Keywords: Analysis/Processing, Machine Learning/Artificial Intelligence, co-registration, distortion correction, voxelmorph

Motivation: The co-registration between diffusion and T1-weighted data is important for various diffusion analyses, which is challenging due to the geometric distortion in diffusion images.

Goal(s): To achieve accurate and efficient co-registration between diffusion data and T1w image.

Approach: A self-supervised deep learning-based framework VoxelMorph was used to non-linearly align distorted diffusion b=0 image to T1w image. Our proposal was systematically and quantitatively compared to other linear and non-linear  transformations. The benefit was also demonstrated.

Results: VoxelMorph achieved comparable co-registration accuracy compared to NiftyReg and seconds processing time, which was 40 times faster than NiftyReg, or even 300 times faster by leveraging transfer learning.

Impact: Our proposal achieved fast and accurate co-registration between distorted diffusion data and T1w image, which has a great potential to benefit various diffusion MRI data analyses for neuroscientific studies, including region-of-interest specific quantification and surface-based analysis.

Keywords: Analysis/Processing, Quantitative Imaging, Quantitative Mapping, T1ρ mapping, Nonlinear Least Squares, Bi-exponential models, Deep Learning

Motivation: The nonlinear least squares (NLS)-based estimation of mono- and bi-exponential T1ρ maps in the knee joint is highly time-consuming. Deep-learning (DL) methods are faster alternatives.

Goal(s): However, DL requires substantial training data, which is usually obtained using NLS on acquired data. This paper introduces self-supervised DL models that leverage synthetic target data for training, eliminating the need for scanned or NLS data to be used as reference.

Approach: We have tested five different DL models, each utilizing a distinct activation function, and compared them against NLS.

Results: The proposed models are 25-200x faster than the NLS method, with errors close to NLS.

Impact: This study compared five different self-supervised DL models for estimating mono- and bi-exponential T1ρ maps in the knee joint. These models are faster alternatives to NLS, potentially replacing it to produce reference maps.

Keywords: Analysis/Processing, Data Processing, magnetic resonance imaging, diffusion tensor imaging, self-supervised learning, transfer learning

Motivation: MRI with high resolution and/or acceleration factor suffers from intrinsic low signal-to-noise ratio. Supervised learning-based denoising significantly improves image quality, but requires high-SNR data as training targets.

Goal(s): To denoise images using noisy image repetitions without additional acquisition.

Approach: Noise2Average trains CNN to map each noisy image to its residual compared to the average of all noisy images at iteration 1 and all denoised images from iteration k-1 at iteration k. The images from opposite phase-encoding directions of EPI or different echo times of ME-MPRAGE are noisy repetitions.

Results: Noise2Average outperforms BM4D, AONLM and Noise2Noise in terms of image quality and DTI metrics.

Impact: By reducing the requirement for training data and time, Noise2Average substantially increases the feasibility and accessibility of deep learning based denoising methods for MRI and potentially benefits a wider range of clinical and neuroscientific studies.

Keywords: Other AI/ML, Machine Learning/Artificial Intelligence, Super-resolution, Heart

Motivation: Three-dimensional (3D) whole-heart MRI requires long scan times and the sequence used to acquire such scans is susceptible to banding artifacts.

Goal(s): The goal of this study was to develop an unsupervised super-resolution neural network for 3D whole-heart MRI.

Approach: The data used in this study was acquired using a modified Relaxation-Enhanced Angiography without Contrast and Triggering (REACT) sequence. A neural network referred to hereafter as the Super-resolution Neural Network (SRNN) was developed to super-resolve 3D MRI data.

Results: The SRNN allows us to acquire lower-resolution scans, thus decreasing scan time, and provides improved image quality after performing super-resolution.

Impact: The results of this study show that super-resolution offers a viable option to decrease scan time and improve overall image quality in 3D whole-heart MRI.

Keywords: Analysis/Processing, Brain, Unsupervised Anomaly Detection

Motivation: Unsupervised anomaly detection (UAD) approaches on T1 contrast-enhanced (T1CE) images are currently not feasible as T1CE images of healthy individuals are not typically available.

Goal(s): In this work, we aim to eliminate the need for large labeled datasets that are required for manual anomaly detection on T1CE images.

Approach: Using deep learning (DL), we synthesized healthy T1CE images from non-contrast images available in public datasets. We also synthesized healthy-anomalous paired images and forced the DL network to learn the healthy reconstruction. The anomalies were localized by subtracting the reconstruction from the input image.

Results: The proposed method achieves state-of-the art dice score coefficients.

Impact: This work opens up new avenues of research in unsupervised anomaly detection on T1CE images which has been infeasible due to lack of healthy post-contrast images. We also propose a novel contrastive learning paradigm using synthesis of healthy-anomalous image pairs.

Keywords: Data Processing, Software Tools

Motivation: The lack of standardization in MRI metadata increases radiologist workload.

Goal(s): To demonstrate an approach to standardize the image contrast and the body part examined information in the DICOM header and to understand the benefits of self-supervised pretraining on the metadata standardization task in the presence of label noise.

Approach: Masked language modeling was used for pretraining. At the fine-tuning stage, a transformer model was used to predict the image contrast and body part from both the text and numerical DICOM tags.

Results: Pretraining improves robustness to label noise, with there being no loss in performance at 20% label noise.

Impact: Pretraining using masked language modeling is effective at rendering a metadata stanardization system robust to label noise. Such a system can be used to standardize MRI metadata and therefore reduce radiologist workload. Future work should investigate class conditional label noise.

Keywords: AI/ML Image Reconstruction, Low-Field MRI

Motivation: Restoring the structure that is barely visible on the MR images is a major challenge for self supervised enhancement using one input, especially in low-field MR imaging applications.

Goal(s): To improve the image quality and the visibility of some clinically relevant structure in certain contrast in MR images

Approach: We proposed a self-supervised learning framework using the shareable information from other image contrasts. More specifically, two mutual modulations with a cyclic consistency constraint are introduced to guide the training. 

Results: Preliminary results on 0.25T spine MR images suggest that our method can achieve superior results compared to other self-supervised methods.

Impact: The work shows the feasibility of adopting the multiple contrast information to improve the MR images with poor quality without acquiring low resolution/high resolution pairs. It leads to more accurate diagnoses.

Keywords: AI/ML Image Reconstruction, Image Reconstruction

Motivation: Self-supervised learning via data undersampling (SSDU) uses single contrast images in reconstruction, but a typical protocol contains multiple contrasts that provide additional information.

Goal(s): Our goal is to improve self-supervised image reconstruction fidelity by jointly reconstructing multi-contrast images.

Approach: We modify SSDU by concatenating independently under-sampled contrasts along the channel dimension in a VarNet architecture.

Results: Joint multi-contrast SSDU reconstructs with higher SSIM and lower NMSE than single contrast supervised and self-supervised methods.

Impact: Joint multi-contrast SSDU produces higher quality reconstructions than single-contrast methods, without fully-sampled training data. Accelerated multi-contrast imaging protocols will benefit from higher diagnostic quality or higher acceleration factors.

Keywords: AI Diffusion Models, Quantitative Imaging, parallel imaging; apparent diffusion coefficient; IVIM; parameter estimation

Motivation: The distribution of reconstructed MRI signals, used as input for quantitative MRI with self-supervised deep learning, depends on the number of receiver coils. Current loss functions do not account for this, leading to bias.

Goal(s): Develop a non-central chi likelihood (NLC) loss that accounts for the distribution of MRI measures in the most common scenario of parallelised acquisitions.

Approach: Implement and evaluate the NLC loss and compare its performance against the MSE and Rician likelihood loss in simulated data.

Results: The NLC improves performance compared to the Rician likelihood and MSE loss for the mono-exponential ADC model in simulated data.

Impact: The NLC loss permits fast inference of parameters from MRI signals reconstructed from parallelised acquisitions and may reduce bias compared to the Rician and MSE loss. The NLC loss is widely applicable due to the abundance of parallelised MRI acquisitions.

Keywords: AI/ML Image Reconstruction, Machine Learning/Artificial Intelligence

Motivation: In training-data-limited settings, weak supervision –cooperatively utilizing under-sampled and fully-sampled datasets– can be advantageous.

Goal(s): To compare weakly-supervised multi-coil Magnetic Resonance (MR) image reconstruction against reconstruction using only under-sampled or fully-sampled datasets in high- and low-data regimes.

Approach: Pretrain a Variational Network (VarNet) in a self-supervised manner by minimizing L1 loss in k-space using a 4x under-sampled dataset. Transfer the pre-trained weights to another VarNet and fine-tune it using a smaller, fully sampled dataset by optimizing MS-SSIM loss in image space.

Results: We demonstrate improvements in reconstruction quality in the high-data regime as well as enhanced robustness of reconstruction in the low-data regime.

Impact: Multi-coil MR image reconstruction exploiting both under-sampled and fully-sampled datasets is achievable with transfer learning and fine-tuning. Our proposed methodology can provide improved reconstruction quality and robustness.

Keywords: AI/ML Image Reconstruction, Image Reconstruction, Self-supervised AI

Motivation: Supervised deep learning (SDL) has limitations due to data dependency, and self-supervised frameworks like DIP struggle with noise and artifacts.

Goal(s): Introducing PEARL, a novel self-supervised accelerated parallel MRI approach.

Approach: PEARL leverages joint deep decoders coupling with cross-fusion schemes based on multi-parameter priors to achieve enhanced reconstruction.

Results: PEARL outperforms the existing methods, demonstrating notable improvement under highly accelerated acquisition.

Impact: This study emphasizes the significance and potential of self-supervised learning in addressing critical MRI challenges.

Keywords: AI Diffusion Models, Liver

Motivation: Abdominal DW-MRI suffers from low SNR and motion artifacts, compromising ADC reliability.

Goal(s): Improve ADC estimation in low SNR abdominal DW-MRI using a single-image acquisition per b-value. 

Approach: A novel self-supervised training approach using denoising diffusion probabilistic models (ssDDPM). Tailored for multi-b-value DW-MRI images, it requires only a single gradient image per b-value for denoising, which reduces scan time. 

Results: ssDDPM demonstrated superior lesion conspicuity in low b-value images and quantitative ADC maps in comparison to competing methods. In lesion versus normal tissue, a logistic classifier had improved sensitivity from 0.93 to 0.98, and specificity from 0.88 to 0.97, over non-denoised NEX1 images. 

Impact: ssDDPM enhances abdominal DW-MRI ADC accuracy from single acquisitions, reducing scan times and patient discomfort. This gives promise to an earlier, precise tumor detection and monitoring, impacting clinical care and healthcare efficiency.

Keywords: AI/ML Image Reconstruction, Machine Learning/Artificial Intelligence, PET/MR, unsupervised learning

Motivation: PET/MRI enables MR-assisted PET denoising for low-count PET. Recent work has demonstrated the potential of unsupervised deep learning (uDL) denoising method with the advantage of not requiring large amount of training data. We believe the performance of uDL denoising with MR guidance can be further improved.

Goal(s): To demonstrate the efficacy of a novel cost function in enhancing the quality of denoised low-count PET images using uDL techniques.

Approach: We utilized dNet with MRI as input and low-count PET as target, along with a novel cost function consisting of Bowsher prior and mean square error.

Results: Our method outperformed all other denoising methods.

Impact: The impact of the study is the potential for significant improvements in low-count PET image quality through advanced denoising techniques, which could enhance diagnostic accuracy while reducing radiation exposure for patients.

Keywords: Analysis/Processing, Machine Learning/Artificial Intelligence

Motivation: Noise in 7T MRA images can make it challenging to diagnose vascular diseases, and conventional denoisers tend to over-smooth in a way that blur entire image or does not preserve vascular information.

Goal(s): Our goal was to reduce noise in 7T MRA images while maintaining vascular information using deep learning method.

Approach: We devised an unsupervised denoising model using multiple slice information and cycleGAN-based neural networks.

Results: Our approach not only suppressed noise in 7T MRA images, but it also successfully preserved vessel information among the compared models.

Impact: We developed a denoising method in 7T MRA images while preserving vascular information. Clinically, our findings will help diagnose vascular-related diseases. High SNR is preserved by averaging adjacent slices, and we contribute to increase usability of 7T MRI.

Keywords: Analysis/Processing, Liver, Machine learning, quantitative imaging

Motivation: Extensive recent work has been devoted to quantitative MRI, but practical implementation of quantitative parameter mapping is hindered by the lack of tools for easy visualization and automated analysis. 

Goal(s): We demonstrate the applicability of a self-supervised contrastive pretraining framework for organ segmentation in automated analysis of free-breathing 3D liver T1 mapping.

Approach: A DL model is pretrained to learn T1 contrast information from multi-contrast images acquired for T1 parameter mapping.

Results: With few labeled examples, an organ segmentation framework was developed, and its utility in interpreting parameter maps was demonstrated.

Impact: Multi-contrast information from images typically acquired for parameter estimation in quantitative MRI can be leveraged to pretrain organ segmentation models with self-supervision, enabling automated analysis of quantitative parameter maps.

Keywords: AI/ML Image Reconstruction, Image Reconstruction, Self-supervised learning

Motivation: Addressing limitations in parallel imaging, particularly GRAPPA’s challenges like noise amplification and dependence on linear k-space value combinations.

Goal(s): To enhance k-space interpolation accuracy with a transformer network, thereby improving the quality of clinical imaging.

Approach: We employ a novel self-supervised transformer network with an attention mechanism - TransGRAPPA, exploiting latent features for nonlinear interpolation of missing k-space points.

Results: TransGRAPPA outperforms GRAPPA and RAKI in terms of NRMSE, PSNR, SSIM, and noise reduction, showcasing enhanced capabilities on fastMRI’s multi-coil knee dataset.

Impact: The study presents a innovative reconstruction method using transformer network to explore k-space point relationships with limited training data, offering potential improvements in MR image quality and scan speed, and more efficient and accurate diagnostics in medical imaging.