Jia Ding

Bo Wu, Sid Wang, Yunhao Tang et al.

Reinforcement Learning (RL) has become the most effective post-training approach for improving the capabilities of Large Language Models (LLMs). In practice, because of the high demands on latency and memory, it is particularly challenging to develop an efficient RL framework that reliably manages policy models with hundreds to thousands of billions of parameters. In this paper, we present LlamaRL, a fully distributed, asynchronous RL framework optimized for efficient training of large-scale LLMs with various model sizes (8B, 70B, and 405B parameters) on GPU clusters ranging from a handful to thousands of devices. LlamaRL introduces a streamlined, single-controller architecture built entirely on native PyTorch, enabling modularity, ease of use, and seamless scalability to thousands of GPUs. We also provide a theoretical analysis of LlamaRL's efficiency, including a formal proof that its asynchronous design leads to strict RL speed-up. Empirically during the Llama 3 post-training, by leveraging best practices such as colocated model offloading, asynchronous off-policy training, and distributed direct memory access for weight synchronization, LlamaRL achieves significant efficiency gains -- up to 10.7x speed-up compared to DeepSpeed-Chat-like systems on a 405B-parameter policy model. Furthermore, the efficiency advantage continues to grow with increasing model scale, demonstrating the framework's suitability for future large-scale RL training.

Dengbo Chen, Ziwei Zhao, Kexin Zhang et al.

Despite significant strides in medical foundation models, the ultrasound domain lacks a comprehensive solution capable of bridging low-level Ultrasound Grounded Perception (e.g., segmentation, localization) and high-level Ultrasound Comprehensive Interpretation (e.g., diagnosis, reasoning). To bridge this gap, we propose UMind-VL, a unified foundation model designed to synergize pixel-level structural understanding with complex clinical reasoning. We first introduce UMind-DS, a large-scale multimodal dataset comprising 1.2 million ultrasound image-text pairs across 16 anatomical regions, enriching standard data with pixel-level annotations and clinician-validated rationales. Architecturally, UMind-VL incorporates a lightweight Dynamic Convolutional Mask Decoder that generates masks via dynamic kernels conditioned on LLM outputs. This design, combined with task-specific tokens, unifies segmentation, detection, geometric measurement, and diagnosis tasks within a single framework. Extensive evaluations demonstrate that UMind-VL significantly outperforms existing generalist multimodal models and achieves performance on par with, or superior to, state-of-the-art specialist models across segmentation, detection, keypoint localization, and diagnostic reasoning benchmarks, while maintaining strong generalization ability. We demonstrate the capability of UMind-VL in Figure 1.

Lufan Chang, Wenjing Zhuang, Richeng Wu et al.

To conduct a radiomics or deep learning research experiment, the radiologists or physicians need to grasp the needed programming skills, which, however, could be frustrating and costly when they have limited coding experience. In this paper, we present DARWIN, a flexible research platform with a graphical user interface for medical imaging research. Our platform is consists of a radiomics module and a deep learning module. The radiomics module can extract more than 1000 dimension features(first-, second-, and higher-order) and provided many draggable supervised and unsupervised machine learning models. Our deep learning module integrates state of the art architectures of classification, detection, and segmentation tasks. It allows users to manually select hyperparameters, or choose an algorithm to automatically search for the best ones. DARWIN also offers the possibility for users to define a custom pipeline for their experiment. These flexibilities enable radiologists to carry out various experiments easily.

Dong Wang, Kexin Zhang, Jia Ding et al.

Skeletal bone age assessment (BAA), as an essential imaging examination, aims at evaluating the biological and structural maturation of human bones. In the clinical practice, Tanner and Whitehouse (TW2) method is a widely-used method for radiologists to perform BAA. The TW2 method splits the hands into Region Of Interests (ROI) and analyzes each of the anatomical ROI separately to estimate the bone age. Because of considering the analysis of local information, the TW2 method shows accurate results in practice. Following the spirit of TW2, we propose a novel model called Anatomical Local-Aware Network (ALA-Net) for automatic bone age assessment. In ALA-Net, anatomical local extraction module is introduced to learn the hand structure and extract local information. Moreover, we design an anatomical patch training strategy to provide extra regularization during the training process. Our model can detect the anatomical ROIs and estimate bone age jointly in an end-to-end manner. The experimental results show that our ALA-Net achieves a new state-of-the-art single model performance of 3.91 mean absolute error (MAE) on the public available RSNA dataset. Since the design of our model is well consistent with the well recognized TW2 method, it is interpretable and reliable for clinical usage.

Jia Ding, Aoxue Li, Zhiqiang Hu et al.

Early detection of pulmonary cancer is the most promising way to enhance a patient's chance for survival. Accurate pulmonary nodule detection in computed tomography (CT) images is a crucial step in diagnosing pulmonary cancer. In this paper, inspired by the successful use of deep convolutional neural networks (DCNNs) in natural image recognition, we propose a novel pulmonary nodule detection approach based on DCNNs. We first introduce a deconvolutional structure to Faster Region-based Convolutional Neural Network (Faster R-CNN) for candidate detection on axial slices. Then, a three-dimensional DCNN is presented for the subsequent false positive reduction. Experimental results of the LUng Nodule Analysis 2016 (LUNA16) Challenge demonstrate the superior detection performance of the proposed approach on nodule detection(average FROC-score of 0.891, ranking the 1st place over all submitted results).