Spatiotemporal (four-dimensional) infant-dedicated brain atlases are essential for neuroimaging analysis of early dynamic brain development. However, due to the substantial technical challenges in the acquisition and processing of infant brain MR images, 4D atlases densely covering the dynamic brain development during infancy are still scarce. Few existing ones generally have fuzzy tissue contrast and low spatiotemporal resolution, leading to degraded accuracy of atlas-based normalization and subsequent analyses. To address this issue, in this paper, we construct a 4D structural MRI atlas for infant brains based on the UNC/UMN Baby Connectome Project (BCP) dataset, which features a high spatial resolution, extensive age-range coverage, and densely sampled time points. Specifically, 542 longitudinal T1w and T2w scans from 240 typically developing infants up to 26-month of age were utilized for our atlas construction. To improve the co-registration accuracy of the infant brain images, which typically exhibit dynamic appearance with low tissue contrast, we employed the state-of-the-art registration method and leveraged our generated reliable brain tissue probability maps in addition to the intensity images to improve the alignment of individual images. To achieve consistent region labeling on both infant and adult brain images for facilitating region-based analysis across ages, we mapped the widely used Desikan cortical parcellation onto our atlas by following an age-decreasing mapping manner. Meanwhile, the typical subcortical structures were manually delineated to facilitate the studies related to the subcortex. Compared with the existing infant brain atlases, our 4D atlas has much higher spatiotemporal resolution and preserves more structural details, and thus can boost accuracy in neurodevelopmental analysis during infancy. <<<

Deep Spiking Neural Networks (SNNs) present optimization difficulties for gradient-based approaches due to discrete binary activation and complex spatial-temporal dynamics. Considering the huge success of ResNet in deep learning, it would be natural to train deep SNNs with residual learning. Previous Spiking ResNet mimics the standard residual block in ANNs and simply replaces ReLU activation layers with spiking neurons, which suffers the degradation problem and can hardly implement residual learning. In this paper, we propose the spike-element-wise (SEW) ResNet to realize residual learning in deep SNNs. We prove that the SEW ResNet can easily implement identity mapping and overcome the vanishing/exploding gradient problems of Spiking ResNet. We evaluate our SEW ResNet on ImageNet, DVS Gesture, and CIFAR10-DVS datasets, and show that SEW ResNet outperforms the state-of-the-art directly trained SNNs in both accuracy and time-steps. Moreover, SEW ResNet can achieve higher performance by simply adding more layers, providing a simple method to train deep SNNs. To our best knowledge, this is the first time that directly training deep SNNs with more than 100 layers becomes possible. <<<

Uncertainty estimation in medical image registration enables surgeons to evaluate the operative risk based on the trustworthiness of the registered image data thus of paramount importance for practical clinical applications. Despite the recent promising results obtained with deep unsupervised learning-based registration methods, reasoning about uncertainty of unsupervised registration models remains largely unexplored. In this work, we propose a predictive module to learn the registration and uncertainty in correspondence simultaneously. Our framework introduces empirical randomness and registration error based uncertainty prediction. We systematically assess the performances on two MRI datasets with different ensemble paradigms. Experimental results highlight that our proposed framework significantly improves the registration accuracy and uncertainty compared with the baseline. <<<

In this work, we propose a theoretical framework based on maximum profile likelihood for pairwise and groupwise registration. By an asymptotic analysis, we demonstrate that maximum profile likelihood registration minimizes an upper bound on the joint entropy of the distribution that generates the joint image data. Further, we derive the congealing method for groupwise registration by optimizing the profile likelihood in closed form, and using coordinate ascent, or iterative model refinement. We also describe a method for feature based registration in the same framework and demonstrate it on groupwise tractographic registration. In the second part of the article, we propose an approach to deep metric registration that implements maximum likelihood registration using deep discriminative classifiers. We show further that this approach can be used for maximum profile likelihood registration to discharge the need for well-registered training data, using iterative model refinement. We demonstrate that the method succeeds on a challenging registration problem where the standard mutual information approach does not perform well. <<<

PURPOSE: The accuracy of minimally invasive, intracranial neurosurgery can be challenged by deformation of brain tissue - e.g., up to 10 mm due to egress of cerebrospinal fluid during neuroendoscopic approach. We report an unsupervised, deep learning-based registration framework to resolve such deformations between preoperative MR and intraoperative CT with fast runtime for neurosurgical guidance.METHOD: The framework incorporates subnetworks for MR and CT image synthesis with a dual-channel registration subnetwork (with synthesis uncertainty providing spatially varying weights on the dual-channel loss) to estimate a diffeomorphic deformation field from both the MR and CT channels. An end-to-end training is proposed that jointly optimizes both the synthesis and registration subnetworks. The proposed framework was investigated using three datasets: (1) paired MR/CT with simulated deformations; (2) paired MR/CT with real deformations; and (3) a neurosurgery dataset with real deformation. Two state-of-the-art methods (Symmetric Normalization and VoxelMorph) were implemented as a basis of comparison, and variations in the proposed dual-channel network were investigated, including single-channel registration, fusion without uncertainty weighting, and conventional sequential training of the synthesis and registration subnetworks.RESULTS: The proposed method achieved: (1) Dice coefficient = 0.82±0.07 and TRE = 1.2 ± 0.6 mm on paired MR/CT with simulated deformations; (2) Dice coefficient = 0.83 ± 0.07 and TRE = 1.4 ± 0.7 mm on paired MR/CT with real deformations; and (3) Dice = 0.79 ± 0.13 and TRE = 1.6 ± 1.0 mm on the neurosurgery dataset with real deformations. The dual-channel registration with uncertainty weighting demonstrated superior performance (e.g., TRE = 1.2 ± 0.6 mm) compared to single-channel registration (TRE = 1.6 ± 1.0 mm, p < 0.05 for CT channel and TRE = 1.3 ± 0.7 mm for MR channel) and dual-channel registration without uncertainty weighting (TRE = 1.4 ± 0.8 mm, p < 0.05). End-to-end training of the synthesis and registration subnetworks also improved performance compared to the conventional sequential training strategy (TRE = 1.3 ± 0.6 mm). Registration runtime with the proposed network was ∼3 s.CONCLUSION: The deformable registration framework based on dual-channel MR/CT registration with spatially varying weights and end-to-end training achieved geometric accuracy and runtime that was superior to state-of-the-art baseline methods and various ablations of the proposed network. The accuracy and runtime of the method may be compatible with the requirements of high-precision neurosurgery. <<<

This paper addresses the scalability challenge of architecture search by formulating the task in a differentiable manner. Unlike conventional approaches of applying evolution or reinforcement learning over a discrete and non-differentiable search space, our method is based on the continuous relaxation of the architecture representation, allowing efficient search of the architecture using gradient descent. Extensive experiments on CIFAR-10, ImageNet, Penn Treebank and WikiText-2 show that our algorithm excels in discovering high-performance convolutional architectures for image classification and recurrent architectures for language modeling, while being orders of magnitude faster than state-of-the-art non-differentiable techniques. Our implementation has been made publicly available to facilitate further research on efficient architecture search algorithms. <<<

Non-rigid cortical registration is an important and challenging task due to the geometric complexity of the human cortex and the high degree of inter-subject variability. A conventional solution is to use a spherical representation of surface properties and perform registration by aligning cortical folding patterns in that space. This strategy produces accurate spatial alignment, but often requires high computational cost. Recently, convolutional neural networks (CNNs) have demonstrated the potential to dramatically speed up volumetric registration. However, due to distortions introduced by projecting a sphere to a 2D plane, a direct application of recent learning-based methods to surfaces yields poor results. In this study, we present SphereMorph, a diffeomorphic registration framework for cortical surfaces using deep networks that addresses these issues. SphereMorph uses a UNet-style network associated with a spherical kernel to learn the displacement field and warps the sphere using a modified spatial transformer layer. We propose a resampling weight in computing the data fitting loss to account for distortions introduced by polar projection, and demonstrate the performance of our proposed method on two tasks, including cortical parcellation and group-wise functional area alignment. The experiments show that the proposed SphereMorph is capable of modeling the geometric registration problem in a CNN framework and demonstrate superior registration accuracy and computational efficiency. The source code of SphereMorph will be released to the public upon acceptance of this manuscript at https://github.com/voxelmorph/spheremorph. <<<

Biomedical imaging is a driver of scientific discovery and a core component of medical care and is being stimulated by the field of deep learning. While semantic segmentation algorithms enable image analysis and quantification in many applications, the design of respective specialized solutions is non-trivial and highly dependent on dataset properties and hardware conditions. We developed nnU-Net, a deep learning-based segmentation method that automatically configures itself, including preprocessing, network architecture, training and post-processing for any new task. The key design choices in this process are modeled as a set of fixed parameters, interdependent rules and empirical decisions. Without manual intervention, nnU-Net surpasses most existing approaches, including highly specialized solutions on 23 public datasets used in international biomedical segmentation competitions. We make nnU-Net publicly available as an out-of-the-box tool, rendering state-of-the-art segmentation accessible to a broad audience by requiring neither expert knowledge nor computing resources beyond standard network training. <<<

Brain development in the first 2 years after birth is extremely dynamic and likely plays an important role in neurodevelopmental disorders, including autism and schizophrenia. Knowledge regarding this period is currently quite limited. We studied structural brain development in healthy subjects from birth to 2. Ninety-eight children received structural MRI scans on a Siemens head-only 3T scanner with magnetization prepared rapid gradient echo T1-weighted, and turbo spin echo, dual-echo (proton density and T2 weighted) sequences: 84 children at 2-4 weeks, 35 at 1 year and 26 at 2 years of age. Tissue segmentation was accomplished using a novel automated approach. Lateral ventricle, caudate, and hippocampal volumes were also determined. Total brain volume increased 101% in the first year, with a 15% increase in the second. The majority of hemispheric growth was accounted for by gray matter, which increased 149% in the first year; hemispheric white matter volume increased by only 11%. Cerebellum volume increased 240% in the first year. Lateral ventricle volume increased 280% in the first year, with a small decrease in the second. The caudate increased 19% and the hippocampus 13% from age 1 to age 2. There was robust growth of the human brain in the first two years of life, driven mainly by gray matter growth. In contrast, white matter growth was much slower. Cerebellum volume also increased substantially in the first year of life. These results suggest the structural underpinnings of cognitive and motor development in early childhood, as well as the potential pathogenesis of neurodevelopmental disorders. <<<

Dongming Wei, Sahar Ahmad, Yuyu Guo, Liyun Chen, Yunzhi Huang, Lei Ma, Zhengwang Wu, Gang Li, Li Wang, Weili Lin, Pew-Thian Yap, Dinggang Shen, Qian Wang <<<

Deformable registration is fundamental to longitudinal and population-based image analyses. However, it is challenging to precisely align longitudinal infant brain MR images of the same subject, as well as cross-sectional infant brain MR images of different subjects, due to fast brain development during infancy. In this paper, we propose a recurrently usable deep neural network for the registration of infant brain MR images. There are three main highlights of our proposed method. (i) We use brain tissue segmentation maps for registration, instead of intensity images, to tackle the issue of rapid contrast changes of brain tissues during the first year of life. (ii) A single registration network is trained in a one-shot manner, and then recurrently applied in inference for multiple times, such that the complex deformation field can be recovered incrementally. (iii) We also propose both the adaptive smoothing layer and the tissue-aware anti-folding constraint into the registration network to ensure the physiological plausibility of estimated deformations without degrading the registration accuracy. Experimental results, in comparison to the state-of-the-art registration methods, indicate that our proposed method achieves the highest registration accuracy while still preserving the smoothness of the deformation field. The implementation of our proposed registration network is available online https://github.com/Barnonewdm/ACTA-Reg-Net. <<<

We propose an echo planar imaging (EPI) distortion correction method (DR-BUDDI), specialized for diffusion MRI, which uses data acquired twice with reversed phase encoding directions, often referred to as blip-up blip-down acquisitions. DR-BUDDI can incorporate information from an undistorted structural MRI and also use diffusion-weighted images (DWI) to guide the registration, improving the quality of the registration in the presence of large deformations and in white matter regions. DR-BUDDI does not require the transformations for correcting blip-up and blip-down images to be the exact inverse of each other. Imposing the theoretical "blip-up blip-down distortion symmetry" may not be appropriate in the presence of common clinical scanning artifacts such as motion, ghosting, Gibbs ringing, vibrations, and low signal-to-noise. The performance of DR-BUDDI is evaluated with several data sets and compared to other existing blip-up blip-down correction approaches. The proposed method is robust and generally outperforms existing approaches. The inclusion of the DWIs in the correction process proves to be important to obtain a reliable correction of distortions in the brain stem. Methods that do not use DWIs may produce a visually appealing correction of the non-diffusion weighted images, but the directionally encoded color maps computed from the tensor reveal an abnormal anatomy of the white matter pathways. <<<

To develop and validate a radiomics model for evaluating pathologic complete response (pCR) to neoadjuvant chemoradiotherapy in patients with locally advanced rectal cancer (LARC). We enrolled 222 patients (152 in the primary cohort and 70 in the validation cohort) with clinicopathologically confirmed LARC who received chemoradiotherapy before surgery. All patients underwent T2-weighted and diffusion-weighted imaging before and after chemoradiotherapy; 2,252 radiomic features were extracted from each patient before and after treatment imaging. The two-sample test and the least absolute shrinkage and selection operator regression were used for feature selection, whereupon a radiomics signature was built with support vector machines. Multivariable logistic regression analysis was then used to develop a radiomics model incorporating the radiomics signature and independent clinicopathologic risk factors. The performance of the radiomics model was assessed by its calibration, discrimination, and clinical usefulness with independent validation. The radiomics signature comprised 30 selected features and showed good discrimination performance in both the primary and validation cohorts. The individualized radiomics model, which incorporated the radiomics signature and tumor length, also showed good discrimination, with an area under the receiver operating characteristic curve of 0.9756 (95% confidence interval, 0.9185-0.9711) in the validation cohort, and good calibration. Decision curve analysis confirmed the clinical utility of the radiomics model. Using pre- and posttreatment MRI data, we developed a radiomics model with excellent performance for individualized, noninvasive prediction of pCR. This model may be used to identify LARC patients who can omit surgery after chemoradiotherapy. . <<<

During the first 2 postnatal years, cortical thickness of the human brain develops dynamically and spatially heterogeneously and likely peaks between 1 and 2 y of age. The striking development renders this period critical for later cognitive outcomes and vulnerable to early neurodevelopmental disorders. However, due to the difficulties in longitudinal infant brain MRI acquisition and processing, our knowledge still remains limited on the dynamic changes, peak age, and spatial heterogeneities of cortical thickness during infancy. To fill this knowledge gap, in this study, we discover the developmental regionalization of cortical thickness, i.e., developmentally distinct regions, each of which is composed of a set of codeveloping cortical vertices, for better understanding of the spatiotemporal heterogeneities of cortical thickness development. We leverage an infant-dedicated computational pipeline, an advanced multivariate analysis method (i.e., nonnegative matrix factorization), and a densely sampled longitudinal dataset with 210 serial MRI scans from 43 healthy infants, with each infant being scheduled to have 7 longitudinal scans at around 1, 3, 6, 9, 12, 18, and 24 mo of age. Our results suggest that, during the first 2 y, the whole-brain average cortical thickness increases rapidly and reaches a plateau at about 14 mo of age and then decreases at a slow pace thereafter. More importantly, each discovered region is structurally and functionally meaningful and exhibits a distinctive developmental pattern, with several regions peaking at varied ages while others keep increasing in the first 2 postnatal years. Our findings provide valuable references and insights for early brain development. <<<