Temporal Neural Cellular Automata: Application to modeling of contrast enhancement in breast MRI

Daniel M. Lang; Julia A. Schnabel; Karim Lekadir; Richard Osuala; Rickmer Braren; Veronika Spieker

REVIEW 4 major objections 5 minor 28 references

Temporal Neural Cellular Automata: Application to modeling of contrast enhancement in breast MRI

T0 review · 4 major / 5 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 Eight-second neural-cell steps beat U-Net and diffusion on breast MRI

desk verdict TeNCA's irregular-time training trick is real, but the paper needs to verify its 8-second divisibility assumption and add error bars before the temporal claims hold. read the letter →

arxiv 2506.18720 v1 pith:LLOO5IK2 submitted 2025-06-23 eess.IV cs.CV

Daniel M. Lang , Richard Osuala , Veronika Spieker , Karim Lekadir , Rickmer Braren , Julia A. Schnabel This is my paper · ORCID

classification eess.IVcs.CV

keywords neuralcellularautomatatemporalimagesynthesisdynamiccontrast-enhancedMRIbreastsyntheticcontrastenhancementconsistencylightweightmodelsmulti-centermedicalimaging

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 introduces TeNCA, a temporal neural cellular automaton that learns to evolve a pre-contrast breast MRI into post-contrast phases by treating each update step as eight seconds of physical time. The authors claim that, trained on a combined multi-center dataset of 1,604 cases, TeNCA matches ground-truth post-contrast sequences better than a U-Net and a latent-diffusion baseline under LPIPS, SSIM, MS-SSIM, and PSNR, while using only about 13,000 parameters. They further claim the model is temporally stable: it does not degrade in later phases the way the baselines do, and its intermediate states evolve continuously. This matters because synthetic contrast enhancement could shorten breast MRI protocols and remove contrast-agent injections, but only if generated sequences are trustworthy over time. The paper argues that conditioning the loss at every acquisition time forces the automaton to learn physiologically plausible contrast kinetics rather than a single static output.

What carries the argument

The core object is the TeNCA transition function $F$, defined by $S_{t+1} = F(S_t) = S_t + (\partial S/\partial t)\,\Delta t$ with $\Delta t = 8$ seconds, so the number of update steps to a phase is fixed by its acquisition time. Perception uses two learnable 3x3 convolutional kernels for neighbor communication plus an identity kernel for the cell's own state, all feeding a two-layer MLP; the state has 24 channels, one visible (the image) and the rest hidden. The mechanism that carries the argument is the loss in Eq. (3), computed at every time $t_i$ where ground truth exists, which forces the hidden state to encode a continuous transition rather than just the final contrast phase.

What would settle it

Train or evaluate TeNCA on a cohort whose acquisition times differ from the training distribution, then compare a held-out middle time point (for example at 90 seconds) against the model's state after 90/8 steps; if per-phase image metrics at the held-out time are substantially worse than at the training times, or if protocol-stratified metrics show the advantage disappears for protocols with long or irregular spacing, the temporal-consistency claim is falsified.

Watch

Extended reading notes

Core claim

The central claim is that a neural cellular automaton whose update rule is tied to a fixed time increment $\Delta t = 8$ seconds can be trained on temporally sparse, non-uniformly spaced DCE-MRI acquisitions and produce intermediate states aligned with ground-truth post-contrast images. TeNCA initializes the visible state with the pre-contrast image, runs a shared transition function $F$, and computes the image loss at every step $t_i/\Delta t$ for which a ground-truth acquisition exists. This adaptive sequential loss makes the temporal dimension part of the learning objective, not just a means to reach a final image. On the combined MAMA-MIA and Duke test data, TeNCA improves over both reference methods on image-level metrics while using far fewer parameters than the diffusion baseline; the paper attributes the baseline's poor pixel-level scores to hallucination, a known failure mode of diffusion models.

Load-bearing premise

The load-bearing premise is that an 8-second fixed update step maps directly onto real contrast-agent kinetics and that one transition function learned from many centers can represent all protocols; if timing or dynamics vary by protocol in ways the model does not see, the intermediate states are not physiologically meaningful.

Editorial extensions

If this is right

Generating all post-contrast phases from one pre-contrast scan requires no contrast injection and no extra acquisition time, so a reader could inspect several synthetic phases and recover some washout kinetics lost in abbreviated protocols.
Because the model has about 13,000 parameters and uses iterative local convolutions, it can run on modest hardware, making synthetic DCE-MRI feasible in resource-constrained screening settings.
Per-phase metrics show TeNCA maintains accuracy in later post-contrast phases where the U-Net baseline degrades, implying washout behavior is modeled rather than ignored.
A fixed 8-second time step standardizes heterogeneous protocols: the same learned transition function can be queried at any time point, enabling interpolation between acquired phases.

Reading between the lines

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

A testable extension is to withhold one intermediate acquisition time during training and compare TeNCA's state at that time against the held-out ground truth; success would directly validate the physical-time assumption.
If the fixed 8-second step truly encodes physical time, the same training scheme could apply to other dynamic imaging modalities with sparse temporal sampling, such as cine MRI, 4D CT, or perfusion imaging.
The paper does not verify the physiological meaning of TeNCA's hidden channels, so an interesting next step would be probing whether those channels encode contrast arrival or washout rates.
The comparison with diffusion suggests that local, continuity-enforcing inductive biases may matter more than raw generative capacity for medical temporal synthesis; this is an editorial reading, not a claim of the paper.

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.

Desk Editor's Note

read the letter

Let me give you the short version: TeNCA is a genuine contribution to the NCA-for-image-synthesis line, and the paper is honest enough to report that the diffusion baseline wins on distribution metrics. But the paper's central temporal-consistency claim has a gap that the authors don't address, and the evaluation is thinner than the prose suggests.

What's new: the training strategy that applies the NCA loss at irregularly spaced time points, with each update step mapped to Δt=8 seconds. Prior NCA temporal work used uniform or dense sampling; this adaptive-loss approach is a real extension. The model is also strikingly cheap (13k parameters) and the authors evaluate on a diverse multi-center dataset with code released. Credit where due: they report FID/FRD even though CC-Net beats them there, and they explain the pixel-level vs distribution tradeoff.

The soft spots, in order of importance.

First, the Δt=8s assumption. Algorithm 1 conditions the loss only when the accumulated time 'equals t_i'. For that to work, every acquisition time in MAMA-MIA and Duke must be an exact multiple of 8 seconds. The paper never says that. If a t_i is not divisible by 8, that ground-truth phase is silently ignored; if the implementation rounds, the rounding error (up to 4s) is unreported and could matter in early phases where contrast uptake changes fast. This isn't a nitpick; it's the mechanism that grounds the iterative updates in physical time, and it's unverified.

Second, the evaluation. There are no error bars or significance tests on any metric. PSNR is a monotone transform of MSE, which is exactly TeNCA's training loss, so the PSNR comparison is stacked. Distribution metrics favor CC-Net, and the authors' explanation (hallucination) is plausible but not quantified. Temporal consistency is shown via videos and per-phase plots, but the per-phase plots inherit the divisibility problem.

Third, minor: 'prove superiority' overclaims, and the U-Net baseline is a small variant that may not be well tuned.

Overall: the core idea is sound and the paper is readable. The stress-test concern is real and needs to be addressed before the temporal claim is credible. But that's a fixable gap, not a fatal flaw.

Who it's for: anyone working on synthetic contrast enhancement or NCA-based temporal modeling. It deserves a serious referee. I'd recommend sending it to peer review with a request that the authors clarify the time-step handling and add statistical validation.

Referee Report

4 major / 5 minor

Summary. The paper introduces TeNCA (Temporal Neural Cellular Automata), an NCA-based method for synthesizing post-contrast breast MRI phases from a pre-contrast image. The key idea is to reinterpret each NCA update step as a fixed physical time increment (Δt = 8 s) and to compute the training loss at intermediate states corresponding to observed acquisition times, thereby learning a continuous, temporally conditioned contrast-enhancement trajectory. The method is trained on a combined MAMA-MIA and Duke breast MRI dataset and compared with a U-Net baseline and the CC-Net latent diffusion model. The authors report that TeNCA achieves the best LPIPS, SSIM, MS-SSIM, and PSNR on the test set with about 13,000 parameters, while CC-Net achieves better FID and FRD. They further claim temporal stability and sequential consistency based on per-phase metric plots and qualitative videos.

Significance. If fully substantiated, the paper would make a useful contribution: a parameter-efficient architecture for synthetic contrast enhancement with explicit temporal conditioning, trained and evaluated on a multi-center public dataset, with code provided. The strengths include the use of external baselines, per-phase evaluation in Figure 3, the very small parameter count of TeNCA, and the transparent reporting that CC-Net attains better distribution-level metrics. These features make the work of interest to the medical image synthesis community. However, the current evidence falls short of the paper's strong 'surpassing current SOTA' and 'prove superiority' claims due to missing statistical validation, an under-specified temporal alignment mechanism, and a temporal-consistency evaluation that is only qualitative.

major comments (4)

[Section 3, Eq. (3) and Algorithm 1] The loss is computed only at update step t_i / Δt (Eq. 3), and Algorithm 1 applies the loss only when 't equals t_i'. With Δt = 8 s, this mechanism is well-defined only if every acquisition time in the combined training set is an exact integer multiple of 8 s. The paper neither verifies this condition nor states a rounding rule or tolerance for the equality check. If some observed t_i are not multiples of 8 s, the corresponding ground-truth phase is silently excluded from training, and the per-phase results in Figure 3 are not produced by the claimed conditioning mechanism. If rounding is used, the resulting timing error (up to 4 s) is unreported and could be significant for early phases where contrast uptake is fast. This is a load-bearing internal-consistency issue for the paper's central temporal-modeling claim and must be clarified.
[Table 1 and Figure 3] The reported improvements over the U-Net baseline are small (LPIPS 0.12 vs. 0.13, SSIM 0.89 vs. 0.88, MS-SSIM 0.93 vs. 0.92, PSNR 32.26 dB vs. 31.93 dB), yet no error bars, confidence intervals, or significance tests are provided. The reader cannot tell whether these differences are consistent across the 300 test cases or driven by a few outliers. In addition, PSNR is a monotone transform of MSE, which is TeNCA's training loss, so the PSNR advantage is partly by construction; the paper should report metrics that are not direct transforms of the training objectives, or at least discuss this confound. Without statistical support, the claim of 'surpassing current SOTA' is not established.
[Section 4.4, 'Temporal stability' and 'Sequential consistency'] The temporal consistency claim is supported only by per-phase mean metric plots (Figure 3) and qualitative example videos provided online. Per-phase fidelity averaged over the test set is not a temporal consistency metric: a model could have good per-phase metrics while producing non-physiological jumps between phases. To support 'prove TeNCA's superiority in learning temporal patterns', the paper needs a quantitative temporal measure, such as the smoothness of predicted intensity trajectories, the displacement or variance between consecutive predicted phases, or a comparison of kinetic curves against the ground-truth enhancement dynamics. Without such a metric, the central temporal-consistency claim is not evidenced.
[Table 1 and Section 4.4] CC-Net outperforms TeNCA on both distribution measures (FID 21.28 vs. 27.83 and FRD 20.00 vs. 48.68). The authors explain this as CC-Net 'hallucinating' realistic-looking images, which is plausible, but it also means that the overall 'superiority' claim is not uniformly true. The paper should temper the abstract and conclusion wording ('surpassing current SOTA performance', 'prove TeNCA's superiority') to reflect that TeNCA is better on pixel- and perception-level fidelity while CC-Net is better on distribution-level realism, and should discuss which of these properties is more clinically relevant for synthetic contrast-enhanced MRI.

minor comments (5)

[Section 1] There are grammatical errors in the first paragraph: 'To address this limitations' and 'Wide adoption ... is hindered by its high costs and lengthy acquisition times' should be corrected.
[Section 4.2] The final hyperparameters for TeNCA (number of update steps N_steps, training epochs, learning rate, batch size) are not reported. The code link helps, but for reproducibility the paper should state the chosen values, especially N_steps, which directly determines the maximum simulated time horizon.
[Section 4.1] The dataset description does not specify how many Duke cases were used, how they were selected, or whether the random validation split was stratified by center or protocol. Since the paper emphasizes diversity across protocols, this information is needed to assess the generality claims.
[Algorithm 1] The indentation of the loops in Algorithm 1 is ambiguous: it is unclear whether the inner loop over i is intended to run inside the update-step loop or only after it. The pseudocode should be restructured so that the scope of the loss accumulation is unambiguous.
[Author affiliations and abstract] There is a typo in the affiliation list ('Klinkum' should be 'Klinikum'), and the abstract uses 'prove' in 'prove TeNCA's superiority in learning temporal patterns'; given the current evidence, a more measured phrase such as 'provide evidence for' would be more appropriate.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: TeNCA's central claims are measured against held-out external benchmarks and independent baselines.

full rationale

The paper's derivation chain is self-contained. The core method defines an NCA update rule and a supervised loss (Eq. 3) at intermediate states conditioned at acquisition times; this is a training objective, not a prediction masquerading as a result. The central performance claims are evaluated on the held-out MAMA-MIA test set with standard image and distribution metrics, compared against a U-Net baseline and CC-Net. CC-Net is the authors' prior work, but it is used as a retrained baseline for comparison, not as a load-bearing premise or as a fitted input; no result is forced by construction. No uniqueness theorem is imported, and no ansatz is smuggled in via self-citation. The only substantive concern is whether all acquisition times are exact multiples of Δt=8s, which would affect the validity of the temporal conditioning mechanism; that is an implementation-correctness question, not a circularity, because the conclusion does not reduce to an assumption of the method. Therefore no circular step is identified.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The central results rely on modeling assumptions about how time and image appearance are linked: the fixed step duration, the ability of a local cell process to represent contrast kinetics, the pooling of heterogeneous protocols, and the intensity normalization. No new physical entities are introduced, and the only hand-set values are standard hyperparameters and the temporal resolution.

free parameters (3)

Temporal resolution Delta_t = 8 seconds
Chosen by hand as the NCA step duration; maps acquisition times to update steps and controls the number of intermediate states. Not derived from physiology or data.
NCA state channels = 24
Padding of the input image to a channel size of 24, plus an MLP hidden size of 128; set by empirical hyperparameter search.
Maximum update steps = 128 (from 1024 s / 8 s)
The number of update steps is implicitly set by the maximum acquisition time considered (1024 seconds).

assumptions (4)

domain assumption A single NCA transition function with local 3x3 perception and an MLP update can model contrast enhancement dynamics in breast MRI.
Core modeling assumption of the method; the network is expected to learn the underlying kinetics from paired examples without an explicit physiological model.
ad hoc to paper The acquisition time t_i maps linearly to update steps as t_i / Delta_t with Delta_t = 8 s.
Used in Section 3 and Algorithm 1 to decide when to compute loss; assumes fixed step duration equals real time and that all sites share the same time base.
domain assumption Multi-center data with different protocols can be pooled and modeled with time as the only conditioning variable.
Section 4.1 combines MAMA-MIA and Duke with varying numbers of post-contrast phases; no protocol or center identifier is given to the model, so protocol differences must be captured implicitly.
domain assumption Linear intensity rescaling based on pre-contrast percentiles preserves the contrast uptake signal.
Section 4.1 normalizes intensities to [0,1] using the 0.02 and 99.98 percentiles of the pre-contrast image; this could distort uptake magnitude across patients if pre-contrast baselines vary.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Temporal Neural Cellular Automata: Application to modeling of contrast enhancement in breast MRI." pith.science (2026). https://pith.science/paper/LLOO5IK2

@misc{pith2026250618720,
  author       = {Pith},
  title        = {Pith review of: Temporal Neural Cellular Automata: Application to modeling of contrast enhancement in breast MRI},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LLOO5IK2}},
  note         = {Machine review of arXiv:2506.18720}
}

read the original abstract

Synthetic contrast enhancement offers fast image acquisition and eliminates the need for intravenous injection of contrast agent. This is particularly beneficial for breast imaging, where long acquisition times and high cost are significantly limiting the applicability of magnetic resonance imaging (MRI) as a widespread screening modality. Recent studies have demonstrated the feasibility of synthetic contrast generation. However, current state-of-the-art (SOTA) methods lack sufficient measures for consistent temporal evolution. Neural cellular automata (NCA) offer a robust and lightweight architecture to model evolving patterns between neighboring cells or pixels. In this work we introduce TeNCA (Temporal Neural Cellular Automata), which extends and further refines NCAs to effectively model temporally sparse, non-uniformly sampled imaging data. To achieve this, we advance the training strategy by enabling adaptive loss computation and define the iterative nature of the method to resemble a physical progression in time. This conditions the model to learn a physiologically plausible evolution of contrast enhancement. We rigorously train and test TeNCA on a diverse breast MRI dataset and demonstrate its effectiveness, surpassing the performance of existing methods in generation of images that align with ground truth post-contrast sequences.

Figures

Figures reproduced from arXiv: 2506.18720 by the authors.

**Figure 1.** Overview of TeNCA. For each step, our NCA backbone transitions the images gradually to reflect the next time point. During Training, intermediate states are conditioned at all time points with a given ground truth DCE-MRI available. clinical center might capture only two post-contrast sequences shortly after injection, whereas another may employ five sequences to also illustrate contrast washout. To provide a comp… view at source ↗

**Figure 2.** Example test set results, involving a (predicted) post-contrast image and a subtraction between the pre- and the post-contrast image, highlighting contrast uptake. TeNCA successfully models detailed structures, outperforming the U-Net. Additionally, it avoids hallucination of artifacts, a limitation evident in the first example of CC-Net [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗

**Figure 3.** Mean image metric values for the test set across all post-contrast phases. TeNCA maintains consistent performance throughout all phases, while other methods exhibit a noticeable decline in later phases. Sequential consistency Example videos illustrating the temporal evolution of contrast enhancement are provided online2 . Notably, TeNCA exhibits the best sequential consistency, characterized by a continuous evoluti… view at source ↗

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

28 extracted references · 23 canonical work pages

[1]

Advances in Neural Information Pro- cessing Systems 37, 134614–134644 (2025)

Aithal, S.K., Maini, P., Lipton, Z., Kolter, J.Z.: Understanding hallucinations in diffusion models through mode interpolation. Advances in Neural Information Pro- cessing Systems 37, 134614–134644 (2025)

work page 2025
[2]

In: International Conference on Medical Image Computing and Computer-Assisted Intervention

Deutges, M., Sadafi, A., Navab, N., Marr, C.: Neural Cellular Automata for Lightweight, Robust and Explainable Classification of White Blood Cell Images. In: International Conference on Medical Image Computing and Computer-Assisted Intervention. pp. 693–702. Springer (2024)

work page 2024
[3]

In: International Conference on Medical Image Computing and Computer-Assisted Intervention

Elbatel, M., Kamnitsas, K., Li, X.: An Organism Starts with a Single Pix-Cell: A Neural Cellular Diffusion for High-Resolution Image Synthesis. In: International Conference on Medical Image Computing and Computer-Assisted Intervention. pp. 656–666. Springer (2024)

work page 2024
[4]

Radiologic Clinics of North America 59(1), 85 (2020)

Gao, Y., Reig, B., Heacock, L., Bennett, D.L., Heller, S.L., Moy, L.: Magnetic res- onance imaging in screening of breast cancer. Radiologic Clinics of North America 59(1), 85 (2020)

work page 2020
[5]

arXiv preprint arXiv:2406.13844 (2024) 10 D.M

Garrucho, L., Reidel, C.A., Kushibar, K., Joshi, S., Osuala, R., Tsirikoglou, A., Bobowicz, M., del Riego, J., Catanese, A., Gwoździewicz, K., et al.: MAMA-MIA: A Large-Scale Multi-Center Breast Cancer DCE-MRI Benchmark Dataset with Expert Segmentations. arXiv preprint arXiv:2406.13844 (2024) 10 D.M. Lang et al

arXiv 2024
[6]

Physical Review E 100(3), 032402 (2019)

Gilpin, W.: Cellular automata as convolutional neural networks. Physical Review E 100(3), 032402 (2019)

work page 2019
[7]

Ad- vances in neural information processing systems30 (2017)

Heusel, M., Ramsauer, H., Unterthiner, T., Nessler, B., Hochreiter, S.: GANs trained by a two time-scale update rule converge to a local nash equilibrium. Ad- vances in neural information processing systems30 (2017)

work page 2017
[8]

In: International Conference on In- formation Processing in Medical Imaging

Kalkhof, J., González, C., Mukhopadhyay, A.: Med-NCA: Robust and lightweight segmentation with neural cellular automata. In: International Conference on In- formation Processing in Medical Imaging. pp. 705–716. Springer (2023)

work page 2023

Show all 28 references

[9]

In: International Conference on Medical Image Computing and Computer-Assisted Intervention

Kalkhof, J., Mukhopadhyay, A.: M3D-NCA: Robust 3D Segmentation with Built- In Quality Control. In: International Conference on Medical Image Computing and Computer-Assisted Intervention. pp. 169–178. Springer (2023)

2023
[10]

IEEE journal of translational engineering in health and medicine 11, 32–43 (2022)

Kim, E., Cho, H.H., Kwon, J., Oh, Y.T., Ko, E.S., Park, H.: Tumor-attentive segmentation-guided GAN for synthesizing breast contrast-enhanced mri without contrast agents. IEEE journal of translational engineering in health and medicine 11, 32–43 (2022)

2022
[11]

Journal of Clinical Oncology32(22), 2304–2310 (2014)

Kuhl, C.K., Schrading, S., Strobel, K., Schild, H.H., Hilgers, R.D., Bieling, H.B.: Abbreviated breast magnetic resonance imaging (MRI): first postcontrast sub- tracted images and maximum-intensity projection—a novel approach to breast cancer screening with MRI. Journal of Cli...

2014
[12]

Leithner, D., Moy, L., Morris, E.A., Marino, M.A., Helbich, T.H., Pinker, K.: Ab- breviated MRI of the breast: does it provide value? Journal of Magnetic Resonance Imaging 49(7), e85–e100 (2019)

2019
[13]

bmj388 (2025)

Lekadir, K., Frangi, A.F., Porras, A.R., Glocker, B., Cintas, C., Langlotz, C.P., Weicken, E., Asselbergs, F.W., Prior, F., Collins, G.S., et al.: FUTURE-AI: Inter- national consensus guideline for trustworthy and deployable artificial intelligence in healthcare. bmj388 (2025)

2025
[14]

In: 2021 IEEE 18th International Symposium on Biomedical Imaging (ISBI)

Manzanera, O.E.M., Ellis, S., Baltatzis, V., Nair, A., Le Folgoc, L., Desai, S., Glocker, B., Schnabel, J.A.: Patient-specific 3d cellular automata nodule growth synthesis in lung cancer without the need of external data. In: 2021 IEEE 18th International Symposium on Biomedica...

2021
[15]

arXiv preprint arXiv:2501.02447 (2025)

Mittal, A., Kalkhof, J., Mukhopadhyay, A., Bhavsar, A.: Medsegdiffnca: Diffusion models with neural cellular automata for skin lesion segmentation. arXiv preprint arXiv:2501.02447 (2025)

2025 arXiv
[16]

Distill 5(2), e23 (2020)

Mordvintsev, A., Randazzo, E., Niklasson, E., Levin, M.: Growing neural cellular automata. Distill 5(2), e23 (2020)

2020
[17]

Radiology 307(3), e222211 (2023)

Müller-Franzes, G., Huck, L., Tayebi Arasteh, S., Khader, F., Han, T., Schulz, V., Dethlefsen, E., Kather, J.N., Nebelung, S., Nolte, T., et al.: Using machine learning to reduce the need for contrast agents in breast MRI through synthetic images. Radiology 307(3), e222211 (2023)

2023
[18]

arXiv preprint arXiv:2409.18872 (2024)

Osuala, R., Joshi, S., Tsirikoglou, A., Garrucho, L., Pinaya, W.H., Lang, D.M., Schnabel, J.A., Diaz, O., Lekadir, K.: Simulating Dynamic Tumor Contrast En- hancement in Breast MRI using Conditional Generative Adversarial Networks. arXiv preprint arXiv:2409.18872 (2024)

2024 arXiv
[19]

In: International Con- ference on Medical Image Computing and Computer-Assisted Intervention

Osuala, R., Lang, D.M., Verma, P., Joshi, S., Tsirikoglou, A., Skorupko, G., Kushibar, K., Garrucho, L., Pinaya, W.H., Diaz, O., et al.: Towards learning con- trast kinetics with multi-condition latent diffusion models. In: International Con- ference on Medical Image Computing...

2024
[20]

In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

Pajouheshgar, E., Xu, Y., Zhang, T., Süsstrunk, S.: DyNCA: Real-time dynamic texture synthesis using neural cellular automata. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 20742–20751 (2023) Temporal Neural Cellular Automata 11

2023
[21]

arXiv preprint arXiv:2410.22265 (2024)

Ranem, A., Kalkhof, J., Mukhopadhyay, A.: NCA-Morph: Medical Image Regis- tration with Neural Cellular Automata. arXiv preprint arXiv:2410.22265 (2024)

2024 arXiv
[22]

PLOS Computational Biology 20(4), e1011589 (2024)

Richardson, A.D., Antal, T., Blythe, R.A., Schumacher, L.J.: Learning spatio- temporal patterns with Neural Cellular Automata. PLOS Computational Biology 20(4), e1011589 (2024)

2024
[23]

In: Medical image computing and computer-assisted intervention–MICCAI 2015: 18th international conference, Munich, Germany, Oc- tober 5-9, 2015, proceedings, part III 18

Ronneberger, O., Fischer, P., Brox, T.: U-Net: Convolutional networks for biomed- ical image segmentation. In: Medical image computing and computer-assisted intervention–MICCAI 2015: 18th international conference, Munich, Germany, Oc- tober 5-9, 2015, proceedings, part III 18....

2015
[24]

British journal of cancer 119(4), 508–516 (2018)

Saha, A., Harowicz, M.R., Grimm, L.J., Kim, C.E., Ghate, S.V., Walsh, R., Mazurowski, M.A.: A machine learning approach to radiogenomics of breast can- cer: a study of 922 subjects and 529 DCE-MRI features. British journal of cancer 119(4), 508–516 (2018)

2018
[25]

In: Deep Breast Workshop on AI and Imaging for Diagnostic and Treatment Challenges in Breast Care

Schreiter, H., Eberle, J., Kapsner, L.A., Hadler, D., Ohlmeyer, S., Erber, R., Emons, J., Laun, F.B., Uder, M., Wenkel, E., et al.: Virtual dynamic contrast enhanced breast MRI using 2D U-Net Architectures. In: Deep Breast Workshop on AI and Imaging for Diagnostic and Treatmen...

2024
[26]

In: The Thrity-Seventh Asilomar Conference on Signals, Systems & Computers, 2003

Wang, Z., Simoncelli, E.P., Bovik, A.C.: Multiscale structural similarity for im- age quality assessment. In: The Thrity-Seventh Asilomar Conference on Signals, Systems & Computers, 2003. vol. 2, pp. 1398–1402. Ieee (2003)

2003
[27]

In: Proceedings of the IEEE/CVF International Conference on Computer Vision

Zhang, L., Rao, A., Agrawala, M.: Adding conditional control to text-to-image diffusion models. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. pp. 3836–3847 (2023)

2023
[28]

In: Proceedings of the IEEE conference on computer vision and pattern recognition

Zhang, R., Isola, P., Efros, A.A., Shechtman, E., Wang, O.: The unreasonable effectiveness of deep features as a perceptual metric. In: Proceedings of the IEEE conference on computer vision and pattern recognition. pp. 586–595 (2018)

2018

Pith tools

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

[1] [1]

Advances in Neural Information Pro- cessing Systems 37, 134614–134644 (2025)

Aithal, S.K., Maini, P., Lipton, Z., Kolter, J.Z.: Understanding hallucinations in diffusion models through mode interpolation. Advances in Neural Information Pro- cessing Systems 37, 134614–134644 (2025)

work page 2025

[2] [2]

In: International Conference on Medical Image Computing and Computer-Assisted Intervention

Deutges, M., Sadafi, A., Navab, N., Marr, C.: Neural Cellular Automata for Lightweight, Robust and Explainable Classification of White Blood Cell Images. In: International Conference on Medical Image Computing and Computer-Assisted Intervention. pp. 693–702. Springer (2024)

work page 2024

[3] [3]

In: International Conference on Medical Image Computing and Computer-Assisted Intervention

Elbatel, M., Kamnitsas, K., Li, X.: An Organism Starts with a Single Pix-Cell: A Neural Cellular Diffusion for High-Resolution Image Synthesis. In: International Conference on Medical Image Computing and Computer-Assisted Intervention. pp. 656–666. Springer (2024)

work page 2024

[4] [4]

Radiologic Clinics of North America 59(1), 85 (2020)

Gao, Y., Reig, B., Heacock, L., Bennett, D.L., Heller, S.L., Moy, L.: Magnetic res- onance imaging in screening of breast cancer. Radiologic Clinics of North America 59(1), 85 (2020)

work page 2020

[5] [5]

arXiv preprint arXiv:2406.13844 (2024) 10 D.M

Garrucho, L., Reidel, C.A., Kushibar, K., Joshi, S., Osuala, R., Tsirikoglou, A., Bobowicz, M., del Riego, J., Catanese, A., Gwoździewicz, K., et al.: MAMA-MIA: A Large-Scale Multi-Center Breast Cancer DCE-MRI Benchmark Dataset with Expert Segmentations. arXiv preprint arXiv:2406.13844 (2024) 10 D.M. Lang et al

arXiv 2024

[6] [6]

Physical Review E 100(3), 032402 (2019)

Gilpin, W.: Cellular automata as convolutional neural networks. Physical Review E 100(3), 032402 (2019)

work page 2019

[7] [7]

Ad- vances in neural information processing systems30 (2017)

Heusel, M., Ramsauer, H., Unterthiner, T., Nessler, B., Hochreiter, S.: GANs trained by a two time-scale update rule converge to a local nash equilibrium. Ad- vances in neural information processing systems30 (2017)

work page 2017

[8] [8]

In: International Conference on In- formation Processing in Medical Imaging

Kalkhof, J., González, C., Mukhopadhyay, A.: Med-NCA: Robust and lightweight segmentation with neural cellular automata. In: International Conference on In- formation Processing in Medical Imaging. pp. 705–716. Springer (2023)

work page 2023