This paper introduces a multi-task learning approach for fetal biometric estimation from ultrasound images, classifying regions (head, abdomen, femur) and estimating parameters. The model, a U-Net architecture with a classification head, achieved a mean absolute error of 1.08 mm for head circumference, 1.44 mm for abdomen circumference, and 1.10 mm for femur length, with 99.91% classification accuracy. The researchers are affiliated with MBZUAI. Why it matters: This research demonstrates advancements in automated fetal health monitoring using AI, potentially improving prenatal care and diagnostics in the region.
A new study uses the UNet++ deep learning model and Sentinel-2 satellite data to monitor mangrove dynamics in the UAE from 2017 to 2024. The model achieved a mean Intersection over Union (mIoU) of 87.8% on the validation set. Results indicate a significant increase in mangrove area, primarily in Abu Dhabi, contributing to enhanced carbon sequestration across the UAE.
This paper introduces a self-supervised contrastive learning method for segmenting the left ventricle in echocardiography images when limited labeled data is available. The approach uses contrastive pretraining to improve the performance of UNet and DeepLabV3 segmentation networks. Experiments on the EchoNet-Dynamic dataset show the method achieves a Dice score of 0.9252, outperforming existing approaches, with code available on Github.
A new brain tumor segmentation method based on convolutional neural networks is proposed for the BraTS-GoAT challenge. The method employs the MedNeXt architecture and model ensembling to segment tumors in brain MRI scans from diverse populations. Experiments on the unseen validation set demonstrate promising results with an average DSC of 85.54%.
The paper introduces the Unscented Autoencoder (UAE), a novel deep generative model based on the Variational Autoencoder (VAE) framework. The UAE uses the Unscented Transform (UT) for a more informative posterior representation compared to the reparameterization trick in VAEs. It replaces Kullback-Leibler (KL) divergence with the Wasserstein distribution metric and demonstrates competitive performance in Fréchet Inception Distance (FID) scores.
The paper introduces the Prism Hypothesis, which posits a correspondence between an encoder's feature spectrum and its functional role, with semantic encoders capturing low-frequency components and pixel encoders retaining high-frequency information. Based on this, the authors propose Unified Autoencoding (UAE), a model that harmonizes semantic structure and pixel details using a frequency-band modulator. Experiments on ImageNet and MS-COCO demonstrate that UAE effectively unifies semantic abstraction and pixel-level fidelity, achieving state-of-the-art performance.
Pascal Fua from EPFL presented an approach to implementing convolutional neural nets that output complex 3D surface meshes. The method overcomes limitations in converting implicit representations to explicit surface representations. Applications include single view reconstruction, physically-driven shape optimization, and bio-medical image segmentation. Why it matters: This research advances geometric deep learning by enabling end-to-end trainable models for 3D surface mesh generation, with potential impact on various applications in computer vision and biomedical imaging in the region.