Author: Kasra Arnavaz

Authors: Irene Balelli (INRIA)*; Santiago Silva (INRIA); Marco Lorenzi (INRIA)

Poster

Abstract: We propose a novel federated learning paradigm to model data variability among heterogeneous clients in multi-centric studies. Our method is expressed through a hierarchical Bayesian latent variable model, where client-specific parameters are assumed to be realization from a global distribution at the master level, which is in turn estimated to account for data bias and variability across clients. We show that our framework can be effectively optimized through expectation maximization over latent master’s distribution and clients’ parameters. We tested our method on the analysis of multi-modal medical imaging data and clinical scores from distributed clinical datasets of patients affected by Alzheimer’s disease. We demonstrate that our method is robust when data is distributed either in iid and non-iid manners: it allows to quantify the variability of data, views and centers, while guaranteeing high-quality data reconstruction as compared to the state-of-the-art autoencoding models and federated learning schemes.

Authors: Aiham Taleb (Digital Health Center, Hasso-plattner-institute)*; Christoph Lippert (Hasso Plattner Insitute for Digital Engineering, Universität Potsdam); Tassilo Klein (SAP); Moin Nabi (SAP)

Poster

Abstract:

Self-supervised learning approaches leverage unlabeled samples to acquire generic knowledge about different concepts, hence allowing for annotation-efficient downstream task learning.
In this paper, we propose a novel self-supervised method that leverages multiple imaging modalities.
We introduce the multimodal puzzle task, which facilitates representation learning from multiple image modalities. The learned modality-agnostic representations are obtained by confusing image modalities at the data-level.
Together with the Sinkhorn operator, with which we formulate the puzzle solving optimization as permutation matrix inference instead of classification, they allow for efficient solving of multimodal puzzles with varying levels of complexity.
In addition, we also propose to utilize generation techniques for multimodal data augmentation used for self-supervised pretraining, instead of downstream tasks directly. This aims to circumvent quality issues associated with synthetic images, while improving data-efficiency and the representations learned by self-supervised methods.
Our experimental results show that solving our multimodal puzzles yields better semantic representations, compared to treating each modality independently. Our results also highlight the benefits of exploiting synthetic images for self-supervised pretraining.
We showcase our approach on three segmentation tasks, and we outperform many solutions and our results are competitive to state-of-the-art.

Authors: Yuanyuan Lyu (Z2Sky Technologies Inc.)*; Haofu Liao (University of Rochester); Heqin Zhu (University of Science and Technology of China); S. Kevin Zhou (CAS)

Poster

Abstract:

Spinal surgery planning necessitates automatic segmentation of vertebrae in cone-beam computed tomography (CBCT), an intraoperative imaging modality that is widely used in intervention. However, CBCT images are of low-quality and artifact-laden due to noise, poor tissue contrast, and the presence of metallic objects, causing vertebra segmentation, even manually, a demanding task. In contrast, there exists a wealth of artifact-free, high quality CT images with vertebra annotations. This motivates us to build a CBCT vertebra segmentation model using unpaired CT images with annotations. To overcome the \textit{domain and artifact gaps} between CBCT and CT, it is a must to address the \textit{three heterogeneous tasks} of vertebra segmentation, artifact reduction and modality translation all together. To this, we propose a novel \textit{anatomy-aware artifact disentanglement and segmentation network} (\textbf{A$^3$DSegNet}) that intensively leverages knowledge sharing of these three tasks to promote learning. Specifically, it takes a random pair of CBCT and CT images as the input and manipulates the synthesis and segmentation via different decoding combinations from the disentangled latent layers. Then, by proposing various forms of consistency among the synthesized images and among segmented vertebrae, the learning is achieved without paired (i.e., anatomically identical) data. Finally, we stack 2D slices together and build 3D networks on top to obtain final 3D segmentation result. Extensive experiments on a large number of clinical CBCT (21,364) and CT (17,089) images show that the proposed \textbf{A$^3$DSegNet} performs significantly better than state-of-the-art
competing methods trained independently for each task and, remarkably, it achieves an average Dice coefficient of 0.926 for unpaired 3D CBCT vertebra segmentation.

Authors: Pierre-Emmanuel Poulet (Inria)*; Stanley Durrleman (INRIA)

Poster

Abstract: Disease modeling techniques summarize the possible trajectories of progression from multimodal and longitudinal data. These techniques often assume that individuals form a homogeneous cluster, thus ignoring possible disease subtypes within the population. We extend a non-linear mixed-effect model used for disease course mapping with a mixture framework. We jointly estimate model parameters and subtypes with a tempered version of a stochastic approximation of the Expectation Maximisation algorithm. We show that our model recovers the ground truth parameters from synthetically generated data, in contrast to the naive solution consisting in post hoc clustering of individual parameters from a one-class model. Applications to Alzheimer’s disease data allows the unsupervised identification of disease subtypes associated with regional atrophy and cognitive decline.

Authors: Axel Elaldi (New York University)*; Neel Dey (New York University); Heejong Kim (New York University); Guido Gerig (NYU)

Poster

Abstract: We present a rotation-equivariant self-supervised learning framework for the sparse deconvolution of non-negative scalar fields on the unit sphere. Spherical signals with multiple peaks naturally arise in Diffusion MRI (dMRI), where each voxel consists of one or more signal sources corresponding to anisotropic tissue structure such as white matter. Due to spatial and spectral partial voluming, clinically-feasible dMRI struggles to resolve crossing-fiber white matter configurations, leading to extensive development in spherical deconvolution methodology to recover underlying fiber directions. However, these methods are typically linear and struggle with small crossing-angles and partial volume fraction estimation. In this work, we improve on current methodologies by nonlinearly estimating fiber structures via self-supervised spherical convolutional networks with guaranteed equivariance to spherical rotation. We perform validation via extensive single and multi-shell synthetic benchmarks demonstrating competitive performance against common baselines. We further show improved downstream performance on fiber tractography measures on the Tractometer benchmark dataset. Finally, we show downstream improvements in terms of tractography and partial volume estimation on a multi-shell dataset of human subjects.

Authors: Huai Chen (Shanghai Jiao Tong University); Jieyu Li (Shanghai Jiao Tong University); Renzhen Wang (Xi’an Jiaotong University); Yi-Jie Huang (Shanghai Jiao Tong University); Fanrui Meng (Shanghai Jiaotong University); Deyu Meng (Xi’an Jiaotong University); Qing Peng (Department of Ophthalmology, Shanghai Tenth People’s Hospital, Tongji University, Shanghai); Lisheng Wang (Shanghai Jiao Tong University)*

Poster

Abstract: Local discriminative representation is needed in many medical image analysis tasks such as identifying sub-types of lesion or segmenting detailed components of anatomical structures. However, the commonly applied supervised representation learning methods require a large amount of annotated data, and unsupervised discriminative representation learning distinguishes different images by learning a global feature, both of which are not suitable for localized medical image analysis tasks. In order to avoid the limitations of these two methods, we introduce local discrimination into unsupervised representation learning in this work. The model contains two branches: one is an embedding branch which learns an embedding function to disperse dissimilar pixels over a low-dimensional hypersphere; and the other is a clustering branch which learns a clustering function to classify similar pixels into the same cluster. These two branches are trained simultaneously in a mutually beneficial pattern, and the learnt local discriminative representations are able to well measure the similarity of local image regions. These representations can be transferred to enhance various downstream tasks. Meanwhile, they can also be applied to cluster anatomical structures from unlabeled medical images under the guidance of topological priors from simulation or other structures with similar topological characteristics. The effectiveness and usefulness of the proposed method are demonstrated by enhancing various downstream tasks and clustering anatomical structures in retinal images and chest X-ray images.