Pith. sign in

REVIEW 4 major objections 6 minor 36 references

An Intrinsically Explainable Approach to Detecting Vertebral Compression Fractures in CT Scans via Neurosymbolic Modeling

T0 review · 4 major / 6 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read A two-rule pipeline detects spine fractures in CT scans as accurately as a black-box deep network while showing the evidence behind each call.

desk verdict A transparent height-ratio rule for VCF detection that works on a held-out VerSe19 split, but the 'predefined' rule is actually a fitted model and the segmentation layer is not validated on fractured vertebrae. read the letter →

arxiv 2412.17258 v1 pith:DGQPIWQ7 submitted 2024-12-23 cs.LG cs.CVeess.IV

Blanca Inigo , Yiqing Shen , Benjamin D. Killeen , Michelle Song , Axel Krieger , Christopher Bradley , Mathias Unberath This is my paper · ORCID
classification cs.LGcs.CVeess.IV
keywords vertebralcompressionfractureexplainableAIneurosymbolicmodelingCTimaginginterpretablemachinelearningheightmaprule-baseddetectionopportunisticscreening
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

The paper sets out to show that vertebral compression fractures can be detected in CT scans by a pipeline that is transparent end to end: a deep network performs only the segmentation, and the diagnosis itself is a fixed rule set over vertebral height ratios. On the VerSe19 dataset, the rule set reaches 96 percent accuracy and 91 percent sensitivity, matching the best masked-DenseNet black box on sensitivity and surpassing it on accuracy, precision, and F1. The authors argue this matters because opportunistic screening is a high-stakes setting where clinicians will only adopt AI advice they can verify, and the rules map directly onto the height-loss criteria radiologists already use. If the claim holds, explainability does not have to be bought at the cost of performance in this task.

What carries the argument

The load-bearing object is the vertebral body height map: a 2D grid that records the height of the vertebral body in every grid column, computed in a consistent local coordinate system so that the maps are comparable across patients and scanners. Height maps convert a high-dimensional CT volume into a small set of anatomically meaningful numbers, the mean and standard deviation of height over seven regions of the vertebral body, which is what makes a fixed rule set possible. The decision mechanism is the two-condition rule of Eq. 3, and the rules are discovered rather than guessed using RuleFit with LASSO, which selects the thresholds and yields a linear model whose coefficients are the rule weights.

What would settle it

Run the same pipeline on VerSe20 or another external CT cohort, stratify by Genant fracture grade, and re-rank the interpretable rule set against the masked DenseNet; if the accuracy edge disappears whenever segmentation errors on severely deformed vertebrae are accounted for, the claim that the transparent rules themselves carry the performance is falsified.

Watch

Extended reading notes

Core claim

The central discovery is that three mutually exclusive conditions, effectively two thresholds, reproduce what a deep classifier learns. A vertebra is flagged as fractured when the average height of the antero-centric section is at most 91 percent of the posterior section's average height, or when the central section's average height is at most 81 percent of the central height of the reference vertebra, defined as the vertebra with the tallest central section. These thresholds are not hand-set; they are distilled with the RuleFit algorithm from a gradient-boosting ensemble over measurements extracted from per-vertebra 2D height maps. The maps themselves come from a preprocessing chain: TotalSegmentator labels the vertebral bodies, marching cubes builds meshes, k-means on surface normals detects the main surfaces, the mesh is reoriented to a consistent coordinate system, and heights are projected onto a grid. On 1,460 thoracolumbar vertebrae from VerSe19 with Genant grading, this rule set reports an F1 of 0.81, accuracy of 0.96, and recall of 0.91, compared with 0.77, 0.95, and 0.91 for masked DenseNet. The paper explicitly notes a common failure mode in which vertebral shape suggests a compression fracture even though the label does not confirm one, and it reports excluding vertebrae with metal implants and scans with only one annotated vertebra. Its claim is that height measurements taken across the whole axial plane, instead of only the classical anterior, middle, and posterior points, are what let two simple ratios capture real deformations.

Load-bearing premise

The load-bearing premise is that the automated height maps remain accurate on fractured and severely deformed vertebrae, since the fixed 91 percent and 81 percent ratios can only separate fractures from normal anatomy if the segmentation, surface detection, and reorientation steps have not silently corrupted the heights.

Editorial extensions

If this is right

  • Every positive prediction comes with the height-map regions that triggered it, so a clinician can check the anatomical evidence instead of trusting a saliency map.
  • Because the diagnostic layer is a fixed threshold rule, the model does not need its rule layer retrained when applied to a new CT scan, only the preprocessing.
  • At matched sensitivity the interpretable rule set exceeds the masked-DenseNet baseline in accuracy, precision, and F1, supporting the paper's claim that transparency and performance are not in tension here.
  • Height measurements taken across the whole axial plane, rather than at the classical anterior, middle, and posterior points, are what let the two-rule set generalize across different deformation patterns.

Reading between the lines

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

  • Because the evaluation excluded metal-implant vertebrae and single-vertebra scans, a real-world screening cohort that contains those cases is the natural stress test of the 96 percent accuracy figure.
  • The thresholds' portability rests on the accuracy of the segmentation and surface-reorientation steps for severely deformed vertebrae; auditing that preprocessing on fracture cases is the next step the paper leaves implicit.
  • If the rules are as general as claimed, the same recipe of ratio thresholds over anatomically registered height maps should transfer to other CT-based bone morphology scoring, such as hip or wrist compression assessment.
  • The visible height-map reasoning creates a concrete way to study trust: show radiologists the rule's trigger regions and measure whether their accept or reject decisions on flagged vertebrae improve.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 6 minor

Summary. The paper proposes a neurosymbolic pipeline for detecting vertebral compression fractures (VCFs) in CT volumes. A deep-learning segmentation module (TotalSegmentator) extracts vertebral bodies; a shape-based algorithm builds per-vertebra height maps, defines seven anatomical regions of interest, and computes pairwise height ratios. A rule set is then obtained via RuleFit (GradientBoosting + LASSO), yielding the two-condition rule in Eq. (3): a vertebra is positive if the antero-centric height is below 91% of the posterior height, or, failing that, the central height is below 81% of the reference vertebra's central height. On VerSe19, the method is reported to achieve 96% accuracy and 91% sensitivity, compared with 95% accuracy and 91% sensitivity for a masked DenseNet baseline. The paper argues that this transparent rule-based model can match or surpass a black-box deep network while providing explanations for each prediction.

Significance. If the reported result is robust, the paper makes a useful demonstration: a compact, readable height-ratio rule can be competitive with a deep-learning black box on vertebra-level VCF detection from CT, and the visualization of the reasoning (Figure 2) is a genuine practical advantage for clinician-facing screening tools. The use of a public benchmark (VerSe19) with Genant grade labels, the explicit final rule, and the direct comparison to DenseNet and ResNeXt are strengths. However, the central claim is not yet fully established because the evaluation lacks uncertainty quantification, the measurement layer feeding the rule is not validated on fractured vertebrae, and the fitted thresholds are presented as intrinsic explanations without robustness analysis. These are addressable within the scope of the manuscript, but they are load-bearing.

major comments (4)
  1. [Section 2.2, Eq. (3)] The rule in Eq. (3) is presented as the model's explanation, but the thresholds 0.91 and 0.81, the selected number of rules, and the coefficients are estimated parameters from GradientBoosting and LASSO fitted on the training split. This makes the 'predefined' rule set a fitted hypothesis rather than a clinical a priori rule. The generalizability claim would be materially stronger if the paper reported the stability of these thresholds across bootstrap or cross-validation runs, and ideally validated the rule on an independent cohort or pre-registered it. Without such analysis, the 96% test accuracy could reflect overfitting of the rule thresholds to the VerSe19 dataset family.
  2. [Section 2.1, Figure 1] The height-map extraction pipeline is not validated on fractured vertebrae. TotalSegmentator segmentation, k-means surface detection (k=6), mesh reorientation, and grid projection are chained together to produce the A0/P and C/Cref ratios that feed Eq. (3), but compression fractures alter exactly the vertebral shape and surface geometry that these steps must recover. The paper reports no failure audit for segmentation or reorientation on fractured cases, and it does not include an ablation using the ground-truth VerSe19 segmentations instead of the automatic segmentation output. Without this, the reported 96% accuracy and 0.81 F1 measure the combined system on VerSe19 only; the claim that the approach is intrinsically explainable and transferable to new scans is not yet supported by a demonstrated measurement layer.
  3. [Section 3, Table 1] The headline comparison to DenseNet rests on a one-point accuracy gap (96% versus 95%) with identical sensitivity (91%) and F1 scores of 0.81 versus 0.77. No confidence intervals, bootstrap estimates, or significance tests are reported, and the test-set size after exclusions is not given. It is therefore impossible to rule out chance as an explanation for the difference. At minimum, the authors should report confidence intervals for all metrics and a paired statistical test (for example, McNemar's test for vertebra-level binary decisions, or a cluster bootstrap by patient to account for within-patient correlations).
  4. [Section 3, Benchmark details] The comparison between the neurosymbolic method and the deep-learning baselines is not fully apples-to-apples as described. The text states that the DL models are trained by taking advantage of 'the available vertebral body segmentations' to mask inputs, whereas the proposed method uses automatic TotalSegmentator segmentations. If the DL baselines use ground-truth segmentations while the proposed method uses automatic ones, this asymmetry should be stated explicitly and its effect discussed. In addition, the exclusion criteria (vertebrae with foreign material, scans with only one annotated vertebra) are described qualitatively, but the number of excluded scans and vertebrae and the per-split counts are not reported. Uneven exclusions across train, validation, and test splits could bias the comparison and should be quantified.
minor comments (6)
  1. [Abstract and Section 3] The text contains the typo 'VFC' in the rule description; it should read 'VCF'.
  2. [Eq. (3)] The notation uses C for both the central section of the target vertebra and the central section of the reference vertebra; using C_ref for the latter would eliminate ambiguity.
  3. [Section 2.2] The phrase 'the 3 rules with the highest stratification power' is not defined; please specify the selection criterion, the total number of candidate rules generated by the tree ensemble, and how the final three were chosen.
  4. [Section 3, Dataset] The paper does not state how the four Genant grades are binarized into positive and negative VCF labels; please clarify whether mild fractures are considered negative and whether moderate and severe are both positive.
  5. [References] Reference [35] appears to be mis-cited for the Youden J statistic; the listed author and title do not match a 2008 work on Youden's index and should be corrected.
  6. [Table 1] The ResNeXt unmasked row shows sensitivity 0.97 with a much lower precision of 0.45 and F1 of 0.62; a brief comment on this class-imbalance behavior would help readers interpret the baseline results.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the rule set is a fitted transparent model evaluated on a held-out VerSe19 test set, not a prediction equivalent to its training input.

full rationale

The paper's derivation chain is not circular. The height-map extraction in Section 2.1 converts CT volumes into anatomical height ratios using TotalSegmentator, k-means surface detection, mesh reorientation, and grid projection; none of these steps assumes the VCF label or the final decision rule. The rule set in Eq. 3 is explicitly the output of RuleFit with GradientBoosting and LASSO (Section 2.2): the thresholds 0.91 and 0.81 and the coefficient vector are estimated on the training split and then evaluated on the VerSe19 test split. This is a transparent fitted classifier, not a first-principles derivation, so its test-set performance is independent evidence rather than a tautology. The abstract's phrase 'predefined 2-rule set' is imprecise because the thresholds are learned, but that is a framing issue, not a circular step. The cited prior work from the same group (refs 16, 17, 18, 21) is related interpretability research and is not load-bearing for the central quantitative claim. The main limitations, including the lack of validation of segmentation and reorientation specifically on fractured vertebrae and the learned nature of the thresholds, are correctness or generalizability risks, not circularity.

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

The analysis rests on standard external tools and clinical assumptions rather than new physical or architectural entities. Free parameters are the fitted rule thresholds and selected design choices; axioms are the reliability of segmentation, reorientation, and the Genant grading labels.

free parameters (6)
  • Anterior-posterior ratio threshold = 0.91
    Chosen by RuleFit on the training split; presented as part of the rule set but is a fitted threshold, not a clinical constant (Eq. 3).
  • Central-to-reference ratio threshold = 0.81
    Also selected by RuleFit/LASSO on training data (Eq. 3), again presented as a rule threshold.
  • Rule coefficients in Eq. 3 = 1.49471001, 0.36870275, -4.10884354
    Estimated by a sparse linear model with LASSO; these weights define the positive and negative decision and are fit to data.
  • k-means cluster count = 6
    Hand-chosen to identify main surfaces for mesh reorientation; no sensitivity analysis is reported (Section 2.1).
  • Number of anatomical ROIs = 7
    The seven regions of interest are defined by the authors without an ablation or clinical justification beyond general anatomy (Section 2.1).
  • Number of rules used in linear model = 3
    RuleFit's top three rules by stratification power are selected; the selection criterion and stability are not detailed (Section 2.2, Eq. 2).
assumptions (4)
  • domain assumption TotalSegmentator provides vertebral body segmentations accurate enough for height-ratio fracture detection.
    Section 2.1: 'the vertebral bodies are segmented using TotalSegmentator'; segmentation errors propagate directly into height maps, but no per-case validation on fractured vertebrae is reported.
  • domain assumption The mesh reorientation procedure yields a consistent local coordinate system across patients.
    Section 2.1 describes k-means on surface normals and alignment of posterior and inferior surfaces, but provides no validation that the inferred planes are anatomically correct for deformed or fractured vertebrae.
  • domain assumption Genant semiquantitative grading in the public VerSe19 ground truth is a correct and reliable reference standard.
    Section 3 uses the public Genant grades as the ground truth; label noise in fracture grading would directly affect the reported accuracy and sensitivity.
  • domain assumption Height loss in anterior or central vertebral body regions is the relevant morphological signature of VCF.
    The rule set in Eq. 3 relies on this clinical assumption; fracture morphologies not captured by height ratios would be missed.

how reviews work

0 comments
Cite this review

Pith. "Pith review of An Intrinsically Explainable Approach to Detecting Vertebral Compression Fractures in CT Scans via Neurosymbolic Modeling." pith.science (2026). https://pith.science/paper/DGQPIWQ7

@misc{pith2026241217258,
  author       = {Pith},
  title        = {Pith review of: An Intrinsically Explainable Approach to Detecting Vertebral Compression Fractures in CT Scans via Neurosymbolic Modeling},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/DGQPIWQ7}},
  note         = {Machine review of arXiv:2412.17258}
}
read the original abstract

Vertebral compression fractures (VCFs) are a common and potentially serious consequence of osteoporosis. Yet, they often remain undiagnosed. Opportunistic screening, which involves automated analysis of medical imaging data acquired primarily for other purposes, is a cost-effective method to identify undiagnosed VCFs. In high-stakes scenarios like opportunistic medical diagnosis, model interpretability is a key factor for the adoption of AI recommendations. Rule-based methods are inherently explainable and closely align with clinical guidelines, but they are not immediately applicable to high-dimensional data such as CT scans. To address this gap, we introduce a neurosymbolic approach for VCF detection in CT volumes. The proposed model combines deep learning (DL) for vertebral segmentation with a shape-based algorithm (SBA) that analyzes vertebral height distributions in salient anatomical regions. This allows for the definition of a rule set over the height distributions to detect VCFs. Evaluation of VerSe19 dataset shows that our method achieves an accuracy of 96% and a sensitivity of 91% in VCF detection. In comparison, a black box model, DenseNet, achieved an accuracy of 95% and sensitivity of 91% in the same dataset. Our results demonstrate that our intrinsically explainable approach can match or surpass the performance of black box deep neural networks while providing additional insights into why a prediction was made. This transparency can enhance clinician's trust thus, supporting more informed decision-making in VCF diagnosis and treatment planning.

Figures

Figures reproduced from arXiv: 2412.17258 by the authors.

Figure 1
Figure 1. (a) The height map extraction process relies on vertebral body seg￾mentation with TotalSegmentator [28] to obtain height maps in a consistent local coordinate system. (b) Drawing on real diagnostic decision-making, we define 7 regions of the vertebral body to inform the severity rating. 2 METHODS 2.1 Data processing and parameter extraction overview This method aims to provide an interpretable decision-making pipeli… view at source ↗
Figure 2
Figure 2. (a) Our neurosymbolic reasoning strategy evaluates the relative height of the vertebral body in specific regions to identify cases of VCF. (b) Example model output. A common failure mode occurs when the vertebral body’s shape suggests a compression fracture, even if one is not labeled. Black-box model benchmark We trained two backbone deep learning mod￾els, ResNeXt50 and DenseNet, as a benchmark in our study. For ea… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

36 extracted references · 29 canonical work pages

  1. [1]

    Prevention and management of osteoporosis

    James A. Simon et al. “Prevention and management of osteoporosis”. In: Clin. Cornerstone 5 (Jan. 2003), S5–S12. issn: 1098-3597. doi: 10.1016/ S1098-3597(03)90042-1

    work page 2003

  2. [2]

    Evaluation and Management of Vertebral Com- pression Fractures

    Daniela Alexandru et al. “Evaluation and Management of Vertebral Com- pression Fractures”. In: Permanente Journal 16.4 (2012), p. 46. doi: 10. 7812/tpp/12-037

    work page 2012

  3. [3]

    Clinical Characteristics of Elderly People with Osteoporotic Vertebral Compression Fracture Based on a 12-Year Single- Center Experience in Korea

    Seung-Kwan Lee et al. “Clinical Characteristics of Elderly People with Osteoporotic Vertebral Compression Fracture Based on a 12-Year Single- Center Experience in Korea”. In: Geriatrics 7.6 (Oct. 2022), p. 123. issn: 2308-3417. doi: 10.3390/geriatrics7060123

    work page doi:10.3390/geriatrics7060123 2022

  4. [4]

    Diagnosis and Management of Vertebral Compression Fracture

    Daniel Alsoof et al. “Diagnosis and Management of Vertebral Compression Fracture”. In: Am. J. Med. 135.7 (July 2022), pp. 815–821. issn: 0002-

    work page 2022

  5. [5]

    Percutaneous vertebroplasty for pain relief and spinal stabilization

    J. D. Barr et al. “Percutaneous vertebroplasty for pain relief and spinal stabilization”. In: Spine 25.8 (Apr. 2000), pp. 923–928. issn: 0362-2436. doi: 10.1097/00007632-200004150-00005. eprint: 10767803

    work page doi:10.1097/00007632-200004150-00005 2000

  6. [6]

    Diagnosis and management of vertebral com- pression fractures

    JASON McCARTHY et al. “Diagnosis and management of vertebral com- pression fractures”. In: American family physician 94.1 (2016), pp. 44–50

    work page 2016

  7. [7]

    Vertebral compression fractures: a review of current management and multimodal therapy

    Cyrus C Wong et al. “Vertebral compression fractures: a review of current management and multimodal therapy”. In: Journal of multidisciplinary healthcare (2013), pp. 205–214

    work page 2013

  8. [8]

    Opportunistic Screening Techniques for Analysis of CT Scans

    Oliver Chaudry Klaus Engelke et al. “Opportunistic Screening Techniques for Analysis of CT Scans”. In:National Library of Medicine (NLM) (2023)

    work page 2023

Show all 36 references
  1. [9]

    In silico simulation: a key enabling technology for next-generation intelligent surgical systems

    Benjamin D. Killeen et al. “In silico simulation: a key enabling technology for next-generation intelligent surgical systems”. In: Prog. Biomed. Eng. 5.3 (May 2023), p. 032001. issn: 2516-1091. doi: 10.1088/2516- 1091/ acd28b

    2023 doi

  2. [10]

    MRI Cross-Modality Image-to-Image Translation

    Qianye Yang et al. “MRI Cross-Modality Image-to-Image Translation”. In: Sci. Rep. 10.3753 (Feb. 2020), pp. 1–18. issn: 2045-2322. doi: 10 . 1038/s41598-020-60520-6

    2020

  3. [11]

    Compression fractures detection on CT

    Amir Bar et al. “Compression fractures detection on CT”. In: Proceed- ings Volume 10134, Medical Imaging 2017: Computer-Aided Diagnosis . Vol. 10134. SPIE, Mar. 2017, pp. 1036–1043. doi: 10.1117/12.2249635

    2017 doi

  4. [12]

    Opportunistic Identification of Ver- tebral Compression Fractures on CT Scans of the Chest and Abdomen, Using an AI Algorithm, in a Real-Life Setting

    Magnus Grønlund Bendtsen et al. “Opportunistic Identification of Ver- tebral Compression Fractures on CT Scans of the Chest and Abdomen, Using an AI Algorithm, in a Real-Life Setting”. In: Calcif. Tissue Int. 114.5 (May 2024), pp. 468–479. issn: 1432-0827. doi: 10.1007/s00223-...

    2024 doi

  5. [13]

    Opportunistic Screening: Radiology Scientific Expert Panel

    Perry J. Pickhardt et al. “Opportunistic Screening: Radiology Scientific Expert Panel”. In: Radiology (May 2023). url: https : / / pubs . rsna . org/doi/full/10.1148/radiol.222044

    2023 doi

  6. [14]

    Explainable AI in medical imaging: An overview for clinical practitioners – Saliency-based XAI approaches

    Katarzyna Borys et al. “Explainable AI in medical imaging: An overview for clinical practitioners – Saliency-based XAI approaches”. In: Eur. J. Radiol. 162 (May 2023), p. 110787. issn: 0720-048X. doi: 10.1016/j. ejrad.2023.110787

    2023

  7. [15]

    Revisiting the Trustworthiness of Saliency Methods in Radiology AI

    Jiajin Zhang et al. “Revisiting the Trustworthiness of Saliency Methods in Radiology AI”. In: Radiology: Artificial Intelligence 6.1 (Jan. 2024). doi: 10.1148/ryai.220221

    2024 doi

  8. [16]

    Interpretable Severity Scoring of Pelvic Trauma Through Au- tomated Fracture Detection and Bayesian Inference

    Chen H.and Dreizin D.and Gomez C.and Zapaishchykova A.and & Un- berath M. “Interpretable Severity Scoring of Pelvic Trauma Through Au- tomated Fracture Detection and Bayesian Inference.” In: IEEE Transac- tions on Medical Imaging (2024)

    2024

  9. [17]

    Toward automated interpretable AAST grading for blunt splenic injury

    Chen H.and Unberath M. et al. “Toward automated interpretable AAST grading for blunt splenic injury.” In: Emergency radiology (2022)

    2022

  10. [18]

    An interpretable approach to automated sever- ity scoring in pelvic trauma

    Zapaishchykova A. et al. “An interpretable approach to automated sever- ity scoring in pelvic trauma.” In: Medical Image Computing and Com- puter Assisted Intervention–MICCAI (2021). doi: 10.1007/s10140-022- 02099-1

    2021 doi

  11. [19]

    An interpretable Algorithm for uveal melanoma subtyping from whole slide cytology images

    Chen H. et al. “An interpretable Algorithm for uveal melanoma subtyping from whole slide cytology images.” In: (2021)

    2021

  12. [20]

    Explainable AI Enhances Glaucoma Referrals, Yet the Human-AI Team Still Falls Short of the AI Alone

    Gomez C. et al. “Explainable AI Enhances Glaucoma Referrals, Yet the Human-AI Team Still Falls Short of the AI Alone.” In: (May 2024)

    2024

  13. [21]

    Explain- able AI decision support improves accuracy during telehealth strep throat screening

    Gomez C.and Smith B. L.and Zayas A.and Unberath M. et al. “Explain- able AI decision support improves accuracy during telehealth strep throat screening.” In: npj Digital Medicine 7.18 (2024). doi: https://doi.org/ 10.1038/s41746-024-00962-8

    2024 doi

  14. [22]

    Vertebral Body Compression Fractures and Bone Density: Automated Detection and Classification on CT Images

    Joseph E. Burns et al. “Vertebral Body Compression Fractures and Bone Density: Automated Detection and Classification on CT Images”. In: RSNA 284.3 (2017). doi: 10.1148/radiol.2017162100

    2017 doi

  15. [23]

    Automatic detection of osteoporotic vertebral frac- tures in routine thoracic and abdominal MDCT

    Thomas Baum et al. “Automatic detection of osteoporotic vertebral frac- tures in routine thoracic and abdominal MDCT”. In: European Radiology 24.4 (2014). doi: 10.1007/s00330-013-3089-2

    2014 doi

  16. [24]

    Diagnosis of Osteoporotic Vertebral Fractures: Im- portance of Recognition and Description by Radiologists

    Leon Lenchik et al. “Diagnosis of Osteoporotic Vertebral Fractures: Im- portance of Recognition and Description by Radiologists”. In: American Journal of Roentgenology (Oct. 2024)

    2024

  17. [25]

    Spinal Vertebral Fracture Detection and Fracture Level Assessment Based on Deep Learning

    Yuhang Wang et al. “Spinal Vertebral Fracture Detection and Fracture Level Assessment Based on Deep Learning”. In: CMC 79 (2024). doi: 10.32604/cmc.2024.047379

    2024

  18. [26]

    Explainable medical imaging AI needs human-centered design: guidelines and evidence from a systematic review

    H. Chen et al. “Explainable medical imaging AI needs human-centered design: guidelines and evidence from a systematic review.” In: NPJ digital medicine 5.156 (2022). doi: https://doi.org/10.1038/s41746- 022- 00699-2

    2022 doi

  19. [27]

    Recognizing and reporting osteoporotic verte- bral fractures

    Mikayel Grigoryan et al. “Recognizing and reporting osteoporotic verte- bral fractures”. In: PMC (2003)

    2003

  20. [28]

    TotalSegmentator: Robust Segmentation of 104 Anatomic Structures in CT Images

    Jakob Wasserthal et al. “TotalSegmentator: Robust Segmentation of 104 Anatomic Structures in CT Images”. In: Radiology: Artificial Intelligence (July 2023). url: https://pubs.rsna.org/doi/10.1148/ryai.230024

    2023 doi

  21. [29]

    nnU-Net: a self-configuring method for deep learning- based biomedical image segmentation

    Fabian Isensee et al. “nnU-Net: a self-configuring method for deep learning- based biomedical image segmentation”. In: Nat. Methods 18 (Feb. 2021), pp. 203–211. issn: 1548-7105. doi: 10.1038/s41592-020-01008-z

    2021 doi

  22. [30]

    Marching cubes: A high resolution 3D surface construction algorithm

    William E. Lorensen et al. “Marching cubes: A high resolution 3D surface construction algorithm”. In: Proceedings of the 14th Annual Conference on Computer Graphics and Interactive Techniques (SIGGRAPH ’87) 21.4 (1987)

    1987

  23. [31]

    Predictive Learning via Rule Ensembles

    Jerome H. Friedman et al. “Predictive Learning via Rule Ensembles”. In: Annals of Applied Statistics (2008)

    2008

  24. [32]

    VerSe: A Vertebrae Labelling and Segmentation Benchmark for Multi-Detector CT Images

    Anjany Sekuboyina et al. “VerSe: A Vertebrae Labelling and Segmentation Benchmark for Multi-Detector CT Images”. In: Medical Image Analysis (2021)

    2021

  25. [33]

    A Vertebral Segmentation Dataset with Frac- ture Grading

    Maximilian T. L¨ offler et al. “A Vertebral Segmentation Dataset with Frac- ture Grading”. In: radiology: Artificial Intelligence (2020)

    2020

  26. [34]

    VerSe: A Vertebrae Labelling and Segmentation Benchmark for Multi-detector CT Images

    Anjany Sekuboyina et al. “VerSe: A Vertebrae Labelling and Segmentation Benchmark for Multi-detector CT Images”. In: arXiv (Jan. 2020). doi: 10.1016/j.media.2021.102166. eprint: 2001.09193

    2020

  27. [35]

    Youden Index and optimal cut-point estimated from observations affected by a lower limit of detection

    Anjany Sekuboyina et al. “Youden Index and optimal cut-point estimated from observations affected by a lower limit of detection”. In: 2008 (2008)

    2008

  28. [9343]

    doi: 10.1016/j.amjmed.2022.02.035

    2022 doi

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.