Fully Automated Analysis of the Anatomic and Mechanical Axes From Pediatric Standing Lower Limb Radiographs Using Convolutional Neural Networks
Bibliographic record
Abstract
BACKGROUND: Lower limb alignment is the quantification of a set of parameters that are commonly measured radiographically to test for and track a wide range of skeletal pathologies. Determining limb alignment is a commonly performed yet laborious task in the pediatric orthopaedic setting and is therefore an interesting goal for automation. METHODS: We employ a machine learning approach using convolutional neural networks (CNNs) to segment pediatric weight-bearing lower limb radiographs. The results are then used with custom Matlab code to extract anatomic landmarks and to determine lower limb alignment parameters. RESULTS: Measurements obtained from the automated workflow proposed here were compared with manual measurements performed by orthopaedic surgery fellows. Mechanical axis deviation was determined within a mean of 2.02 mm. Lateral distal femoral angle and medial proximal tibial angle were determined with a mean deviation of 1.73 and 2.90 degrees, respectively. The calculation speed for the full set of mechanical and anatomic axis parameters was found to be ~2 seconds per radiograph. CONCLUSIONS: The CNN-based approach proposed in this work was shown to produce results comparable to orthopaedic surgery fellows at fast calculation speed. Although further work is needed to validate these results against radiographs and measurements from other centers, we see this as a promising start and a functional path that can be employed in further research. CLINICAL RELEVANCE: CNNs are a promising approach to automating commonly performed, repetitive tasks, especially those pertaining to image processing. The time savings are particularly important in clinical research applications where large sets of radiographs are routinely available and require analysis. With further development of these algorithms, we anticipate significantly improved agreement with expert-measured results and the calculation speed.
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 | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.003 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 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".