Pith. sign in

REVIEW 5 major objections 5 minor 40 references

Anatomically Consistent Cross-Contrast Super-Resolution of Anisotropic Brain T2w MRI

T0 review · 5 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read The paper claims that anisotropic T2w brain MRI can be super-resolved to near-isotropic quality using only the patient's isotropic T1c as anatomical reference, with no isotropic T2w ground truth, improving downstream tumor segmentation.

desk verdict The paper has a load-bearing framing problem: the acquired T2w is never used at inference, so it is really T1c-to-T2w synthesis, and the ablation claim is contradicted by its own table; still worth a careful referee. read the letter →

arxiv 2608.08401 v1 pith:L3OQDRFV submitted 2026-08-09 cs.CV

Mengqi Shen , Haicheng Wang , Meghna Trivedi , Tony J. Wang , Yuanguang Xu , Yingyan Zeng , Yading Yuan This is my paper · ORCID
classification cs.CV
keywords MRIsuper-resolutionanisotropicreconstructioncross-contrastguidanceprojectionoptimizationgenerativeadversarialnetworksinter-planeconsistencytumorsegmentationzero-shottransfer
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 tries to establish that the missing inter-plane detail in an anisotropic T2-weighted brain MRI can be recovered by borrowing anatomy from the same patient's isotropic T1c image, without ever needing an isotropic T2w scan as ground truth. This matters because clinical T2w is routinely acquired with thick slices, and the resulting blur and stair-stepping obscure small structures and weaken 3D analysis for tumor segmentation and radiotherapy planning. The proposed method trains a view-independent patch generator on axial T1c-to-T2w slices, applies it along all three anatomical axes, and fuses the three outputs by a projection-based optimization that balances per-view fidelity against inter-plane self-consistency. On a metastasis cohort the fused volume improved mean-label tumor segmentation Dice from 0.330 to 0.465, and on a glioma cohort without retraining from 0.473 to 0.563, supporting the claim.

What carries the argument

The load-bearing mechanism is the projection-based fusion objective. It combines a per-axis inter-plane variance penalty $L_{\text{proj}}(P)=\sum_a \|(P-\bar{P}^a)\odot M\|_F^2$ with a view-fidelity term $L_{\text{dir}}(P)=\sum_a \|(P-P^*_a)\odot M\|_F^2$, optimized over the masked volume $P$ with Adam. Because $L_{\text{proj}}$ vanishes only if the volume is constant along every axis, it acts as a strong 3D coherence prior rather than a physical model of slice acquisition. Around this sits the view-independent patched generator VIP-GAN: a U-Net with mask conditioning, coarsened skip connections, and no deepest skip, trained on 128x128 axial patches without resampling, which lets one frozen checkpoint serve all three anatomical views.

What would settle it

Acquire thin-slice isotropic T2w for a subset of the same subjects, run VIPP-SR on the thick-slice versions, and compare the fused output against the real isotropic T2w in inter-plane high-frequency bands; if the optimization step moves the fused volume away from the real isotropic T2w relative to the simple three-view average, the self-consistency prior is not recovering anatomy.

Watch

Extended reading notes

Core claim

VIPP-SR's central claim is that cross-view agreement can substitute for missing isotropic supervision. A single generator, trained only on high-resolution axial T1c-to-T2w pairs, is applied unchanged to axial, coronal, and sagittal slices of the T1c volume; each resulting volume is sharp in its own plane and unreliable elsewhere. The fused estimate is obtained by minimizing a weighted sum of a self-consistency term, which penalizes the squared distance of every masked voxel from its slice mean along each axis, and a data-fidelity term, which keeps the result close to the three view-specific volumes. The paper reports that this fusion, without any isotropic T2w ground truth, improves downstream single-channel tumor segmentation over the acquired anisotropic T2w baseline on both cohorts, and attributes the improvement mainly to the inter-plane self-consistency term.

Load-bearing premise

The load-bearing premise is that penalizing how far each voxel is from the average of its slice along every axis is a valid stand-in for anatomical consistency; a real brain is not constant from slice to slice, so in practice this term is a smoothing prior, not a measured relationship between anisotropic and isotropic anatomy.

Editorial extensions

If this is right

  • Retrospective cohorts could receive near-isotropic T2w without rescanning, as long as a co-registered isotropic T1c exists.
  • The same frozen generator can serve a new tumor population at zero training cost; only the per-subject voxel-space fusion step is run at deployment.
  • Segmentation gains concentrate on small focal structures such as tumor, enhancing tissue, and resection cavity, so tasks that depend on those structures benefit most.
  • The method avoids iterative backpropagation through a large generative prior; fusion is voxel-space and linear-time per update, keeping per-case compute low.
  • Future super-resolution designs should treat explicit cross-view 3D consistency as a first-class objective rather than an implicit side effect of slice-wise generation.

Reading between the lines

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

  • Editor's inference: because the self-consistency term penalizes deviation from slice-constancy, it likely smooths genuine inter-plane high frequencies; testing an alpha/beta sweep on a dataset with true isotropic T2w would reveal whether very thin structures are over-smoothed.
  • Editor's inference: the reported ablation numbers leave room for another reading, since the fidelity-only variant reaches comparable or higher mean Dice on the training cohort; a systematic weight sweep would clarify whether self-consistency itself or multi-view averaging drives the gain.
  • Editor's inference: replacing the slice-constancy penalty with a slice-profile or motion-aware forward model could be tested within the same framework, and the paper's edge-consistency and downstream-segmentation metrics would quantify the change.
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

5 major / 5 minor

Summary. The paper proposes VIPP-SR, a two-stage framework for producing an isotropic T2-weighted (T2w) volume from an anisotropic T2w acquisition using a co-registered isotropic T1c volume. Stage 1 trains a view-independent GAN (VIP-GAN) on axial T1c-to-T2w patch pairs. Stage 2 applies the frozen generator along axial, coronal, and sagittal views of the T1c volume and fuses the three view-specific outputs through a projection-based optimization that balances a slice-mean self-consistency penalty (Eq. 10) against a view-fidelity penalty (Eq. 11). The method is evaluated on BraTS-MET (held-out test set) and BraTS-GLI (zero-shot cross-cohort) using image-quality metrics (Canny-Dice, SSIM, Lap-RMSE, HFEN) and downstream nnU-Net segmentation. The paper claims that VIPP-SR improves mean-label Dice over the real anisotropic T2w baseline (from 0.330 to 0.465 on BraTS-MET and from 0.473 to 0.563 on BraTS-GLI) and that ablation studies identify inter-plane self-consistency as the main source of the gain.

Significance. If the central claims were supported, the paper would offer a practically valuable approach for retrospective neuro-oncology cohorts where isotropic T2w ground truth is unavailable. The cross-cohort, zero-shot evaluation is a strength, and the intention to release code is commendable. However, the paper's framing as super-resolution is not supported by the algorithm, which does not use the acquired T2w volume at inference. The reported ablation contradicts the headline claim, and the primary structural-fidelity metric is circular. These issues are load-bearing, so the significance of the contribution as stated cannot be accepted without fundamental revision.

major comments (5)
  1. [Sec. 3.4-3.5, Algorithm 1, Eq. (12)] The patient's anisotropic T2w volume Y is never used in the main VIPP-SR objective. Eq. (12) optimizes L*(P) = alpha*L_proj + beta*L_dir, where L_dir depends only on the T1c-derived view volumes P*_hd, P*_wd, P*_hw. The only term containing Y, L_aux, is introduced in Sec. 3.5.3 and used solely in the v3 ablation. Algorithm 1 confirms that Stage 2 operates on X, M, and the frozen generator, with no reference to Y. Consequently, the final output is a pure function of the T1c volume, making the method a T1c-to-T2w synthesis rather than a super-resolution of the patient's acquired T2w. This directly contradicts the framing in Sec. 1 and the abstract, and it confounds the reported segmentation gains because the output carries all of the T1c anatomy.
  2. [Table 3, Sec. 5.2, Abstract] The abstract and Sec. 5.2 claim that ablation studies identify inter-plane self-consistency as the main source of the downstream gain. Table 3 shows the opposite: the fidelity-only ablation v1 achieves mean-label Dice 0.484, higher than VIPP-SR's 0.465, and v1 also gives the highest edema Dice (0.529). The text explicitly acknowledges that 'v1 gives the highest mean-label Dice (0.484) and edema Dice (0.529)' but still concludes that self-consistency is the main contributor. This is a direct contradiction between the reported numbers and the paper's central attribution claim, so the ablation conclusion is not supported.
  3. [Eq. (13), Sec. 4.2, Sec. 5.1] The structural-fidelity metric Canny-Dice computes the Dice coefficient between the synthetic T2w edge map and the T1c edge map. Since T1c is the generator's input, high scores can be obtained by copying T1c edges into the output, and the metric is circular as evidence of T2w fidelity. The paper's claims in Sec. 5.1 that the synthetic volumes 'agree with the subject's own T1c anatomy' and that high Canny-Dice indicates recovered inter-plane structure are therefore not established; an explicit ablation or a non-circular reference (e.g., an independent modality or manual annotations) is needed.
  4. [Sec. 3.5.1, Eq. (10)] The projection term L_proj penalizes the squared distance of each voxel from its slice mean along each of the three axes. The paper itself notes that a perfectly self-consistent volume is constant along each axis, i.e., constant over the whole brain mask. This is a strong spatial-flatness prior, not a physically motivated model of anatomical inter-plane consistency for a real brain T2w volume, which has substantial structure along all axes. Consequently, the interpretation of the fusion stage as enforcing 'anatomical consistency' is not supported, and the mechanism by which it might help segmentation remains unclear.
  5. [Sec. 5.2, Tables 3 and 4] No baseline is reported in which the isotropic T1c volume alone is used as the input to the downstream nnU-Net segmentation. Because VIPP-SR's output is a function of T1c and the trained generator, the observed improvement over the real T2w baseline could be due to the T1c information alone rather than to any property of the proposed reconstruction. Without a T1c-only channel baseline, the central segmentation claim is confounded and cannot be interpreted as evidence that the reconstructed T2w is superior to the acquired anisotropic T2w.
minor comments (5)
  1. [Eq. (1) and Sec. 3.4] There are several apparent OCR artifacts in the typesetting, for example 'bY M=1 ≈Y iso M=1' in Eq. (1) and 'P ∗ ∗hd' in Sec. 3.4; these should be corrected before publication.
  2. [Sec. 3.2.2, Eqs. (2)-(3)] When H or W is not a multiple of P, the stride formulas produce non-integer values; the implementation detail of how non-integer strides are rounded or otherwise handled should be stated explicitly.
  3. [Sec. 4.2] The image-quality metrics are computed on the single largest-mask-area slice from each anatomical plane rather than over the full volume; this limited sampling should be stated as a potential source of bias.
  4. [Sec. 5.2] The statement that the segmentation results are 'single-fold nnU-Net results without cross-validation, we treat the gains as indicative rather than formally significant' is an important caveat that should also appear in the abstract, which currently states the results 'validate' the method.
  5. [References [14] and [22]] References [14] and [22] appear to be the same paper (identical title and venue); this duplication should be resolved.

Circularity Check

2 steps flagged · score 6.0 of 10

VIPP-SR's Stage 2 objective (Eq. 12 + Algorithm 1) never uses the patient's anisotropic T2w, so the 'super-resolution' output is by construction T1c-to-T2w synthesis; its Canny-Dice structural metric also references the T1c input, making a core validation self-referential.

  1. self definitional [Algorithm 1 (Stage 2), Sec. 3.5.3, Sec. 3.5.4 Eq. (12)]
    "Stage 2 below is run per subject. Reuses Gθ unchanged. ... for each view v∈{hw,hd,wd} do Slice X along axis v ... Apply Gθ per patch ... recombine into P∗v. ... P←P−η∇_P L⋆(P) via Adam, using (10)–(11). (Algorithm 1) An optional anchor against the patient's anisotropic T2w, Laux(P)=∥(P−Y)⊙M∥2F ... is used only in the ablation variant v3 below. (Sec. 3.5.3)"

    The inference objective is L⋆(P)=αL_proj(P)+βL_dir(P) (Eq. 12). L_dir (Eq. 11) measures fidelity to P∗_a, the three view-specific generator outputs produced by applying Gθ to the T1c volume X; L_proj (Eq. 10) is a slice-mean flatness penalty. No term in L⋆ involves the acquired T2w Y. The only Y-dependent term, L_aux, is explicitly optional and used only in ablation v3, not in the reported VIPP-SR rows. Therefore the returned bY=P is, by Algorithm 1, a pure function of X, M, and the trained Gθ. The abstract's claim that VIPP-SR 'restores the inter-plane resolution of an existing anisotropic T2w volume' is thus not a derivation from Y but T1c-to-T2w synthesis relabeled as super-resolution.

  2. self definitional [Sec. 4.2(ii), Eq. (13)]
    "The reference is the T1c edge map, a different modality from the output. The metric is Canny-Dice [36], defined as the Dice similarity coefficient between the synthetic-T2w and T1c Canny edge maps. ... CannyDice= 2|E_I ∩ E_T1c| / (|E_I|+|E_T1c|)."

    The synthetic T2w being scored is generated from the T1c volume: Stage 2 slices X, applies Gθ, and fuses the resulting views. Comparing its Canny edges to the T1c edge map therefore measures input-output edge agreement, not agreement with an independent anatomical reference. The paper itself notes in Sec. 3.3.1 that full-resolution skips can allow 'direct copying of T1c edges into the output,' confirming the mechanism. High Canny-Dice is expected for any generator that preserves T1c structure, so this metric cannot validate recovery of the anisotropic T2w's inter-plane detail; it is a self-referential consistency check between the input and the output rather than an external structural fidelity test.

full rationale

The central circularity is in the method's definition: although the paper frames VIPP-SR as super-resolving the patient's anisotropic T2w, the per-subject inference objective (Eq. 12) contains only the projection term and the T1c-derived view-fidelity term; the acquired T2w Y is used only in the optional auxiliary term L_aux, which is confined to the v3 ablation. Algorithm 1 therefore returns a volume that is, by construction, a function of the T1c volume and the trained generator. This makes the headline comparison against the real-T2w segmentation baseline a comparison of T1c-derived synthesis against the T2w input, rather than evidence of T2w super-resolution; the lack of a T1c-only segmentation baseline leaves the confound uncontrolled. The image-quality evaluation is similarly self-referential: Canny-Dice compares the synthetic T2w's edges to the T1c edge map, which is the generator's input, so high scores are partly forced by input-output edge fidelity rather than by recovery of T2w structure. The downstream segmentation benchmark itself is external and not circular, which is why the score is not higher; however, its interpretation as validation of SR is undermined by the T1c-only construction. No self-citation chain or imported uniqueness theorem is used, and the paper is otherwise self-contained. An additional consistency problem, not itself circularity, is that Table 3 shows the fidelity-only ablation v1 (mean Dice 0.484) outperforming the full VIPP-SR (0.465), which undercuts the abstract's claim that inter-plane self-consistency is the main source of the gain. On balance, the two exhibited reductions-by-construction justify a partial circularity score of 6.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The central hyperparameters are chosen on validation rather than derived. The method also assumes spatial alignment with isotropic T1c, orientation-invariant patch-local generation, cross-cohort transfer, and that the projection term in Eq. (10) is a valid anatomical consistency prior. The last assumption is questionable because the term enforces slice constancy. No new physical entities are introduced.

free parameters (5)
  • lambda_rec = 100
    Reconstruction loss weight in Eq. (8), chosen by validation (Sec. 4.3).
  • alpha = 0.7
    Weight on L_proj in Eq. (12), chosen on a validation split (Sec. 3.5.4).
  • beta = 0.3
    Weight on L_dir in Eq. (12), chosen on a validation split (Sec. 3.5.4).
  • patch_size P = 128
    Patch size in Sec. 3.2.2, hand-chosen to balance receptive field and batch parallelism.
  • v3 weights alpha_prime, beta_prime, gamma = 0.9, 0.3, 1
    Weights for the ablation-only anchor variant v3 (Sec. 3.5.4), chosen by hand.
assumptions (4)
  • domain assumption A co-registered, isotropic T1c volume is available for every subject.
    The method's input X is described as co-registered and isotropic (Sec. 3.1); if T1c is anisotropic or misregistered, the anatomical guidance and view transfer are invalid.
  • domain assumption A generator trained on axial T1c-to-T2w patches transfers zero-shot to coronal and sagittal views.
    Sec. 3.3 and 3.4 assume patch-local, orientation-free mapping based on fixed patching and skip modifications; no direct validation of view transfer is provided before fusion.
  • ad hoc to paper The projection penalty L_proj (Eq. 10) is a valid model of anatomical self-consistency.
    The paper itself notes perfect self-consistency under this term means the volume is constant along each axis, i.e., a flatness prior rather than a realistic inter-plane consistency condition.
  • domain assumption T1c-to-T2w correspondence learned on BraTS-MET transfers to BraTS-GLI.
    Zero-shot cross-cohort evaluation assumes the mapping is stable across tumor types and protocols; the residual SNFH gap in Table 4 suggests a partial violation.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Anatomically Consistent Cross-Contrast Super-Resolution of Anisotropic Brain T2w MRI." pith.science (2026). https://pith.science/paper/L3OQDRFV

@misc{pith2026260808401,
  author       = {Pith},
  title        = {Pith review of: Anatomically Consistent Cross-Contrast Super-Resolution of Anisotropic Brain T2w MRI},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/L3OQDRFV}},
  note         = {Machine review of arXiv:2608.08401}
}
read the original abstract

T2-weighted (T2w) brain MRI provides fluid-sensitive soft-tissue contrast that is important for neuro-oncology and radiotherapy planning. However, T2w scans are acquired with anisotropic voxels and appear blurred or stair-stepped on coronal and sagittal views, which obscures small structures and weakens any downstream 3D analysis. We propose VIPP-SR (View-Independent Patched Projection Super-Resolution), a cross-contrast guided super-resolution framework that restores the inter-plane resolution of an existing anisotropic T2w volume without an isotropic ground-truth T2w. VIPP-SR first trains a view-independent patched generator (VIP-GAN) to learn local T1c-to-T2w anatomical correspondence from high-resolution axial slices. The trained generator is then applied to axial, coronal, and sagittal views of the T1c volume to generate three orthogonal T2w estimates. Shape-preserving patching and deepest-skip removal reduce view-specific shortcuts, thereby constraining the generator to learn patch-local representations and enabling the zero-shot inter-plane transfer. Central to VIPP-SR, a projection-based optimization then enforces anatomical consistency across the three view-specific volumes, fusing them by balancing inter-plane self-consistency against per-view data fidelity. The generator is trained on BraTS-MET and evaluated on both the held-out BraTS-MET testing set and the BraTS-GLI cohort without retraining, assessing the cross-cohort generalizability. The results validate that VIPP-SR improves downstream segmentation over the real anisotropic T2w baseline, raising mean-label Dice from 0.330 to 0.465 on BraTS-MET and, zero-shot, from 0.473 to 0.563 on BraTS-GLI and ablation studies identify inter-plane self-consistency as the main source of the gain.

Figures

Figures reproduced from arXiv: 2608.08401 by the authors.

Figure 1
Figure 1. Overview of the proposed VIPP-SR pipeline. Stage 1 trains a view-independent patched GAN on axial T1c→T2w pairs and applies the frozen generator along the three orthogonal axes. Stage 2 fuses the resulting view-specific volumes with an axial-aligned projection objective to pro￾duce the final isotropic T2w estimate. the view-specific outputs as complementary constraints on one unknown 3D volume. It balances self-cons… view at source ↗
Figure 2
Figure 2. View-dependent outputs of the trained VIP-GAN. Top row: patient’s co-registered T1c and the original T2w. Bottom row: the three view-specific reconstructions P ∗ hw, P∗ hd, P∗ wd obtained by applying the same trained generator weights to axial, coronal and sagittal views in stage 1. Each P ∗ v is sharpest in its own slicing plane (boxes), highlighted by red boxes. Ladv = BCE(Dϕ(x, Gθ(˜x)), 1), (7) Lrec = [PITH_FULL… view at source ↗
Figure 3
Figure 3. Structural edge comparison on a representative BraTS-GLI coronal reformat (case 00307, [PITH_FULL_IMAGE:figures/full_fig_p017_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

40 extracted references · 37 canonical work pages

  1. [1]

    Brain metastases resection cavity radio—surgery based on t2-weighted mri: Technique assessment

    Achiraya Teyateeti, Paul D Brown, Anita Mahajan, Nadia N Laack, and Bruce E Pollock. Brain metastases resection cavity radio—surgery based on t2-weighted mri: Technique assessment. Journal of neuro-oncology, 148(1):89–95, 2020

    work page 2020

  2. [2]

    Advanced imaging techniques for radiotherapy planning of gliomas.Cancers, 13(5):1063, 2021

    Antonella Castellano, Michele Bailo, Francesco Cicone, Luciano Carideo, Natale Quartuccio, Pietro Mortini, Andrea Falini, Giuseppe Lucio Cascini, and Giuseppe Minniti. Advanced imaging techniques for radiotherapy planning of gliomas.Cancers, 13(5):1063, 2021

    work page 2021

  3. [3]

    Sean CL Deoni, Jonathan O’Muircheartaigh, Emil Ljungberg, Mathew Huentelman, and Steven CR Williams. Simultaneous high-resolution t2-weighted imaging and quantitative t 2 mapping at low magnetic field strengths using a multiple te and multi-orientation acquisition approach.Magnetic Resonance in Medicine, 88(3):1273–1281, 2022

    work page 2022

  4. [4]

    Gradient-guided isotropic mri reconstruction from anisotropic acquisitions.IEEE transactions on computational imaging, 7:1240–1253, 2021

    Yao Sui, Onur Afacan, Camilo Jaimes, Ali Gholipour, and Simon K Warfield. Gradient-guided isotropic mri reconstruction from anisotropic acquisitions.IEEE transactions on computational imaging, 7:1240–1253, 2021

    work page 2021

  5. [5]

    Alena Uus, Sara Neves Silva, Jordina Aviles Verdera, Kelly Payette, Megan Hall, Kathleen Colford, AyshaLuis, HelenaSousa, ZihanNing, ThomasRoberts, etal. Scanner-basedreal-time three-dimensional brain+ body slice-to-volume reconstruction for t2-weighted 0.55-t low-field fetal magnetic resonance imaging.Pediatric Radiology, pages 1–14, 2025

    work page 2025

  6. [6]

    3d isotropic super-resolution prostate mri using generative adversarial networks and unpaired multiplane slices.Journal of Digital Imaging, 34:1199–1208, 2021

    Yucheng Liu, Yulin Liu, Rami Vanguri, Daniel Litwiller, Michael Liu, Hao-Yun Hsu, Richard Ha, Hiram Shaish, and Sachin Jambawalikar. 3d isotropic super-resolution prostate mri using generative adversarial networks and unpaired multiplane slices.Journal of Digital Imaging, 34:1199–1208, 2021

    work page 2021

  7. [7]

    A counterexample to a conjecture of Larman and Rogers on sets avoiding distance 1

    Joseph Paul Cohen, Luke Luck, and Sina Honari. Distribution matching losses can hallucinate features in medical image translation.arXiv preprint arXiv:1808.07299, 2018

    work page Pith review arXiv 2018

  8. [8]

    Greve, Koen Van Leemput, Bruce Fischl, and Adrian V

    Juan Eugenio Iglesias, Benjamin Billot, Douglas N. Greve, Koen Van Leemput, Bruce Fischl, and Adrian V. Dalca. Synthsr: resolution-agnostic mri synthesis for clinical neuroimaging. Science Advances, 2023. 19

    work page 2023

Show all 40 references
  1. [9]

    Shuiping Jiang, Hui Xue, Andrew Glover, Mary Rutherford, Daniel Rueckert, and Joseph V. Hajnal. Mri of moving subjects using multislice snapshot images with volume reconstruction (svr): application to fetal, neonatal, and adult brain studies.IEEE Transactions on Medical Imagin...

    2007

  2. [10]

    Estroff, and Simon K

    Ali Gholipour, Judy A. Estroff, and Simon K. Warfield. Robust super-resolution volume recon- struction from slice acquisitions: application to fetal brain mri.IEEE Transactions on Medical Imaging, 29:1739–1758, 2010

    2010

  3. [11]

    Christodoulou, Yao Xie, Zhiming Zhou, and Debiao Li

    Yao Chen, Feng Shi, Aggelos G. Christodoulou, Yao Xie, Zhiming Zhou, and Debiao Li. Effi- cient and accurate mri super-resolution using a generative adversarial network and 3d multi- level densely connected network. InMedical Image Computing and Computer Assisted Inter- ventio...

    2018

  4. [12]

    Multi-contrast super-resolution mri through a progressive network.IEEE transac- tions on medical imaging, 39(9):2738–2749, 2020

    Qing Lyu, Hongming Shan, Cole Steber, Corbin Helis, Chris Whitlow, Michael Chan, and Ge Wang. Multi-contrast super-resolution mri through a progressive network.IEEE transac- tions on medical imaging, 39(9):2738–2749, 2020

    2020

  5. [13]

    Adaptive latent diffusion model for 3d medical image to image translation: Multi-modal magnetic resonance imaging study

    Jonghun Kim and Hyunjin Park. Adaptive latent diffusion model for 3d medical image to image translation: Multi-modal magnetic resonance imaging study. InProceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, pages 7604–7613, 2024

    2024

  6. [14]

    High-resolution 3d mri with deep gener- ative networks via novel slice-profile transformation super-resolution.IEEE Access, 11:95022– 95036, 2023

    Shuo Lin, Shiqian Lin, Bin Wang, Jiacheng Liu, et al. High-resolution 3d mri with deep gener- ative networks via novel slice-profile transformation super-resolution.IEEE Access, 11:95022– 95036, 2023

    2023

  7. [15]

    Smore: a self-supervised anti-aliasing and super-resolution algorithm for mri using deep learning.IEEE transactions on medical imaging, 40(3):805–817, 2020

    Can Zhao, Blake E Dewey, Dzung L Pham, Peter A Calabresi, Daniel S Reich, and Jerry L Prince. Smore: a self-supervised anti-aliasing and super-resolution algorithm for mri using deep learning.IEEE transactions on medical imaging, 40(3):805–817, 2020

    2020

  8. [16]

    Simple: Simultaneous multi-plane self-supervised learning for isotropic mri restoration from anisotropic data.arXiv preprint arXiv:2408.13065, 2024

    Rotem Benisty, Yevgenia Shteynman, Moshe Porat, Anat Illivitzki, and Moti Freiman. Simple: Simultaneous multi-plane self-supervised learning for isotropic mri restoration from anisotropic data.arXiv preprint arXiv:2408.13065, 2024

    2024 arXiv

  9. [17]

    Inversesr: 3d brain mri super-resolution using a latent diffusion model

    Jueqi Wang, Jacob Levman, Walter Hugo Lopez Pinaya, Petru-Daniel Tudosiu, M Jorge Car- doso, and Razvan Marinescu. Inversesr: 3d brain mri super-resolution using a latent diffusion model. InInternational conference on medical image computing and computer-assisted inter- ventio...

    2023

  10. [18]

    Graham, Ronald Gray, et al

    Walter Hugo Lopez Pinaya, Mark S. Graham, Ronald Gray, et al. Brain imaging generation with latent diffusion models.arXiv preprint arXiv:2209.07162, 2022

    2022 arXiv

  11. [19]

    Registration-based approach for reconstruc- tion of high-resolution in utero fetal mr brain images.Academic radiology, 13(9):1072–1081, 2006

    Francois Rousseau, Orit A Glenn, Bistra Iordanova, Claudia Rodriguez-Carranza, Daniel B Vi- gneron, James A Barkovich, and Colin Studholme. Registration-based approach for reconstruc- tion of high-resolution in utero fetal mr brain images.Academic radiology, 13(9):1072–1081, 2006

    2006

  12. [20]

    Multiscale brain mri super-resolution using deep 3d convolutional networks.Computerized Medical Imaging and Graphics, 77:101647, 2019

    Cong Pham, Alexandre Ducournau, Ronan Fablet, and François Rousseau. Multiscale brain mri super-resolution using deep 3d convolutional networks.Computerized Medical Imaging and Graphics, 77:101647, 2019. 20

    2019

  13. [21]

    An arbitrary scale super-resolution approach for 3d mr images via implicit neural representation

    Qing Wu, Yuwei Li, Yawen Sun, Yan Zhou, Hongjiang Wei, Jingyi Yu, and Yuyao Zhang. An arbitrary scale super-resolution approach for 3d mr images via implicit neural representation. IEEE Journal of Biomedical and Health Informatics, 27(2):1004–1015, 2022

    2022

  14. [22]

    High-resolution 3d mri with deep generative networks via novel slice-profile transformation super-resolution.IEEE Access, 11:95022–95036, 2023

    Jiahao Lin, QI Miao, Chuthaporn Surawech, Steven S Raman, Kai Zhao, Holden H Wu, and Kyunghyun Sung. High-resolution 3d mri with deep generative networks via novel slice-profile transformation super-resolution.IEEE Access, 11:95022–95036, 2023

    2023

  15. [23]

    An enhanced generative adversarial network for 3d brain mri super-resolution

    Jiancong Wang, Benjamin Weyn, Patrick Buehler, et al. An enhanced generative adversarial network for 3d brain mri super-resolution. InProceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), pages 712–721, 2020

    2020

  16. [24]

    Simple: Simultaneous multi-plane self-supervised learning for isotropic mri restoration from anisotropic data

    Rotem Benisty, Yevgenia Shteynman, Moshe Porat, Anat Ilivitzki, and Moti Freiman. Simple: Simultaneous multi-plane self-supervised learning for isotropic mri restoration from anisotropic data. InInternational Conference on Medical Image Computing and Computer-Assisted Inter- v...

    2025

  17. [25]

    Remedios, Pratik Ranjitkar, Moritz Zaiss, Anagha Deshmane, Govind Chavhan, Daniel Rueckert, Michael Ünser, et al

    Samuel W. Remedios, Pratik Ranjitkar, Moritz Zaiss, Anagha Deshmane, Govind Chavhan, Daniel Rueckert, Michael Ünser, et al. Eclare: Efficient cross-planar learning for anisotropic resolution enhancement.arXiv preprint arXiv:2503.11787, 2025

    2025

  18. [26]

    Blind super-resolution of 3d mri via unsupervised domain transformation.IEEE Journal of Biomedical and Health Informatics, 27(3):1409–1418, 2022

    Hexiang Zhou, Yawen Huang, Yuexiang Li, Yi Zhou, and Yefeng Zheng. Blind super-resolution of 3d mri via unsupervised domain transformation.IEEE Journal of Biomedical and Health Informatics, 27(3):1409–1418, 2022

    2022

  19. [27]

    Mamba-enhanced diffusion model for perception-aware blind super-resolution of magnetic resonance imaging.IEEE Journal of Biomedical and Health Informatics, 2025

    Xiaoqiang Zhao, Xiaodong Yang, and Zhaoyang Song. Mamba-enhanced diffusion model for perception-aware blind super-resolution of magnetic resonance imaging.IEEE Journal of Biomedical and Health Informatics, 2025

    2025

  20. [28]

    Multi- modal mr synthesis via modality-invariant latent representation.IEEE transactions on medical imaging, 37(3):803–814, 2017

    Agisilaos Chartsias, Thomas Joyce, Mario Valerio Giuffrida, and Sotirios A Tsaftaris. Multi- modal mr synthesis via modality-invariant latent representation.IEEE transactions on medical imaging, 37(3):803–814, 2017

    2017

  21. [29]

    Image synthesis in multi-contrast mri with conditional generative adversarial networks

    Salman UH Dar, Mahmut Yurt, Levent Karacan, Aykut Erdem, Erkut Erdem, and Tolga Cukur. Image synthesis in multi-contrast mri with conditional generative adversarial networks. IEEE transactions on medical imaging, 38(10):2375–2388, 2019

    2019

  22. [30]

    Mahmut Yurt, Salman U. H. Dar, Aykut Erdem, Erkut Erdem, Kader K. Oguz, and Tolga Cukur. mustgan: Multi-stream generative adversarial networks for mr image synthesis.Medical Image Analysis, 70:101944, 2021

    2021

  23. [31]

    Ea- gans: Edge-aware generative adversarial networks for cross-modality mr image synthesis.IEEE Transactions on Medical Imaging, 38(7):1750–1762, 2019

    Biting Yu, Luping Zhou, Lei Wang, Yinghuan Shi, Jurgen Fripp, and Pierrick Bourgeat. Ea- gans: Edge-aware generative adversarial networks for cross-modality mr image synthesis.IEEE Transactions on Medical Imaging, 38(7):1750–1762, 2019

    2019

  24. [32]

    Unpaired volumetric harmonization of brain mri with conditional latent diffusion.Medical Image Analysis, 107:103849, 2026

    Mengqi Wu, Minhui Yu, Shuaiming Jing, Pew-Thian Yap, Zhengwu Zhang, and Mingxia Liu. Unpaired volumetric harmonization of brain mri with conditional latent diffusion.Medical Image Analysis, 107:103849, 2026

    2026

  25. [33]

    Jhaveri, and Thippa Reddy Gadekallu

    Yinghua Li, Weiao Hao, Hao Zeng, Longguang Wang, Jian Xu, Sidheswar Routray, Rutvij H. Jhaveri, and Thippa Reddy Gadekallu. Cross-scale texture supplementation for reference-based medical image super-resolution.IEEE Journal of Biomedical and Health Informatics, 2025. Early Access. 21

    2025

  26. [34]

    The 2024 brain tumor segmentation (brats) challenge: glioma segmentation on post-treatment mri.arXiv preprint arXiv:2405.18368, 2024

    Maria Correia de Verdier, Rachit Saluja, Louis Gagnon, Dominic LaBella, Ujjwall Baid, Nourel Hoda Tahon, Martha Foltyn-Dumitru, Jikai Zhang, Maram Alafif, Saif Baig, et al. The 2024 brain tumor segmentation (brats) challenge: glioma segmentation on post-treatment mri.arXiv pre...

    2024 arXiv

  27. [35]

    Image quality assessment: fromerrorvisibilitytostructuralsimilarity.IEEE Transactions on Image Processing, 13(4):600– 612, 2004

    Zhou Wang, Alan C Bovik, Hamid R Sheikh, and Eero P Simoncelli. Image quality assessment: fromerrorvisibilitytostructuralsimilarity.IEEE Transactions on Image Processing, 13(4):600– 612, 2004

    2004

  28. [36]

    A computational approach to edge detection.IEEE Transactions on Pattern Analysis and Machine Intelligence, PAMI-8(6):679–698, 1986

    John Canny. A computational approach to edge detection.IEEE Transactions on Pattern Analysis and Machine Intelligence, PAMI-8(6):679–698, 1986

    1986

  29. [37]

    Jaeger, Simon A

    Fabian Isensee, Paul F. Jaeger, Simon A. A. Kohl, Jens Petersen, and Klaus H. Maier-Hein. nnu-net: a self-configuring method for deep learning-based biomedical image segmentation. Nature Methods, 18:203–211, 2021

    2021

  30. [38]

    Target-guided diffusion models for unpaired cross-modality medical image translation.IEEE Journal of Biomedical and Health Informatics, 2024

    Yimin Luo, Qinyu Yang, Ziyi Liu, Zenglin Shi, Weimin Huang, Guoyan Zheng, and Jun Cheng. Target-guided diffusion models for unpaired cross-modality medical image translation.IEEE Journal of Biomedical and Health Informatics, 2024

    2024

  31. [39]

    Unpaired image-to-image trans- lation using cycle-consistent adversarial networks

    Jun-Yan Zhu, Taesung Park, Phillip Isola, and Alexei A Efros. Unpaired image-to-image trans- lation using cycle-consistent adversarial networks. InProceedings of the IEEE international conference on computer vision, pages 2223–2232, 2017

    2017

  32. [40]

    Xie et al

    H. Xie et al. Synthesizing high-resolution magnetic resonance imaging using parallel cycle- consistent generative adversarial networks for fast magnetic resonance imaging.Medical Physics, 2022. 22

    2022

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.