MRI Contrast Enhancement Kinetics World Model

TL;DR

This work proposes a new MRI Contrast Enhancement Kinetics World model (MRI CEKWorld) with SpatioTemporal Consistency Learning (STCL) and proposes Latent Difference Learning (LDL) which extends the unobserved intervals by interpolation and constrains smooth variations in the latent space among interpolated sequences.

Abstract

Clinical MRI contrast acquisition suffers from inefficient information yield, which presents as a mismatch between the risky and costly acquisition protocol and the fixed and sparse acquisition sequence. Applying world models to simulate the contrast enhancement kinetics in the human body enables continuous contrast-free dynamics. However, the low temporal resolution in MRI acquisition restricts the training of world models, leading to a sparsely sampled dataset. Directly training a generative model to capture the kinetics leads to two limitations: (a) Due to the absence of data on missing time, the model tends to overfit to irrelevant features, leading to content distortion. (b) Due to the lack of continuous temporal supervision, the model fails to learn the continuous kinetics law over time, causing temporal discontinuities. For the first time, we propose MRI Contrast Enhancement Kinetics World model (MRI CEKWorld) with SpatioTemporal Consistency Learning (STCL). For (a), guided by the spatial law that patient-level structures remain consistent during enhancement, we propose Latent Alignment Learning (LAL) that constructs a patient-specific template to constrain contents to align with this template. For (b), guided by the temporal law that the kinetics follow a consistent smooth trend, we propose Latent Difference Learning (LDL) which extends the unobserved intervals by interpolation and constrains smooth variations in the latent space among interpolated sequences. Extensive experiments on two datasets show our MRI CEKWorld achieves better realistic contents and kinetics. Codes will be available at https://github.com/DD0922/MRI-Contrast-Enhancement-Kinetics-World-Model.

Figures (14)

• Figure 1: (a) Task: MRI CEKWorld generates contrast-enhanced sequences that conform to kinetics in the human body after contrast agent injection. (b) Problem: Clinical contrast MRI acquisition presents inefficient information yield with adverse risks and higher cost, but a fixed, sparse sequence. (c) Advantages: Our MRI CEKWorld enables continuous contrast-free dynamics with no contrast agent risks, low cost, and convenience.
• Figure 2: Limitation: MRI Acquisition-induced low temporal resolution in MRI CEKWorld leads to (a) Content distortion and (b) Temporal discontinuities across time.
• Figure 3: Overview framework of the MRI CEKWorld. (a) and (b) shows the training and inference processes. (c) LAL captures region-wise co-occurrence relationships and enforces anatomical consistency by aligning to a patient-level template. (d) LDL constructs a dense time series in the latent space and imposes a second-order difference (denoted as Diff) on adjacent moments for smooth evolution ($p$ and $q$ denote the inference ).
• Figure 4: Visualization results in Abdominal DCE-MRI: The visualization results of our methods on different time points exhibit better spatial reality (zoom-in regions in blue boxes) and temporal continuities than comparisons (connected green boxes), whereas other methods show deviations from realistic kinetics or lack of dynamic consistency.
• Figure 5: Visualization results in Breast DCE-MRI: The visualization results achieves temporal consistent (connected green boxes) with contrast agent kinetics, demonstrating superior fidelity in breast DCE-MRI sequence generation.