No Image, No Problem: End-to-End Multi-Task Cardiac Analysis from Undersampled k-Space

Yundi Zhang; Sevgi Gokce Kafali; Niklas Bubeck; Daniel Rueckert; Jiazhen Pan

cs.CVMar 10, 2026

No Image, No Problem: End-to-End Multi-Task Cardiac Analysis from Undersampled k-Space

Authors: Yundi Zhang, Sevgi Gokce Kafali, Niklas Bubeck, Daniel Rueckert, Jiazhen Pan

Organizations: Chair for AI in Healthcare and Medicine, Technical University of Munich (TUM) and TUM University Hospital, Munich, Germany · Department of Computing, Imperial College London, United Kingdom · Munich Center for Machine Learning, Technical University of Munich, Germany

Abstract

Conventional clinical CMR pipelines rely on a sequential "reconstruct-then-analyze" paradigm, forcing an ill-posed intermediate step that introduces avoidable artifacts and information bottlenecks. This creates a fundamental mathematical paradox: it attempts to recover high-dimensional pixel arrays (i.e., images) from undersampled k-space, rather than directly extracting the low-dimensional physiological labels actually required for diagnosis. To unlock the direct diagnostic potential of k-space, we propose k-MTR (k-space Multi-Task Representation), a k-space representation learning framework that aligns undersampled k-space data and fully-sampled images into a shared semantic manifold. Leveraging a large-scale controlled simulation of 42,000 subjects, k-MTR forces the k-space encoder to restore anatomical information lost to undersampling directly within the latent space, bypassing the explicit inverse problem for downstream analysis. We demonstrate that this latent alignment enables the dense latent space embedded with high-level physiological semantics directly from undersampled frequencies. Across continuous phenotype regression, disease classification, and anatomical segmentation, k-MTR achieves highly competitive performance against state-of-the-art image-domain baselines. By showcasing that precise spatial geometries and multi-task features can be successfully recovered directly from the k-space representations, k-MTR provides a robust architectural blueprint for task-aware cardiac MRI workflows.

Figures & tables

Figure 1: Overview of the k-MTR framework. (a) $S$ multi-view $2D+t$ slices are tokenized, concatenating image ( $Q_{i}^{u}$ ) and k-space ( $Q_{k}^{u}$ ) tokens across slices for encoder input. (b–d) Training pipeline (single slice shown). (b) Unsupervised masked reconstruction of undersampled k-space and image slices. (c) Contrastive alignment between undersampled k-space ( $T^{u}_{k}$ ) and fully-sampled image ( $T_{i}$ ) embeddings. (d) Fine-tuning the pretrained k-space encoder via lightweight task-specific decoders.

Figure 2: 3D t-SNE visualization of the representations after alignment, colored by ground-truth phenotype groups.

Phenotype	Fully-sampled			Undersampled R=4
Phenotype	RN50	ViT	MAE	RN50 u	RN50 ${}_{k}^{u}$	MAE ${}_{k}^{u}$	k-MTR
LVEDV (mL)	$6.58$	$11.53$	$9.03$	$7.35$	$9.57$	$12.88$	$8.15$
LVSV (mL)	$5.68$	$6.93$	$5.86$	$6.55$	$7.03$	$8.42$	$6.50$
LVEF (%)	$2.95$	$3.91$	$3.20$	$3.30$	$3.58$	$6.09$	$3.14$
LVM (g)	$5.51$	$7.85$	$5.87$	$7.01$	$8.00$	$10.09$	$7.20$
RVEDV (mL)	$9.24$	$13.75$	$9.49$	$10.49$	$11.33$	$14.76$	$10.60$

Table 1: Mean absolute error

\downarrow

for phenotype prediction. Image-based baselines are in gray . RN50: ResNet-50. Best are in dark green , second and third in lighter green .

Disease	Method	AUC-ROC $\uparrow$	F1 Score $\uparrow$	Recall $\uparrow$	Precision $\uparrow$	AP $\uparrow$
CAD (7.4%)	ViT	$0.682$	$0.176$	0.842	$0.098$	$0.126$
	MAE	$\underline{0.708}$	$\underline{0.215}$	$0.200$	0.233	$\underline{0.154}$
	k-MTR (R=4)	0.737	0.282	$\underline{0.500}$	$\underline{0.197}$	0.234
High Blood Pressure (25.8%)	ViT	$0.677$	$0.461$	$0.652$	$\underline{0.357}$	$0.355$
	MAE	$\underline{0.691}$	$\underline{0.469}$	$\underline{0.680}$	0.358	$\underline{0.369}$
	k-MTR (R=4)	0.697	0.474	0.781	$0.340$	0.386

Table 2: Disease classification performance. Fully-sampled image baselines are in gray ; k-MTR with undersampled k-space. Positive class ratios are shown as percentages of the cohort. AP: average precision. Best results are bold , second underlined .

Figure 3: Segmentation and Reconstruction Results. (a) Dice Score and example segmentation maps overlayed on fully-sampled images. k-MTR (R=8) is compared against the fully-sampled upper bound (nnU-Net) and undersampled (R=8) image-based baselines (nnU-Net u , LI-Net). (b) Image reconstructions from undersampled k-space compared to k-GIN.

Figure 4: Ablation of k-MTR’s phenotype prediction accuracy across varying undersampling factors, with 16x prediction failures highlighted in red.

Explore similar work

Robustness of breast lesion segmentation under MRI undersampling improves with k-space-aware deep learning
May 21, 2026Lukas T. Rotkopf, Marco Schlimbach, Julius C. Holzschuh +3Magnetic Resonance Imaging Medical Image Segmentation
Institute for AI in Medicine (IKIM), University Hospital Essen, Essen, Germany · Department of Physics, Technical University Dortmund, Dortmund, Germany · 3Division of Radiology, German Cancer Research Center (DKFZ), Heidelberg, Germany +4
Direct low-field MRI super-resolution using undersampled k-space
Feb 28, 2026Daniel Tweneboah Anyimadu, Mohammed M. Abdelsamea, Ahmed Karam EldalyMagnetic Resonance Imaging Medical Image Super-Resolution
Department of Computer Science, University of Exeter, Exeter, United Kingdom · UCL Hawkes Institute, University College London, London, United Kingdom
K-space Gaussian Representation for Parallel MRI
Jul 29, 2026Yu Guan, Mingyu Hu, Jiale Hu +3Accelerated MRI Reconstruction Magnetic Resonance Imaging
School of Advanced Manufacturing and the School of Information Engineering, Nanchang University, Nanchang 330031, China · Key Laboratory of Biomedical Imaging Science and System, Chinese Academy of Sciences, Shenzhen 518055, China · School of Information Engineering, Nanchang University, Nanchang 330031, China +1

May 21, 2026J/K move · Enter open · S save

Robustness of breast lesion segmentation under MRI undersampling improves with k-space-aware deep learning

Lukas T. Rotkopf, Marco Schlimbach, Julius C. Holzschuh, Heinz-Peter Schlemmer+2

Institute for AI in Medicine (IKIM), University Hospital Essen, Essen, Germany · Department of Physics, Technical University Dortmund, Dortmund, Germany · 3Division of Radiology, German Cancer Research Center (DKFZ), Heidelberg, Germany+4

Magnetic Resonance Imaging Medical Image Segmentation 3D Medical Image Segmentation

Open paper AskPDF arXiv

Purpose: To assess whether breast lesion segmentation can be learned directly from acquired MRI k-space, and whether doing so improves robustness when data are accelerated or noisy. Materials and Methods: This retrospective study used public breast dynamic contrast-enhanced MRI (DCE-MRI) datasets with acquired and synthetic k-space, together with a within-dataset synthetic control. We compared four 3D U-Net variants: a hybrid k-space-to-image model, a native k-space model, and magnitude and complex image-space baselines. Models were evaluated under increasing undersampling and added complex Gaussian k-space noise. The primary outcome was patient-level Dice similarity coefficient under cross-validation, with the hybrid model prespecified as the main comparison against the magnitude image-space baseline. Results: At full sampling, the hybrid and image-space models performed similarly. As acceleration increased, the hybrid model retained substantially more segmentation accuracy and significantly outperformed the magnitude image-space baseline across moderate to high undersampling levels. The same pattern was observed when noise was added directly to k-space: the hybrid model degraded more slowly, whereas the image-space baseline failed under heavier noise. This advantage was reproduced in the within-dataset synthetic control. Feature analysis suggested that the k-space stage and image-space stage played complementary roles, with frequency-domain filtering concentrated before image-domain lesion localization. Conclusion: K-space-aware deep learning improves the robustness of breast lesion segmentation under MRI undersampling and k-space noise, while matching image-space methods at full sampling.

Abstract

Explore similar work

Robustness of breast lesion segmentation under MRI undersampling improves with k-space-aware deep learning

Direct low-field MRI super-resolution using undersampled k-space

K-space Gaussian Representation for Parallel MRI