[Paper Review] Motion Correction and Volumetric Reconstruction for Fetal Functional Magnetic Resonance Imaging Data

Daniel Sobotka, Michael Ebner+9 more|arXiv (Cornell University)|Feb 11, 2022

Fetal and Pediatric Neurological Disorders63 references18 citations

TL;DR

This paper proposes a novel motion correction and volumetric reconstruction framework for fetal fMRI that uses outlier-robust reference volume estimation and Huber L2 regularization to improve functional connectivity accuracy. By reducing motion-induced artifacts and signal outliers, the method enhances reproducibility and signal interpretability, decreasing outlier time points from 18.6% to 9.16% across gestational ages 20–25 weeks.

ABSTRACT

Motion correction is an essential preprocessing step in functional Magnetic Resonance Imaging (fMRI) of the fetal brain with the aim to remove artifacts caused by fetal movement and maternal breathing and consequently to suppress erroneous signal correlations. Current motion correction approaches for fetal fMRI choose a single 3D volume from a specific acquisition timepoint with least motion artefacts as reference volume, and perform interpolation for the reconstruction of the motion corrected time series. The results can suffer, if no low-motion frame is available, and if reconstruction does not exploit any assumptions about the continuity of the fMRI signal. Here, we propose a novel framework, which estimates a high-resolution reference volume by using outlier-robust motion correction, and by utilizing Huber L2 regularization for intra-stack volumetric reconstruction of the motion-corrected fetal brain fMRI. We performed an extensive parameter study to investigate the effectiveness of motion estimation and present in this work benchmark metrics to quantify the effect of motion correction and regularised volumetric reconstruction approaches on functional connectivity computations. We demonstrate the proposed framework's ability to improve functional connectivity estimates, reproducibility and signal interpretability, which is clinically highly desirable for the establishment of prognostic noninvasive imaging biomarkers. The motion correction and volumetric reconstruction framework is made available as an open-source package of NiftyMIC.

Motivation & Objective

• To address the challenge of motion artifacts in fetal fMRI caused by fetal movement and maternal respiration.
• To improve functional connectivity estimation by minimizing spurious correlations due to motion.
• To develop a robust, open-source pipeline for motion correction and volumetric reconstruction that preserves signal quality.
• To enable more reliable, reproducible functional connectivity analysis in clinically challenging fetal fMRI data.
• To reduce data loss from motion censoring by correcting motion rather than excluding corrupted time points.

Proposed method

• Proposes a three-component framework: (1) outlier-robust high-resolution reference volume construction using S2V motion correction.
• Applies rigid slice-based motion correction to align individual slices across time points.
• Employs Huber L2 regularization for intra-stack volumetric reconstruction to preserve signal continuity.
• Conducts a parametric study comparing TK1 L2, TV L2, and Huber L2 regularization to identify optimal parameters.
• Uses intensity, outlier, and connectivity-based metrics to evaluate signal quality and functional connectivity.
• Validates motion estimation using synthetic data with known motion patterns to ensure accuracy.

Experimental results

Research questions

• RQ1 Can a robust, outlier-robust reference volume improve motion correction in highly motion-corrupted fetal fMRI data?
• RQ2 Does Huber L2 regularization outperform traditional regularization methods (e.g., TK1 L2, TV L2) in preserving functional connectivity and reducing artifacts?
• RQ3 To what extent does the proposed framework reduce the number of outlier time points in functional connectivity analysis?
• RQ4 How does motion correction affect the reproducibility and interpretability of functional connectivity networks in fetal fMRI?
• RQ5 Can the framework be applied effectively across different acquisition protocols, including axial, coronal, and sagittal sequences?

Key findings

• The proposed framework reduced the outlier ratio from 12.9% to 8.6% (p-value 0.019) in one subject, indicating improved signal quality after motion correction and Huber L2 regularization.
• The slope of the relationship between subject movement and functional connectivity decreased from -0.0029 to -0.0010, showing reduced motion-related bias.
• Huber L2 regularization with a parameter of 0.1 yielded the most reproducible and accurate functional connectivity estimates compared to other regularization methods.
• Between gestational ages 20–25 weeks, the percentage of outlier time points decreased from 18.6% to 9.16%, and from 7.125% to 5% between 26–31 weeks.
• The method effectively reduced spurious high-correlation regions at the brain border, which were artifacts of motion, especially with TK1 L2 or MCFLIRT.
• The framework maintains high performance even with extreme movements (up to 50 mm), though reconstruction quality decreased slightly under extreme EPI and interleaved protocol conditions.