Optimizing Connectivity-Driven Brain Parcellation Using Ensemble Clustering

Region-of-interest extraction in low depth of field images using ensemble clustering and difference of Gaussian approaches

Short-range association fibers (SAFs) are critical for cortical communications but are often underestimated in conventional resolution diffusion magnetic resonance imaging (dMRI) since they locate within a ~1.5 mm thin layer of superficial white matter. With the advent of high-resolution diffusion imaging techniques, this study evaluated the effects of image spatial resolution on SAF reconstruction using two datasets: (1) prospectively acquired dMRI data from 20 healthy subjects, each scanned at 3 resolutions (i.e., 2, 1.5, and 0.96 mm iso.), and (2) retrospectively down-sampled dMRI data from the Human Connectome Project dataset, as well as 20 representative MRtrix3-based tractography pipelines. It was found that lower resolution degraded superficial white matter model fitting, lowered the SAF streamline counts, and reduced global and regional short-range connectivity fraction (SCF), defined as the fraction of SAF connections among all association fiber connections, across all tested methods. Temporal lobe cortical regions exhibited the greatest SCF declines at lower resolutions. Tractography methods differed in resolution sensitivity, with diffusion tensor imaging (DTI)-based single-tissue single-fiber tractography showing greater decreases in SCF than constrained spherical deconvolution (CSD)-based multi-tissue multi-fiber tractography at lower resolutions. Probabilistic, anatomically constrained tractography combined with spherical-deconvolution informed filtering of tractograms was more robust to decreases in resolution. Up-sampling to a nominally higher resolution partially improved model fitting and SCF accuracy across the evaluated pipelines, with the greatest effect observed for DTI. Using the 0.96 mm iso. gSlider data and optimized tractography pipelines from this study, we constructed the first human brain atlas of RSCF. In summary, this study provides a systematic and quantitative evaluation using MRtrix3 of how spatial resolution, fiber models, and tracking methodologies affect SAF reconstruction and structural connectivity estimation, serving as a reference framework for methodological choices. These advances may enhance the characterization of both healthy and diseased human brains across a wide range of neuroscientific and clinical applications.

Endoscopic Anatomy of Transcallosal Hemispherotomy: Laboratory Study with Advanced Three-Dimensional Modeling

Limited sample size hinders the application of deep learning in brain image analysis, and transfer learning is a possible solution. However, most pretrained models are 2D based and cannot be applied directly to 3D brain images. In this study, we propose a novel framework to apply 2D pretrained models to 3D brain images by projecting surface-based cortical morphometry into planar images using computational geometry mapping. Firstly, 3D cortical meshes are reconstructed from magnetic resonance imaging (MRI) using FreeSurfer and projected into 2D planar meshes with topological preservation based on area-preserving geometry mapping. Then, 2D deep models pretrained on ImageNet are adopted and fine-tuned for cortical image classification on morphometric shape metrics. We apply the framework to sex classification on the Human Connectome Project dataset and autism spectrum disorder (ASD) classification on the Autism Brain Imaging Data Exchange dataset. Moreover, a 2-stage transfer learning strategy is suggested to boost the ASD classification performance by using the sex classification as an intermediate task. Our framework brings significant improvement in sex classification and ASD classification with transfer learning. In summary, the proposed framework builds a bridge between 3D cortical data and 2D models, making 2D pretrained models available for brain image analysis in cognitive and psychiatric neuroscience.

We present a novel way to model diffusion magnetic resonance imaging (dMRI) datasets, that benefits from the structural coherence of the human brain while only using data from a single subject. Current methods model the dMRI signal in individual voxels, disregarding the intervoxel coherence that is present. We use a neural network to parameterize a spherical harmonics series (NeSH) to represent the dMRI signal of a single subject from the Human Connectome Project dataset, continuous in both the angular and spatial domain. The reconstructed dMRI signal using this method shows a more structurally coherent representation of the data. Noise in gradient images is removed and the fiber orientation distribution functions show a smooth change in direction along a fiber tract. We showcase how the reconstruction can be used to calculate mean diffusivity, fractional anisotropy, and total apparent fiber density. These results can be achieved with a single model architecture, tuning only one hyperparameter. In this paper we also demonstrate how upsampling in both the angular and spatial domain yields reconstructions that are on par or better than existing methods.

Data clustering is an important task and has found applications in numerous real-world problems. Since no single clustering algorithm is able to identify all different types of cluster shapes and structures, ensemble clustering was proposed to combine different partitions of the same data generated by multiple clustering algorithms. The key idea of most ensemble clustering algorithms is to find a partition that is consistent with most of the available partitions of the input data. One problem with these algorithms is their inability to handle uncertain data pairs, i.e. data pairs for which about half of the partitions put them into the same cluster and the other half do the opposite. When the number of uncertain data pairs is large, they can mislead the ensemble clustering algorithm in generating the final partition. To overcome this limitation, we propose an ensemble clustering approach based on the technique of matrix completion. The proposed algorithm constructs a partially observed similarity matrix based on the data pairs whose cluster memberships are agreed upon by most of the clustering algorithms in the ensemble. It then deploys the matrix completion algorithm to complete the similarity matrix. The final data partition is computed by applying an efficient spectral clustering algorithm to the completed matrix. Our empirical studies with multiple real-world datasets show that the proposed algorithm performs significantly better than the state-of-the-art algorithms for ensemble clustering.

Background: While functional magnetic resonance imaging (fMRI) has become an established noninvasive tool for studying brain activity in both healthy and diseased states, no broad consensus has been reached regarding preprocessing methodology. Furthermore, the relationship between variations in preprocessing and functional connectivity (FC) networks remains incompletely understood. Purpose: The aim of this study was to relate FC to (1) choices in preprocessing methodology and (2) subject motion. Methods: Clinical and MRI data were analyzed from healthy subjects acquired as part of the Autism Brain Imaging Data Exchange (ABIDE). Data were obtained from 508 healthy subjects. Data from subjects in the highest and lowest quartiles for motion were used to calculate the interaction between motion and preprocessing. Data were analyzed across four domains of fMRI preprocessing: (1) pipeline, (2) global signal regression (GSR), (3) bandpass filtering, and (4) anatomic atlas. For the FC network calculated from each preprocessing scheme, overall FC using Pearson correlation, as well as leaf fraction and diameter, was calculated for each subject, and statistical comparison was made across schemes using generalized estimating equations. Results: FC and global network properties were significantly affected by each preprocessing step, and each preprocessing step significantly interacted with subject motion to differentially affect global functional network properties, with GSR having the strongest effect. Conclusion: Preprocessing choices in fMRI studies influence overall FC and global network properties and can have motion-dependent effects. Impact statement Different parameter choices across the four domains of functional magnetic resonance imaging (fMRI) preprocessing analyzed in this study (global signal regression, bandpass filtering, anatomic atlas, and pipeline) were associated with differences in functional connectivity (FC) as well as global network properties in healthy subjects. Some fMRI preprocessing parameter choices resulted in motion-dependent effects on FC and global network properties.

Ensemble clustering has become a hot research field in intelligent information processing and machine learning. Although significant progress has been made in recent years, there are still two challenging issues in the current ensemble clustering research. First of all, most ensemble clustering algorithms tend to explore similarity at the level of object but lack the ability to explore information at the level of cluster. Secondly, many ensemble clustering algorithms only focus on the direct relationship, while ignoring the indirect relationship between clusters. In order to solve these two problems, a link-based meta-clustering algorithm (L-MCLA) have been proposed in this paper. A series of experiment results demonstrate that the proposed algorithm not only produces better clustering effect but is also less influenced by different ensemble sizes.

Ensemble clustering has recently emerged as a powerful tool for aggregating multiple clustering results into a probably better and more robust clustering result. While many ensemble clustering techniques have been developed in recent years, there is still a potential drawback in most of the existing algorithms. They generally neglect the possible noise inside the ensemble of base clusterings, which may exhibit a negative influence on the robustness of the consensus result. To address this, this paper proposes a new ensemble clustering algorithm based on noise-aware graph decomposition. The co-association matrix is first constructed to accumulate information from multiple base clusterings, and then sparsified by means of K-nearest neighbors. By treating the sparsified co-association matrix as the similarity matrix, the corresponding graph is further decomposed into two parts, i.e., the corruption(noise) graph and the good(denoised) graph. In particular, the denoised graph as well as its spectral embedding can be learned simultaneously by solving a trace minimization problem. Finally, upon the learned spectral embedding, the k-means discretization is performed to obtain the final consensus clustering. Experiments on multiple real-world datasets have shown the superiority of the proposed algorithm.

To investigate functional brain networks, many graph-theoretical studies have defined nodes in a graph using an anatomical atlas with about a hundred partitions. Although use of anatomical node definition is popular due to its convenience, functional inhomogeneity within each node may lead to bias or systematic errors in the graph analysis. The current study was aimed to show functional inhomogeneity of a node defined by an anatomical atlas and to show its effects on the graph topology. For this purpose, we compared functional connectivity defined using 138 resting state fMRI data among 90 cerebral nodes from the automated anatomical labeling (AAL), which is an anatomical atlas, and among 372 cerebral nodes defined using a functional connectivity-based atlas as a ground truth, which was obtained using anatomy-constrained hierarchical modularity optimization algorithm (AHMO) that we proposed to evaluate the graph properties for anatomically defined nodes. We found that functional inhomogeneity in the anatomical parcellation induced significant biases in estimating both functional connectivity and graph-theoretical network properties. We also found very high linearity in major global network properties and nodal strength at all brain regions between anatomical atlas and functional atlas with reasonable network-forming thresholds for graph construction. However, some nodal properties such as betweenness centrality did not show significant linearity in some regions. The current study suggests that the use of anatomical atlas may be biased due to its inhomogeneity, but may generally be used in most neuroimaging studies when a single atlas is used for analysis.

Clustering ensemble is an open proposition that aims to solve the limitation of a single cluster algorithm on the diversity of data structures in data partitions. It obtains a consensus model by integrating multiple clustering algorithms. Compared with a single clustering algorithm, it can process the data more accurately and stably, and has a better tolerance for the structure and diversity of the data. This paper proposes to integrate particle swarm optimization for clustering ensemble model. Particularly, the idea is to first use multiple classic clustering algorithms to cluster the original data set to obtain a preliminary clustering result set, and then search for a clustering result that is most consistent with the preliminary results by a particle swarm optimization procedure. In the experiments, we make use of public data sets and micro-precision to evaluate the effectiveness of the proposed clustering ensemble model, and compare it with both base clustering algorithms and other common clustering ensemble algorithms. The evaluation results show that this model can significantly improve the base clustering algorithms and outperforms common clustering ensemble algorithms in most cases.

Clustering ensembles combine multiple partitions of data into a single clustering solution. It is an effective technique for improving the quality of clustering results. Current clustering ensemble algorithms are usually built on the pairwise agreements between clusterings that focus on the similarity via consensus functions, between data objects that induce similarity measures from partitions and re-cluster objects, and between clusters that collapse groups of clusters into meta-clusters. In most of those models, there is a strong assumption on IIDness (i.e., independent and identical distribution), which states that base clusterings perform independently of one another and all objects are also independent. In the real world, however, objects are generally likely related to each other through features that are either explicit or even implicit. There is also latent but definite relationship among intermediate base clusterings because they are derived from the same set of data. All these demand a further investigation of clustering ensembles that explores the interdependence characteristics of data. To solve this problem, a new coupled clustering ensemble (CCE) framework that works on the interdependence nature of objects and intermediate base clusterings is proposed in this article. The main idea is to model the coupling relationship between objects by aggregating the similarity of base clusterings, and the interactive relationship among objects by addressing their neighborhood domains. Once these interdependence relationships are discovered, they will act as critical supplements to clustering ensembles. We verified our proposed framework by using three types of consensus function: clustering-based, object-based, and cluster-based. Substantial experiments on multiple synthetic and real-life benchmark datasets indicate thatCCEcan effectively capture the implicit interdependence relationships among base clusterings and among objects with higher clustering accuracy, stability, and robustness compared to 14 state-of-the-art techniques, supported by statistical analysis. In addition, we show that the final clustering quality is dependent on the data characteristics (e.g., quality and consistency) of base clusterings in terms of sensitivity analysis. Finally, the applications in document clustering, as well as on the datasets with much larger size and dimensionality, further demonstrate the effectiveness, efficiency, and scalability of our proposed models.

Clustering ensemble (CE), renowned for its robust and potent consensus capability, has garnered significant attention from scholars in recent years and has achieved numerous noteworthy breakthroughs. Nevertheless, three key issues persist: (1) the majority of CE selection strategies rely on preset parameters or empirical knowledge as a premise, lacking adaptive selectivity; (2) the construction of co-association matrix is excessively one-sided; (3) the CE method lacks a more macro perspective to reconcile the conflicts among different consensus results. To address these aforementioned problems, a dual-level clustering ensemble algorithm with three consensus strategies is proposed. Firstly, a backward clustering ensemble selection framework is devised, and its built-in selection strategy can adaptively eliminate redundant members. Then, at the base clustering consensus level, taking into account the interplay between actual spatial location information and the co-occurrence frequency, two modified relation matrices are reconstructed, resulting in the development of two consensus methods with different modes. Additionally, at the CE consensus level with a broader perspective, an adjustable Dempster–Shafer evidence theory is developed as the third consensus method in present algorithm to dynamically fuse multiple ensemble results. Experimental results demonstrate that compared to seven other state-of-the-art and typical CE algorithms, the proposed algorithm exhibits exceptional consensus ability and robustness.

Dual-granularity weighted ensemble clustering

In this paper, we study the ensemble clustering problem, where the input is in the form of multiple clustering solutions. The goal of ensemble clustering algorithms is to aggregate the solutions into one solution that maximizes the agreement in the input ensemble. We obtain several new results for this problem. Specifically, we show that the notion of agreement under such circumstances can be better captured using a 2D string encoding rather than a voting strategy, which is common among existing approaches. Our optimization proceeds by first constructing a non-linear objective function which is then transformed into a 0-1 Semidefinite program (SDP) using novel convexification techniques. This model can be subsequently relaxed to a polynomial time solvable SDP. In addition to the theoretical contributions, our experimental results on standard machine learning and synthetic datasets show that this approach leads to improvements not only in terms of the proposed agreement measure but also the existing agreement measures based on voting strategies. In addition, we identify several new application scenarios for this problem. These include combining multiple image segmentations and generating tissue maps from multiple-channel Diffusion Tensor brain images to identify the underlying structure of the brain.