Tammy Riklin-Raviv

Tal Ben-Haim, Ron Moshe Sofer, Gal Ben-Arie et al.

We present a novel deep learning approach to categorical segmentation of lung CTs of COVID-19 patients. Specifically, we partition the scans into healthy lung tissues, non-lung regions, and two different, yet visually similar, pathological lung tissues, namely, ground-glass opacity and consolidation. This is accomplished via a unique, end-to-end hierarchical network architecture and ensemble learning, which contribute to the segmentation and provide a measure for segmentation uncertainty. The proposed framework achieves competitive results and outstanding generalization capabilities for three COVID-19 datasets. Our method is ranked second in a public Kaggle competition for COVID-19 CT images segmentation. Moreover, segmentation uncertainty regions are shown to correspond to the disagreements between the manual annotations of two different radiologists. Finally, preliminary promising correspondence results are shown for our private dataset when comparing the patients' COVID-19 severity scores (based on clinical measures), and the segmented lung pathologies. Code and data are available at our repository: https://github.com/talbenha/covid-seg

Thomas Mendelson, Joshua Francois, Galit Lahav et al.

Accurate delineation of individual cells in microscopy videos is essential for studying cellular dynamics, yet separating touching or overlapping instances remains a persistent challenge. Although foundation-model for segmentation such as SAM have broadened the accessibility of image segmentation, they still struggle to separate nearby cell instances in dense microscopy scenes without extensive prompting. We propose a prompt-free, boundary-aware instance segmentation framework that predicts signed distance functions (SDFs) instead of binary masks, enabling smooth and geometry-consistent modeling of cell contours. A learned sigmoid mapping converts SDFs into probability maps, yielding sharp boundary localization and robust separation of adjacent instances. Training is guided by a unified Modified Hausdorff Distance (MHD) loss that integrates region- and boundary-based terms. Evaluations on both public and private high-throughput microscopy datasets demonstrate improved boundary accuracy and instance-level performance compared to recent SAM-based and foundation-model approaches. Source code is available at: https://github.com/ThomasMendelson/BAISeg.git

Doron Serebro, Tammy Riklin-Raviv

Medical imaging datasets often vary due to differences in acquisition protocols, patient demographics, and imaging devices. These variations in data distribution, known as domain shift, present a significant challenge in adapting imaging analysis models for practical healthcare applications. Most current domain adaptation (DA) approaches aim either to align the distributions between the source and target domains or to learn an invariant feature space that generalizes well across all domains. However, both strategies require access to a sufficient number of examples, though not necessarily annotated, from the test domain during training. This limitation hinders the widespread deployment of models in clinical settings, where target domain data may only be accessible in real time. In this work, we introduce HyDA, a novel hypernetwork framework that leverages domain characteristics rather than suppressing them, enabling dynamic adaptation at inference time. Specifically, HyDA learns implicit domain representations and uses them to adjust model parameters on-the-fly, effectively interpolating to unseen domains. We validate HyDA on two clinically relevant applications - MRI brain age prediction and chest X-ray pathology classification - demonstrating its ability to generalize across tasks and modalities. Our code is available at TBD.

Itay Benou, Tammy Riklin-Raviv

We present DeepTract, a deep-learning framework for estimating white matter fibers orientation and streamline tractography. We adopt a data-driven approach for fiber reconstruction from diffusion weighted images (DWI), which does not assume a specific diffusion model. We use a recurrent neural network for mapping sequences of DWI values into probabilistic fiber orientation distributions. Based on these estimations, our model facilitates both deterministic and probabilistic streamline tractography. We quantitatively evaluate our method using the Tractometer tool, demonstrating competitive performance with state-of-the art classical and machine learning based tractography algorithms. We further present qualitative results of bundle-specific probabilistic tractography obtained using our method. The code is publicly available at: https://github.com/itaybenou/DeepTract.git.

Itay Benou, Tammy Riklin-Raviv

Modern deep neural networks have now reached human-level performance across a variety of tasks. However, unlike humans they lack the ability to explain their decisions by showing where and telling what concepts guided them. In this work, we present a unified framework for transforming any vision neural network into a spatially and conceptually interpretable model. We introduce a spatially-aware concept bottleneck layer that projects "black-box" features of pre-trained backbone models into interpretable concept maps, without requiring human labels. By training a classification layer over this bottleneck, we obtain a self-explaining model that articulates which concepts most influenced its prediction, along with heatmaps that ground them in the input image. Accordingly, we name this method "Spatially-Aware and Label-Free Concept Bottleneck Model" (SALF-CBM). Our results show that the proposed SALF-CBM: (1) Outperforms non-spatial CBM methods, as well as the original backbone, on a variety of classification tasks; (2) Produces high-quality spatial explanations, outperforming widely used heatmap-based methods on a zero-shot segmentation task; (3) Facilitates model exploration and debugging, enabling users to query specific image regions and refine the model's decisions by locally editing its concept maps.

Yael Ben-Guigui, Jacob Goldberger, Tammy Riklin-Raviv

The pressing need to reduce the capacity of deep neural networks has stimulated the development of network dilution methods and their analysis. While the ability of $L_1$ and $L_0$ regularization to encourage sparsity is often mentioned, $L_2$ regularization is seldom discussed in this context. We present a novel framework for weight pruning by sampling from a probability function that favors the zeroing of smaller weights. In addition, we examine the contribution of $L_1$ and $L_2$ regularization to the dynamics of node pruning while optimizing for weight pruning. We then demonstrate the effectiveness of the proposed stochastic framework when used together with a weight decay regularizer on popular classification models in removing 50% of the nodes in an MLP for MNIST classification, 60% of the filters in VGG-16 for CIFAR10 classification, and on medical image models in removing 60% of the channels in a U-Net for instance segmentation and 50% of the channels in CNN model for COVID-19 detection. For these node-pruned networks, we also present competitive weight pruning results that are only slightly less accurate than the original, dense networks.

Or Shwartzman, Harel Gazit, Ilan Shelef et al.

The problem of inter-rater variability is often discussed in the context of manual labeling of medical images. The emergence of data-driven approaches such as Deep Neural Networks (DNNs) brought this issue of raters' disagreement to the front-stage. In this paper, we highlight the issue of inter-rater bias as opposed to random inter-observer variability and demonstrate its influence on DNN training, leading to different segmentation results for the same input images. In fact, lower overlap scores are obtained between the outputs of a DNN trained on annotations of one rater and tested on another. Moreover, we demonstrate that inter-rater bias in the training examples is amplified and becomes more consistent, considering the segmentation predictions of the DNNs' test data. We support our findings by showing that a classifier-DNN trained to distinguish between raters based on their manual annotations performs better when the automatic segmentation predictions rather than the actual raters' annotations were tested. For this study, we used two different datasets: the ISBI 2015 Multiple Sclerosis (MS) challenge dataset, including MRI scans each with annotations provided by two raters with different levels of expertise; and Intracerebral Hemorrhage (ICH) CT scans with manual and semi-manual segmentations. The results obtained allow us to underline a worrisome clinical implication of a DNN bias induced by an inter-rater bias during training. Specifically, we present a consistent underestimate of MS-lesion loads when calculated from segmentation predictions of a DNN trained on input provided by the less experienced rater. In the same manner, the differences in ICH volumes calculated based on outputs of identical DNNs, each trained on annotations from a different source are more consistent and larger than the differences in volumes between the manual and semi-manual annotations used for training.