Validity and accuracy of automatic cobb angle measurement on 3D spinal ultrasonographs for children with adolescent idiopathic scoliosis: SOSORT 2024 award winner
Bibliographic record
Abstract
Abstract Purpose Ultrasonography for scoliosis is a novel imaging method that does not expose children with adolescent idiopathic scoliosis (AIS) to radiation. A single ultrasound scan provides 3D spinal views directly. However, measuring ultrasonograph parameters is challenging, time-consuming, and requires considerable training. This study aimed to validate a machine learning method to measure the coronal curve angle on ultrasonographs automatically. Methods A total of 144 3D spinal ultrasonographs were extracted to train and validate a machine learning model. Among the 144 images, 70 were used for training, and 74 consisted of 144 curves for testing. Automatic coronal curve angle measurements were validated by comparing them with manual measurements performed by an experienced rater. The inter-method intraclass correlation coefficient (ICC 2,1 ), standard error of measurement (SEM), and percentage of measurements within clinical acceptance (≤ 5°) were analyzed. Results The automatic method detected 125/144 manually measured curves. The averages of the 125 manual and automatic coronal curve angle measurements were 22.4 ± 8.0° and 22.9 ± 8.7°, respectively. Good reliability was achieved with ICC 2,1 = 0.81 and SEM = 1.4°. A total of 75% (94/125) of the measurements were within clinical acceptance. The average measurement time per ultrasonograph was 36 ± 7 s. Additionally, the algorithm displayed the predicted centers of laminae to illustrate the measurement. Conclusion The automatic algorithm measured the coronal curve angle with moderate accuracy but good reliability. The algorithm’s quick measurement time and interpretability can make ultrasound a more accessible imaging method for children with AIS. However, further improvements are needed to bring the method to clinical use.
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 | 0.031 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.000 |
| Insufficient payload (model declined to judge) | 0.001 | 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".