Heterogeneity in medical data, e.g., from data collected at different sites and with different protocols in a clinical study, is a fundamental hurdle for accurate prediction using machine learning models, as such models often fail to generalize well. This paper presents a normalizing-flow-based method to perform counterfactual inference upon a structural causal model (SCM) to harmonize such data. We formulate a causal model for observed effects (brain magnetic resonance imaging data) that result from known confounders (site, gender and age) and exogenous noise variables. Our method exploits the bijection induced by flow for harmonization. We can infer the posterior of exogenous variables, intervene on observations, and draw samples from the resultant SCM to obtain counterfactuals. We evaluate on multiple, large, real-world medical datasets to observe that this method leads to better cross-domain generalization compared to state-of-the-art algorithms. Further experiments that evaluate the quality of confounder-independent data generated by our model using regression and classification tasks are provided.
This manuscript describes the first challenge on Federated Learning, namely the Federated Tumor Segmentation (FeTS) challenge 2021. International challenges have become the standard for validation of biomedical image analysis methods. However, the actual performance of participating (even the winning) algorithms on "real-world" clinical data often remains unclear, as the data included in challenges are usually acquired in very controlled settings at few institutions. The seemingly obvious solution of just collecting increasingly more data from more institutions in such challenges does not scale well due to privacy and ownership hurdles. Towards alleviating these concerns, we are proposing the FeTS challenge 2021 to cater towards both the development and the evaluation of models for the segmentation of intrinsically heterogeneous (in appearance, shape, and histology) brain tumors, namely gliomas. Specifically, the FeTS 2021 challenge uses clinically acquired, multi-institutional magnetic resonance imaging (MRI) scans from the BraTS 2020 challenge, as well as from various remote independent institutions included in the collaborative network of a real-world federation (https://www.fets.ai/). The goals of the FeTS challenge are directly represented by the two included tasks: 1) the identification of the optimal weight aggregation approach towards the training of a consensus model that has gained knowledge via federated learning from multiple geographically distinct institutions, while their data are always retained within each institution, and 2) the federated evaluation of the generalizability of brain tumor segmentation models "in the wild", i.e. on data from institutional distributions that were not part of the training datasets.
Resting-state fMRI has been shown to provide surrogate biomarkers for the analysis of various diseases. In addition, fMRI data helps in understanding the brain's functional working during resting state and task-induced activity. To improve the statistical power of biomarkers and the understanding mechanism of the brain, pooling of multi-center studies has become increasingly popular. But pooling the data from multiple sites introduces variations due to hardware, software, and environment. In this paper, we look at the estimation problem of hierarchical Sparsity Connectivity Patterns (hSCPs) in fMRI data acquired on multiple sites. We introduce a simple yet effective matrix factorization based formulation to reduce site-related effects while preserving biologically relevant variations. We leverage adversarial learning in the unsupervised regime to improve the reproducibility of the components. Experiments on simulated datasets display that the proposed method can estimate components with improved accuracy and reproducibility. We also demonstrate the improved reproducibility of the components while preserving age-related variation on a real dataset compiled from multiple sites.
The estimation of sparse hierarchical components reflecting patterns of the brain's functional connectivity from rsfMRI data can contribute to our understanding of the brain's functional organization, and can lead to biomarkers of diseases. However, inter-scanner variations and other confounding factors pose a challenge to the robust and reproducible estimation of functionally-interpretable brain networks, and especially to reproducible biomarkers. Moreover, the brain is believed to be organized hierarchically, and hence single-scale decompositions miss this hierarchy. The paper aims to use current advancements in adversarial learning to estimate interpretable hierarchical patterns in the human brain using rsfMRI data, which are robust to "adversarial effects" such as inter-scanner variations. We write the estimation problem as a minimization problem and solve it using alternating updates. Extensive experiments on simulation and a real-world dataset show high reproducibility of the components compared to other well-known methods.
Heterogeneity in medical imaging data is often tackled, in the context of machine learning, using domain invariance, i.e. deriving models that are robust to domain shifts, which can be both within domain (e.g. demographics) and across domains (e.g. scanner/protocol characteristics). However this approach can be detrimental to performance because it necessitates averaging across intra-class variability and reduces discriminatory power of learned models, in order to achieve better intra- and inter-domain generalization. This paper instead embraces the heterogeneity and treats it as a multi-task learning problem to explicitly adapt trained classifiers to both inter-site and intra-site heterogeneity. We demonstrate that the error of a base classifier on challenging 3D brain magnetic resonance imaging (MRI) datasets can be reduced by 2-3 times, in certain tasks, by adapting to the specific demographics of the patients, and different acquisition protocols. Learning the characteristics of domain shifts is achieved via auxiliary learning tasks leveraging commonly available data and variables, e.g. demographics. In our experiments, we use gender classification and age regression as auxiliary tasks helping the network weights trained on a source site adapt to data from a target site; we show that this approach improves classification accuracy by 5-30 % across different datasets on the main classification tasks, e.g. disease classification.
Deep Learning (DL) has greatly highlighted the potential impact of optimized machine learning in both the scientific and clinical communities. The advent of open-source DL libraries from major industrial entities, such as TensorFlow (Google), PyTorch (Facebook), and MXNet (Apache), further contributes to DL promises on the democratization of computational analytics. However, increased technical and specialized background is required to develop DL algorithms, and the variability of implementation details hinders their reproducibility. Towards lowering the barrier and making the mechanism of DL development, training, and inference more stable, reproducible, and scalable, without requiring an extensive technical background, this manuscript proposes the \textbf{G}ener\textbf{a}lly \textbf{N}uanced \textbf{D}eep \textbf{L}earning \textbf{F}ramework (GaNDLF). With built-in support for $k$-fold cross-validation, data augmentation, multiple modalities and output classes, and multi-GPU training, as well as the ability to work with both radiographic and histologic imaging, GaNDLF aims to provide an end-to-end solution for all DL-related tasks, to tackle problems in medical imaging and provide a robust application framework for deployment in clinical workflows.
Heterogeneity of brain diseases is a challenge for precision diagnosis/prognosis. We describe and validate Smile-GAN (SeMI-supervised cLustEring-Generative Adversarial Network), a novel semi-supervised deep-clustering method, which dissects neuroanatomical heterogeneity, enabling identification of disease subtypes via their imaging signatures relative to controls. When applied to MRIs (2 studies; 2,832 participants; 8,146 scans) including cognitively normal individuals and those with cognitive impairment and dementia, Smile-GAN identified 4 neurodegenerative patterns/axes: P1, normal anatomy and highest cognitive performance; P2, mild/diffuse atrophy and more prominent executive dysfunction; P3, focal medial temporal atrophy and relatively greater memory impairment; P4, advanced neurodegeneration. Further application to longitudinal data revealed two distinct progression pathways: P1$\rightarrow$P2$\rightarrow$P4 and P1$\rightarrow$P3$\rightarrow$P4. Baseline expression of these patterns predicted the pathway and rate of future neurodegeneration. Pattern expression offered better yet complementary performance in predicting clinical progression, compared to amyloid/tau. These deep-learning derived biomarkers offer promise for precision diagnostics and targeted clinical trial recruitment.
Conventional and deep learning-based methods have shown great potential in the medical imaging domain, as means for deriving diagnostic, prognostic, and predictive biomarkers, and by contributing to precision medicine. However, these methods have yet to see widespread clinical adoption, in part due to limited generalization performance across various imaging devices, acquisition protocols, and patient populations. In this work, we propose a new paradigm in which data from a diverse range of acquisition conditions are "harmonized" to a common reference domain, where accurate model learning and prediction can take place. By learning an unsupervised image to image canonical mapping from diverse datasets to a reference domain using generative deep learning models, we aim to reduce confounding data variation while preserving semantic information, thereby rendering the learning task easier in the reference domain. We test this approach on two example problems, namely MRI-based brain age prediction and classification of schizophrenia, leveraging pooled cohorts of neuroimaging MRI data spanning 9 sites and 9701 subjects. Our results indicate a substantial improvement in these tasks in out-of-sample data, even when training is restricted to a single site.
Registration of images with pathologies is challenging due to tissue appearance changes and missing correspondences caused by the pathologies. Moreover, mass effects as observed for brain tumors may displace tissue, creating larger deformations over time than what is observed in a healthy brain. Deep learning models have successfully been applied to image registration to offer dramatic speed up and to use surrogate information (e.g., segmentations) during training. However, existing approaches focus on learning registration models using images from healthy patients. They are therefore not designed for the registration of images with strong pathologies for example in the context of brain tumors, and traumatic brain injuries. In this work, we explore a deep learning approach to register images with brain tumors to an atlas. Our model learns an appearance mapping from images with tumors to the atlas, while simultaneously predicting the transformation to atlas space. Using separate decoders, the network disentangles the tumor mass effect from the reconstruction of quasi-normal images. Results on both synthetic and real brain tumor scans show that our approach outperforms cost function masking for registration to the atlas and that reconstructed quasi-normal images can be used for better longitudinal registrations.