Evaluation metrics for bone segmentation in ultrasound
Bibliographic record
Abstract
Tracked ultrasound is a safe alternative to X-ray for imaging bones. The interpretation of bony structures is challenging as ultrasound has no specific intensity characteristic of bones. Several image segmentation algorithms have been devised to identify bony structures. We propose an open-source framework that would aid in the development and comparison of such algorithms by quantitatively measuring segmentation performance in the ultrasound images. True-positive and false-negative metrics used in the framework quantify algorithm performance based on correctly segmented bone and correctly segmented boneless regions. Ground-truth for these metrics are defined manually and along with the corresponding automatically segmented image are used for the performance analysis. Manually created ground truth tests were generated to verify the accuracy of the analysis. Further evaluation metrics for determining average performance per slide and standard deviation are considered. The metrics provide a means of evaluating accuracy of frames along the length of a volume. This would aid in assessing the accuracy of the volume itself and the approach to image acquisition (positioning and frequency of frame). The framework was implemented as an open-source module of the 3D Slicer platform. The ground truth tests verified that the framework correctly calculates the implemented metrics. The developed framework provides a convenient way to evaluate bone segmentation algorithms. The implementation fits in a widely used application for segmentation algorithm prototyping. Future algorithm development will benefit by monitoring the effects of adjustments to an algorithm in a standard evaluation framework.
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 machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.011 | 0.053 |
| Meta-epidemiology (narrow) | 0.002 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.007 | 0.004 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.003 | 0.002 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.002 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
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 source (direct Gemma or distilled Codex), 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".