of 1:1-Block address otherSupportstart of 1:1-Block address contact otherSupport- [---]*
- [---]*
- [---]*
- [---]*
end of 1:1-Block address contact otherSupport
Status
- Recruiting complete, follow-up complete
- [---]*
- [---]*
- 2019/05/06
- 0
- 10
Trial Publications, Results and other Documents
- [---]*