An optimized process flow for rapid segmentation of cortical bones of the craniofacial skeleton using the level-set method
Bibliographic record
Abstract
Accurate representation of skeletal structures is essential for quantifying structural integrity, for developing accurate models, for improving patient-specific implant design and in image-guided surgery applications. The complex morphology of thin cortical structures of the craniofacial skeleton (CFS) represents a significant challenge with respect to accurate bony segmentation. This technical study presents optimized processing steps to segment the three-dimensional (3D) geometry of thin cortical bone structures from CT images. In this procedure, anoisotropic filtering and a connected components scheme were utilized to isolate and enhance the internal boundaries between craniofacial cortical and trabecular bone. Subsequently, the shell-like nature of cortical bone was exploited using boundary-tracking level-set methods with optimized parameters determined from large-scale sensitivity analysis. The process was applied to clinical CT images acquired from two cadaveric CFSs. The accuracy of the automated segmentations was determined based on their volumetric concurrencies with visually optimized manual segmentations, without statistical appraisal. The full CFSs demonstrated volumetric concurrencies of 0.904 and 0.719; accuracy increased to concurrencies of 0.936 and 0.846 when considering only the maxillary region. The highly automated approach presented here is able to segment the cortical shell and trabecular boundaries of the CFS in clinical CT images. The results indicate that initial scan resolution and cortical-trabecular bone contrast may impact performance. Future application of these steps to larger data sets will enable the determination of the method's sensitivity to differences in image quality and CFS morphology.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".