Pith. sign in

REVIEW 4 major objections 3 minor 49 references

Towards Better Cephalometric Landmark Detection with Diffusion Data Generation

T0 review · 4 major / 3 minor · reviewed 2026-08-15 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Synthetic X-rays lift cephalometric landmark detection to 82.2% success.

desk verdict Solid diffusion-based data generation for a niche medical task; the label-validity caveat is real but fixable, and the empirical gains warrant peer review. read the letter →

arxiv 2505.06055 v1 pith:BJHPU2DV submitted 2025-05-09 cs.CV

Dongqian Guo , Wencheng Han , Pang Lyu , Yuxi Zhou , Jianbing Shen This is my paper · ORCID
classification cs.CV
keywords syntheticdatagenerationdiffusionmodelscephalometriclandmarkdetectionX-rayimagesynthesisanatomicalpriorstext-to-imageconditioninglargevisionmedicalaugmentation
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper claims that a fully synthetic data pipeline can break the data-scarcity bottleneck in cephalometric landmark detection, the task of locating 38 anatomical points on dental X-rays for orthodontic diagnosis. Instead of collecting and manually labeling more patient images, the authors generate new landmark labels from anatomical priors, render X-ray-like images from those labels with a diffusion model, and use the resulting image-label pairs to pre-train landmark detectors. On a real test set, pre-training and fine-tuning with the generated data raises the success detection rate (SDR, landmarks predicted within 2 mm) by 6.5 percentage points to 82.2% with a large vision-transformer backbone, while also improving performance on rare conditions such as deciduous teeth and dentures. If true, this would make large, parameter-heavy detectors usable in a field where human annotation is the limiting resource.

What carries the argument

The load-bearing object is the anatomy-informed landmark topology map. Given a set of 38 landmarks, the AIT module builds a graph with five topological centers (landmarks 2, 4, 11, 12, and 17), assigns each center a unique color, and colors every other node by inverse-distance weighting of those five colors, with gradient-colored edges along connecting lines. This map is fed as a condition into a text-to-image diffusion generator whose text path is fine-tuned on real X-ray images paired with medical descriptions and whose control path injects the topology through zero-convolution layers. The same coordinates that seed the map become the ground-truth labels for the generated image, so the detection labels are exactly the conditions that produced the image.

What would settle it

Show a set of synthetic images to independent radiologists, ask them to localize the 38 landmarks without ever seeing the conditioning labels or topology maps, and measure agreement with the generated labels; systematic disagreement on specific landmarks would indicate that the detector is fitting conditioning artifacts. A second test is to train on synthetic data with one augmentation rule disabled, such as no angle perturbation, and check whether the drop in SDR on the real test set concentrates on the landmarks involved in that rule.

Watch

Extended reading notes

Core claim

The paper's central claim is that the generated data, not any single detection architecture, drives the gain. The authors report that every backbone they test improves after pre-training on synthetic cephalometric X-rays and fine-tuning on real data, and that the largest model benefits most: the largest vision-transformer backbone reaches 82.206% SDR and a mean radial error of 1.365 mm, up from 75.752% SDR without pre-training. They attribute this to three coupled components: the MIRA module, which turns 476 real landmark sets into more than 3,000 anatomically plausible landmark layouts by applying global affine transforms and fixed angle and distance ranges such as the SNA, SNB, and ANB angles; the AIT topology map, which converts the 38 landmarks into a colored graph whose edge gradients and interpolated node colors encode anatomical relationships; and the PDG prompt descriptions, which let the generator vary scanner style, patient character, and oral attributes. The generated images are conditioned on both the topology map and a text prompt, so each synthetic image comes with exactly the landmark labels that were used to produce it.

Load-bearing premise

The whole method assumes the synthesized landmark labels are clinically correct for the generated X-rays, so the detector learns real anatomy rather than artifacts of how the images were made.

Editorial extensions

If this is right

  • A detector pre-trained on the synthetic set and fine-tuned on real data outperforms the same detector trained on real data alone for all eight backbone configurations tested, with the gain growing as model size increases.
  • Rare clinical features with few real examples, such as deciduous teeth, dentures, orthodontic appliances, and missing teeth, each show more than a 7 percentage-point gain in SDR when synthetic pre-training is used.
  • Because labels are generated rather than human-annotated, the pipeline removes the annotation bottleneck and can be scaled to thousands of samples at near-zero marginal labeling cost.
  • Synthetic-only training reaches a 2 mm SDR of 73.841%, close to a standard backbone trained on real data, suggesting generated data alone can serve as a bootstrap when real data are absent.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The generate-labels-then-render pipeline could plausibly transfer to other landmark-heavy medical tasks, such as wrist or spine radiographs, provided the anatomy-specific angle and distance priors are re-derived for each body region.
  • A direct test not reported in the paper would be to have independent clinicians mark landmarks on synthetic images without seeing the conditioning labels; agreement would separate true anatomical fidelity from artifacts of the conditioning signal.
  • If the MIRA angle ranges encode a particular population norm, the generator may under-sample atypical anatomies; re-centering those ranges on diverse population statistics could turn the pipeline into a deliberate data-balancing tool.
  • The reported gains are measured on a test set drawn from the same sources as the training data, so external multi-center evaluation would clarify whether synthetic pre-training improves generalization to unseen scanners, populations, and imaging protocols.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 3 minor

Summary. This paper proposes Anatomy-Informed Cephalometric X-ray Generation (AICG), a pipeline that synthesizes cephalometric X-ray images together with landmark annotations. Landmark labels are first generated via MIRA, which applies global affine and anatomy-informed augmentations to real landmark sets; the landmarks are then encoded as topology-colored graphs (AIT) and combined with text prompts (PDG) to condition a Stable Diffusion/ControlNet generator. The synthetic images and labels are used to pre-train landmark detection networks, which are then fine-tuned on real data. Experiments across eight backbones report consistent gains, with ViT-huge improving from 75.752% to 82.206% SDR (2 mm) after pre-training and fine-tuning, and ablations show each module contributes. The authors also introduce the Prompt-CX dataset of 592 real images with text descriptions and release code and data.

Significance. Assuming the results are reproducible, the paper addresses a real problem: cephalometric landmark datasets are small and annotation is expensive. The proposed pipeline is one of the first to generate both images and annotations for this task, and the consistent improvements across eight detector backbones plus the ablations isolating MIRA/AIT/PDG are strengths. The release of code and the Prompt-CX dataset is a useful community contribution. The main risk is the validity of the automatically generated labels, which is not independently established by the expert study as currently designed; the absence of error bars and external validation also tempers the quantitative claims. If the authors address these points, the paper would be a solid contribution to medical image generation and landmark detection.

major comments (4)
  1. [Section V-D, Table V] The expert study is not blinded and, for synthetic images, raters are shown the exact landmark conditions used to generate the image; hence the high correspondence scores only demonstrate that the generator follows its input condition, not that the MIRA labels are anatomically correct. Please add a blinded validation in which experts mark landmarks on synthetic images without seeing the conditioning labels, or compare MIRA labels against expert re-annotation on a subset of synthetic images.
  2. [Section V-B, Table II] All detection results are from a single run with no error bars or multi-seed statistics. Since the reported gains are 1.9-6.5 percentage points in SDR, the headline improvement for ViT-huge could be within run-to-run variance. Report mean ± standard deviation over at least three seeds and a significance test.
  3. [Section V-B] The test split of 84 images is drawn from the same Prompt-CX dataset used to train the generator and detectors, so the evaluation does not measure generalization to an independent institution or scanner population. The generalizability claim would be substantially strengthened by evaluating on the CL-Detection2023 challenge test set or another external cephalometric dataset.
  4. [Section III-B, Eq. (6)-(7) and Figure 3] The MIRA anatomy-informed ranges (e.g., SNA 79-83°, WITS 0±1 mm) are manually chosen and no sensitivity analysis is provided. Because the anatomical validity of generated labels is load-bearing for the method's claim of annotation-free training, please report detection performance under reasonable variations of these ranges or provide a more rigorous justification of the ranges from the cited orthodontic literature.
minor comments (3)
  1. [Equation (14)] The notation ¯ϵ is used in the loss but ϵθ appears elsewhere; unify the notation.
  2. [Figure 2] The repeated 'Number:3000+' annotations are confusing; make the stage descriptions clearer.
  3. [Table IV] The 'Pre-train Only' rows show lower SDR than 'Without Pre-train' for the same backbone; a brief discussion of why synthetic-only training underperforms would help the reader interpret the fine-tuning results.

Circularity Check

1 steps flagged · score 4.0 of 10

Synthetic-label correctness is validated against the same conditioning landmarks used to generate the images; the main detection claim remains independently grounded on real test data.

  1. self definitional [Section III-A/III-B (Eqs. 4-7) and Section V-D (Table V)]
    "The landmark conditions generate labels based on anatomical prior knowledge, guiding the generator to create X-ray images that match these labels. These labels will serve as ground truth for landmark detection models. ... For synthetic images, the generation conditions (with topology information removed) and text prompts were annotated on the images for expert evaluation."

    The synthetic landmark labels are not independently derived from the generated images; they are the conditioning input L' produced by MIRA (Eqs. 6-7) and then used by the generator (Eq. 4) to synthesize X. The expert study overlays those same conditioning landmark positions on each synthetic image and asks doctors to rate whether the image matches them. A high correspondence score therefore only shows that the diffusion model learned to render its input condition; it does not show that the points coincide with true anatomical structures in the rendered anatomy. The paper's claim that MIRA labels are anatomically accurate and that the generated data provide valid ground truth is thus checked against the very labels used to create the image.

full rationale

The paper's central empirical claim, that pre-training on generated data improves landmark detection, is not circular: it is measured on a real test set with human-annotated ground truth (Table II), and the MIRA angle ranges come from external orthodontic literature (Tweed, Steiner, etc.), not from the paper's own results. The only load-bearing circular step is the expert-study validation of synthetic annotations: because the landmarks shown to the raters are exactly the generation conditions, the high 'Data Correspondence' scores demonstrate condition-following rather than anatomical correctness. This weakens the claim that the generated labels are clinically valid, but it does not make the main SDR result a tautology. Self-citations in the references are not load-bearing. Overall, the derivation is partially self-referential in its synthetic-label validation but retains independent empirical content, so a moderate score of 4 is appropriate.

Assumptions & free parameters 6 free parameters · 4 assumptions · 0 invented entities

The central claim depends on hand-chosen anatomical ranges, design choices for the topology and prompt modules, and a set of unverified assumptions about generator fidelity and annotation correctness. No new physical entities are introduced.

free parameters (6)
  • MIRA anatomy-informed augmentation angle ranges = SNA 81±2 deg, SNB 79±2 deg, ANB 2±2 deg, FHA 29±2 deg, FMA 25±5 deg, WITS d=0±1 mm
    Hand-set ranges from dental literature; they determine anatomical plausibility of generated landmark labels and thus the validity of synthetic ground truth.
  • Global augmentation ranges (scale, rotation, translation) = not specified numerically
    Used to generate diverse landmark labels; ranges chosen by hand and not stated precisely in the paper.
  • Number of synthetic images = 3808 (plus 476, 1904 in ablations)
    Quantity of generated training data used for pre-training; directly affects reported SDR.
  • Topological center landmarks = {L2, L4, L11, L12, L17}
    Chosen as key anatomical landmarks for AIT coloring; influences the condition representation.
  • PDG keyword set and combination rules = 50 keywords, 200+ descriptions, three groups with consistency rules
    Hand-designed by experts; controls diversity of generated images.
  • Number of augmented landmark labels n_l = >3000
    Set by the augmentation budget; affects diversity of synthetic landmark conditions.
assumptions (4)
  • domain assumption The anatomical relationships (SNA, ANB, SNB, WITS, FHA, FMA) from references [8,34,35,45] correctly describe plausible human cephalometric variation.
    MIRA uses these constraints to generate anatomically consistent landmarks; if these priors are wrong or unrepresentative, the synthetic labels mislead the detector.
  • domain assumption The fine-tuned Stable Diffusion + ControlNet can generate realistic cephalometric X-rays that faithfully realize the conditioned landmark topology.
    The entire data generation pipeline assumes the generator's output contains the specified anatomical landmarks; the paper relies on qualitative results and expert study rather than a detection-based verification of generated images.
  • domain assumption The landmark annotations in the Prompt-CX dataset are correct ground truth.
    All training and evaluation use these manual annotations as ground truth; annotation errors would bias both real and synthetic evaluations.
  • domain assumption Generated images from the same scanner distribution as the test set are a useful pre-training distribution for real test images.
    The split is random from the combined datasets; no cross-scanner generalization is tested.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Towards Better Cephalometric Landmark Detection with Diffusion Data Generation." pith.science (2026). https://pith.science/paper/BJHPU2DV

@misc{pith2026250506055,
  author       = {Pith},
  title        = {Pith review of: Towards Better Cephalometric Landmark Detection with Diffusion Data Generation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/BJHPU2DV}},
  note         = {Machine review of arXiv:2505.06055}
}
read the original abstract

Cephalometric landmark detection is essential for orthodontic diagnostics and treatment planning. Nevertheless, the scarcity of samples in data collection and the extensive effort required for manual annotation have significantly impeded the availability of diverse datasets. This limitation has restricted the effectiveness of deep learning-based detection methods, particularly those based on large-scale vision models. To address these challenges, we have developed an innovative data generation method capable of producing diverse cephalometric X-ray images along with corresponding annotations without human intervention. To achieve this, our approach initiates by constructing new cephalometric landmark annotations using anatomical priors. Then, we employ a diffusion-based generator to create realistic X-ray images that correspond closely with these annotations. To achieve precise control in producing samples with different attributes, we introduce a novel prompt cephalometric X-ray image dataset. This dataset includes real cephalometric X-ray images and detailed medical text prompts describing the images. By leveraging these detailed prompts, our method improves the generation process to control different styles and attributes. Facilitated by the large, diverse generated data, we introduce large-scale vision detection models into the cephalometric landmark detection task to improve accuracy. Experimental results demonstrate that training with the generated data substantially enhances the performance. Compared to methods without using the generated data, our approach improves the Success Detection Rate (SDR) by 6.5%, attaining a notable 82.2%. All code and data are available at: https://um-lab.github.io/cepha-generation

Figures

Figures reproduced from arXiv: 2505.06055 by the authors.

Figure 1
Figure 1. The Comparison of Traditional and Our Method. The lack of samples and complex labeling process limit the data scale and diversity. To address this limitation, we propose a conditional data generation method that can simultaneously generate X-ray images and corresponding annotations. This approach enables incorporating large￾scale detection models into this area without introducing overfitting. growth, plan orthodont… view at source ↗
Figure 2
Figure 2. Overview of the Anatomy-Informed Cephalometric X-ray Generation (AICG) Framework. Three primary stages include Condition Generation (highlighted in blue), Image Generation (marked in yellow), and Landmark Detection (stroked in purple), delineating the pipeline from condition preparation through image synthesis to landmark detection. traditional methods in this field [26], [27], [28]. Recently, the deep learning base… view at source ↗
Figure 3
Figure 3. The Proposed MIRA Module. This figure depicts the transfor￾mation of real landmark labels through Global Augmentation and the rules of Anatomy-Informed Augmentation to generate a diverse and anatomically accurate position of cephalometric landmarks. [34], [35], [45]. Each augmentation is based on well￾established relationships between key anatomical landmarks, such as angles and distances. When generating new landma… view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: The Anatomy-Informed Topology (AIT) Module: (a) Shows 38 cephalometric landmarks with their positions and names. (b) Describes the AIT module’s process, highlighting the construction of a graph with critical landmarks and employing Distance-Based Coloring for intuitive…
Figure 5
Figure 5. Figure 5: The Cephalometric Prompt Description Generator. (a) A word cloud depicts the commonality of keywords annotated by medical experts. (b) The Prompt Description Generator module’s schematic creates diverse image descriptions across three categories. D. Cephalometric Promp…
Figure 6
Figure 6. Figure 6: The Structure of Cephalometric X-ray Generator and the Adaptation Process. The adaptation includes two stages: In stage I, we finetune a text-to-image diffusion model to let it adapt to the cephalometric X-ray images domain. In stage II, we freeze the text￾to-image dif…
Figure 7
Figure 7. Figure 7: The Proposed Cephalometric Landmark Detection Pipeline. Part (a) is the backbone, which can be replaced by different networks. Part (b) is the head module. F. Cephalometric Landmark Detection Method Thanks to the scale and diversity of the generated data, we can confid…
Figure 8
Figure 8. Figure 8: The Qualitative Results of the Generated Images with Different Prompts. We use seven different prompts to generate the images [PITH_FULL_IMAGE:figures/full_fig_p008_8.png]
Figure 9
Figure 9. Figure 9: The Qualitative Results of the Generated Images with Different Landmark Conditions. We illustrate four landmark condi￾tions, generating cephalometric X-ray images captured by four scanners based on these landmarks. contains cephalometric images from 102 patients in age…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

49 extracted references · 41 canonical work pages

  1. [1]

    Application of artificial intelligence (ai) in a cephalo- metric analysis: A narrative review,

    M. Kiełczykowski, K. Kami ´nski, K. Perkowski, M. Zadurska, and E. Czochrowska, “Application of artificial intelligence (ai) in a cephalo- metric analysis: A narrative review,”Diagnostics, vol. 13, no. 16, p. 2640, 2023

    work page 2023

  2. [2]

    Missing mri pulse sequence synthesis using multi-modal generative adversarial network,

    A. Sharma and G. Hamarneh, “Missing mri pulse sequence synthesis using multi-modal generative adversarial network,”IEEE Trans. Med. Imag., vol. 39, no. 4, pp. 1170–1183, 2019

    work page 2019

  3. [3]

    Artificial intelligence and machine learning for automated cephalometric landmark identification: A meta-analysis previewed by a systematic review,

    S. Rauniyar, S. Jena, N. Sahoo, P. Mohanty, and B. P. Dash, “Artificial intelligence and machine learning for automated cephalometric landmark identification: A meta-analysis previewed by a systematic review,” Cureus, vol. 15, no. 6, 2023

    work page 2023

  4. [4]

    Hi-net: hybrid-fusion network for multi-modal mr image synthesis,

    T. Zhou, H. Fu, G. Chen, J. Shen, and L. Shao, “Hi-net: hybrid-fusion network for multi-modal mr image synthesis,”IEEE Trans. Med. Imag., vol. 39, no. 9, pp. 2772–2781, 2020

    work page 2020

  5. [5]

    Cross-modality image synthesis via weakly coupled and geometry co-regularized joint dictionary learn- ing,

    Y . Huang, L. Shao, and A. F. Frangi, “Cross-modality image synthesis via weakly coupled and geometry co-regularized joint dictionary learn- ing,”IEEE Trans. Med. Imag., vol. 37, no. 3, pp. 815–827, 2017

    work page 2017

  6. [6]

    Evaluation and comparison of anatomical landmark detection methods for cephalometric x-ray images: a grand challenge,

    C.-W. Wang, C.-T. Huang, M.-C. Hsieh, C.-H. Li, S.-W. Chang, W.-C. Li, R. Vandaele, R. Mar ´ee, S. Jodogne, P. Geurtset al., “Evaluation and comparison of anatomical landmark detection methods for cephalometric x-ray images: a grand challenge,”IEEE Trans. Med. Imag., vol. 34, no. 9, pp. 1890–1900, 2015

    work page 1900

  7. [7]

    Chest x-ray generation and data augmentation for cardiovascular abnormality classification,

    A. Madani, M. Moradi, A. Karargyris, and T. Syeda-Mahmood, “Chest x-ray generation and data augmentation for cardiovascular abnormality classification,” inProc. SPIE, vol. 10574, SPIE, 2018, pp. 415–420

    work page 2018

  8. [8]

    A benchmark for comparison of dental radiography analysis algorithms,

    C.-W. Wang, C.-T. Huang, J.-H. Lee, C.-H. Li, S.-W. Chang, M.-J. Siao, T.-M. Lai, B. Ibragimov, T. Vrtovec, O. Ronnebergeret al., “A benchmark for comparison of dental radiography analysis algorithms,” Med. Image Anal., vol. 31, pp. 63–76, 2016

    work page 2016

Show all 49 references
  1. [9]

    Genera- tion of synthetic chest x-ray images and detection of covid-19: A deep learning based approach,

    Y . Karbhari, A. Basu, Z. W. Geem, G.-T. Han, and R. Sarkar, “Genera- tion of synthetic chest x-ray images and detection of covid-19: A deep learning based approach,”Diagnostics, vol. 11, no. 5, p. 895, 2021

    2021

  2. [10]

    Diverse data generation for retinal layer segmentation with potential structure modelling,

    K. Huang, X. Ma, Z. Zhang, Y . Zhang, S. Yuan, H. Fu, and Q. Chen, “Diverse data generation for retinal layer segmentation with potential structure modelling,”IEEE Trans. Med. Imag, 2024. 11

    2024

  3. [11]

    Generative adversarial nets,

    I. Goodfellow, J. Pouget-Abadie, M. Mirza, B. Xu, D. Warde-Farley, S. Ozair, A. Courville, and Y . Bengio, “Generative adversarial nets,” Proc. Adv. Neural Inf. Process. Syst., vol. 27, 2014

    2014

  4. [12]

    Generation of 3d brain mri using auto-encoding generative adversarial networks,

    G. Kwon, C. Han, and D.-s. Kim, “Generation of 3d brain mri using auto-encoding generative adversarial networks,” inProc. Int. Conf. Med. Image Comput. Comput.-Assist. Intervent.Springer, 2019, pp. 118–126

    2019

  5. [13]

    Gans for medical image synthesis: An empirical study,

    Y . Skandarania, P.-M. Jodoinc, and A. Lalandea, “Gans for medical image synthesis: An empirical study,”arXiv:2105.05318, 2021

    2021 arXiv

  6. [14]

    High- resolution image synthesis with latent diffusion models,

    R. Rombach, A. Blattmann, D. Lorenz, P. Esser, and B. Ommer, “High- resolution image synthesis with latent diffusion models,” inProc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR), 2022, pp. 10 684–10 695

    2022

  7. [15]

    Deep high-resolution representation learning for visual recognition,

    J. Wang, K. Sun, T. Cheng, B. Jiang, C. Deng, Y . Zhao, D. Liu, Y . Mu, M. Tan, X. Wanget al., “Deep high-resolution representation learning for visual recognition,”IEEE Trans. Pattern Anal. Mach. Intell., vol. 43, no. 10, pp. 3349–3364, 2020

    2020

  8. [16]

    Cascaded latent diffusion models for high-resolution chest x-ray synthesis,

    T. Weber, M. Ingrisch, B. Bischl, and D. R ¨ugamer, “Cascaded latent diffusion models for high-resolution chest x-ray synthesis,” inProc. PAKDD. Springer, 2023, pp. 180–191

    2023

  9. [17]

    Cxr-irgen: An integrated vision and language model for the generation of clinically accurate chest x-ray image-report pairs,

    J. Shentu and N. Al Moubayed, “Cxr-irgen: An integrated vision and language model for the generation of clinically accurate chest x-ray image-report pairs,” inProc. IEEE Winter Conf. Appl. Comput. Vis, 2024, pp. 5212–5221

    2024

  10. [18]

    Pie: Simulating disease progression via progressive image editing,

    K. Liang, X. Cao, K.-D. Liao, T. Gao, Z. Chen, and T. Nama, “Pie: Simulating disease progression via progressive image editing,” arXiv:2309.11745, 2023

    2023 arXiv

  11. [19]

    Medm2g: Unifying medical multi-modal generation via cross-guided diffusion with visual invariant,

    C. Zhan, Y . Lin, G. Wang, H. Wang, and J. Wu, “Medm2g: Unifying medical multi-modal generation via cross-guided diffusion with visual invariant,” inProc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR), 2024, pp. 11 502–11 512

    2024

  12. [20]

    Adding conditional control to text- to-image diffusion models,

    L. Zhang, A. Rao, and M. Agrawala, “Adding conditional control to text- to-image diffusion models,” inProc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR), 2023, pp. 3836–3847

    2023

  13. [21]

    Lc-sd: Realistic endoscopic image generation with limited training data,

    J. Kaleta, D. Dall’alba, S. Plotka, and P. Korzeniowski, “Lc-sd: Realistic endoscopic image generation with limited training data,” inAdv. Neural Inf. Process. Syst. Workshop, 2023

    2023

  14. [22]

    Generative ai for medical imaging: extending the monai framework,

    W. H. Pinaya, M. S. Graham, E. Kerfoot, P.-D. Tudosiu, J. Dafflon, V . Fernandez, P. Sanchez, J. Wolleb, P. F. da Costa, A. Patelet al., “Generative ai for medical imaging: extending the monai framework,” arXiv:2307.15208, 2023

    2023 arXiv

  15. [23]

    Knowledge- based landmarking of cephalograms,

    A. Levy-Mandel, A. Venetsanopoulos, and J. Tsotsos, “Knowledge- based landmarking of cephalograms,”Comput. Biomed. Res., vol. 19, no. 3, pp. 282–309, 1986

    1986

  16. [24]

    An image processing system for cephalometric analysis and measurements,

    M. A. Mosleh, M. S. Baba, N. Himazian, and B. M. AL-Makramani, “An image processing system for cephalometric analysis and measurements,” inInt. Symp. Inf. Technol., vol. 4. IEEE, 2008, pp. 1–8

    2008

  17. [25]

    Cephalometric landmarks localization based on histograms of oriented gradients,

    A. A. Pouyan and M. Farshbaf, “Cephalometric landmarks localization based on histograms of oriented gradients,” inProc. Int. Conf. Signal Image Process.IEEE, 2010, pp. 1–6

    2010

  18. [26]

    Automatic cephalo- metric x-ray landmark detection by applying game theory and random forests,

    B. Ibragimov, B. Likar, F. Pernus, and T. Vrtovec, “Automatic cephalo- metric x-ray landmark detection by applying game theory and random forests,” inProc. IEEE 11th Int. Symp. Biomed. Imag. (ISBI), 2014, pp. 1–8

    2014

  19. [27]

    Fully automatic cephalometric evaluation using random forest regression-voting,

    C. Tim, F. Cooteset al., “Fully automatic cephalometric evaluation using random forest regression-voting,” inProc. IEEE 12th Int. Symp. Biomed. Imag. (ISBI), 2015, pp. 16–19

    2015

  20. [28]

    Fully automatic system for accurate localisation and analysis of cephalometric landmarks in lateral cephalograms,

    C. Lindner, C.-W. Wang, C.-T. Huang, C.-H. Li, S.-W. Chang, and T. F. Cootes, “Fully automatic system for accurate localisation and analysis of cephalometric landmarks in lateral cephalograms,”Sci. Rep., vol. 6, no. 1, p. 33581, 2016

    2016

  21. [29]

    Revisiting cephalometric landmark detection from the view of human pose estimation with lightweight super-resolution head,

    Q. Wu, S. Y . Yeo, Y . Chen, and J. Liu, “Revisiting cephalometric landmark detection from the view of human pose estimation with lightweight super-resolution head,”arXiv:2309.17143, 2023

    2023 arXiv

  22. [30]

    Deep learning based cephalometric landmark identification using landmark-dependent multi- scale patches,

    C. Lee, C. Tanikawa, J.-Y . Lim, and T. Yamashiro, “Deep learning based cephalometric landmark identification using landmark-dependent multi- scale patches,”arXiv:1906.02961, 2019

    1906 arXiv

  23. [31]

    Cephalometric landmark detection by attentive feature pyramid fusion and regression- voting,

    R. Chen, Y . Ma, N. Chen, D. Lee, and W. Wang, “Cephalometric landmark detection by attentive feature pyramid fusion and regression- voting,” inProc. Int. Conf. Med. Image Comput. Comput.-Assist. Inter- vent.Springer, 2019, pp. 873–881

    2019

  24. [32]

    Cephalformer: incorporating global structure constraint into visual features for general cephalometric landmark detection,

    Y . Jiang, Y . Li, X. Wang, Y . Tao, J. Lin, and H. Lin, “Cephalformer: incorporating global structure constraint into visual features for general cephalometric landmark detection,” inProc. Int. Conf. Med. Image Comput. Comput.-Assist. Intervent.Springer, 2022, pp. 227–237

    2022

  25. [33]

    Uod: Universal one-shot detection of anatomical landmarks,

    H. Zhu, Q. Quan, Q. Yao, Z. Liu, and S. K. Zhou, “Uod: Universal one-shot detection of anatomical landmarks,” inProc. Int. Conf. Med. Image Comput. Comput.-Assist. Intervent.Springer, 2023, pp. 24–34

    2023

  26. [34]

    The frankfort-mandibular plane angle in orthodontic diagnosis, classification, treatment planning, and prognosis,

    C. H. Tweed, “The frankfort-mandibular plane angle in orthodontic diagnosis, classification, treatment planning, and prognosis,”Am. J. Orthod. Oral Surgery, vol. 32, no. 4, pp. 175–230, 1946

    1946

  27. [35]

    Cephalometrics for you and me,

    C. C. Steiner, “Cephalometrics for you and me,”Am. J. Orthod., vol. 39, no. 10, pp. 729–755, 1953

    1953

  28. [36]

    Learning transferable visual models from natural language supervision,

    A. Radford, J. W. Kim, C. Hallacy, A. Ramesh, G. Goh, S. Agarwal, G. Sastry, A. Askell, P. Mishkin, J. Clarket al., “Learning transferable visual models from natural language supervision,” inInt. Conf. on Mach. Learn.PMLR, 2021, pp. 8748–8763

    2021

  29. [37]

    Deep residual learning for image recognition,

    K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” inProc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR), 2016, pp. 770–778

    2016

  30. [38]

    Simple baselines for human pose estimation and tracking,

    B. Xiao, H. Wu, and Y . Wei, “Simple baselines for human pose estimation and tracking,” inProc. Eur. Conf. Comput. Vis. (ECCV), 2018, pp. 466–481

    2018

  31. [39]

    Cascaded convolutional networks for automatic cephalometric landmark detection,

    M. Zeng, Z. Yan, S. Liu, Y . Zhou, and L. Qiu, “Cascaded convolutional networks for automatic cephalometric landmark detection,”Med. Image Anal., vol. 68, p. 101904, 2021

    2021

  32. [40]

    Cephalometric landmark detection in lateral x-ray images 2023,

    W. Ching-Wei, H. Bingsheng, Z. Hongyuan, M. Hikam, C. Jun, D. Juan, and L. Xuguang, “Cephalometric landmark detection in lateral x-ray images 2023,”https://cl-detection2023. grand-challenge.org/, 2023

    2023

  33. [41]

    Diffusers: State-of-the-art diffusion models,

    P. von Platen, S. Patil, A. Lozhkov, P. Cuenca, N. Lambert, K. Ra- sul, M. Davaadorj, D. Nair, S. Paul, W. Berman, Y . Xu, S. Liu, and T. Wolf, “Diffusers: State-of-the-art diffusion models,”https: //github.com/huggingface/diffusers, 2022

    2022

  34. [42]

    Which images to label for few-shot medical landmark detection?

    Q. Quan, Q. Yao, J. Li, and S. K. Zhou, “Which images to label for few-shot medical landmark detection?” inProc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR), 2022, pp. 20 606–20 616

    2022

  35. [43]

    Huggingface’s transformers: State-of-the-art natural language processing,

    T. Wolf, L. Debut, V . Sanh, J. Chaumond, C. Delangue, A. Moi, P. Cis- tac, T. Rault, R. Louf, M. Funtowiczet al., “Huggingface’s transformers: State-of-the-art natural language processing,”arXiv:1910.03771, 2019

    1910 arXiv

  36. [44]

    An image is worth 16x16 words: Transformers for image recognition at scale,

    A. Dosovitskiy, L. Beyer, A. Kolesnikov, D. Weissenborn, X. Zhai, T. Unterthiner, M. Dehghani, M. Minderer, G. Heigold, S. Gellyet al., “An image is worth 16x16 words: Transformers for image recognition at scale,”arXiv:2010.11929, 2020

    2010 arXiv

  37. [45]

    Floating norms for individualising the anb angle and the wits appraisal in orthodontic cephalometric analysis based on guiding variables,

    E. Paddenberg, P. Proff, and C. Kirschneck, “Floating norms for individualising the anb angle and the wits appraisal in orthodontic cephalometric analysis based on guiding variables,”Journal of Orofacial Orthopedics, vol. 84, no. 1, p. 10, 2021

    2021

  38. [46]

    Synthetic data in machine learning for medicine and healthcare,

    R. J. Chen, M. Y . Lu, T. Y . Chen, D. F. Williamson, and F. Mahmood, “Synthetic data in machine learning for medicine and healthcare,”Nature Biomedical Engineering, vol. 5, no. 6, pp. 493–497, 2021

    2021

  39. [47]

    Data augmentation for medical imaging: A systematic literature review,

    F. Garcea, A. Serra, F. Lamberti, and L. Morra, “Data augmentation for medical imaging: A systematic literature review,”Computers in Biology and Medicine, vol. 152, p. 106391, 2023

    2023

  40. [48]

    Synthetic data generation methods in healthcare: A review on open-source tools and methods,

    V . C. Pezoulas, D. I. Zaridis, E. Mylona, C. Androutsos, K. Apostolidis, N. S. Tachos, and D. I. Fotiadis, “Synthetic data generation methods in healthcare: A review on open-source tools and methods,”Computational and Structural Biotechnology Journal, 2024

    2024

  41. [49]

    Addressing medical imaging limitations with synthetic data generation,

    G. Pengfei, Y . Dong, Z. Can, and X. Daguang, “Addressing medical imaging limitations with synthetic data generation,”NVidia Tech. Blog, 2024

    2024

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.