In this study, as a proof-of-concept, we aim to initiate the development of Radiology Foundation Model, termed as RadFM. We consider three perspectives: dataset construction, model design, and thorough evaluation, concluded as follows: (i), we contribute 4 multimodal datasets with 13M 2D images and 615K 3D scans. When combined with a vast collection of existing datasets, this forms our training dataset, termed as Medical Multi-modal Dataset, MedMD. (ii), we propose an architecture that enables to integrate text input with 2D or 3D medical scans, and generates responses for diverse radiologic tasks, including diagnosis, visual question answering, report generation, and rationale diagnosis; (iii), beyond evaluation on 9 existing datasets, we propose a new benchmark, RadBench, comprising three tasks aiming to assess foundation models comprehensively. We conduct both automatic and human evaluations on RadBench. RadFM outperforms former accessible multi-modal foundation models, including GPT-4V. Additionally, we adapt RadFM for diverse public benchmarks, surpassing various existing SOTAs.

The development of open-source, multilingual medical language models can benefit a wide, linguistically diverse audience from different regions. To promote this domain, we present contributions from the following: First, we construct a multilingual medical corpus, containing approximately 25.5B tokens encompassing 6 main languages, termed as MMedC, enabling auto-regressive domain adaptation for general LLMs; Second, to monitor the development of multilingual medical LLMs, we propose a multilingual medical multi-choice question-answering benchmark with rationale, termed as MMedBench; Third, we have assessed a number of open-source large language models (LLMs) on our benchmark, along with those further auto-regressive trained on MMedC. Our final model, MMed-Llama 3, with only 8B parameters, achieves superior performance compared to all other open-source models on both MMedBench and English benchmarks, even rivaling GPT-4. In conclusion, in this work, We present a large-scale corpus, a benchmark and a series of models to support the development of multilingual medical LLMs.

3D skeleton-based action recognition and motion prediction are two essential problems of human activity understanding. In many previous works: 1) they studied two tasks separately, neglecting internal correlations; and 2) they did not capture sufficient relations inside the body. To address these issues, we propose a symbiotic model to handle two tasks jointly; and we propose two scales of graphs to explicitly capture relations among body-joints and body-parts. Together, we propose symbiotic graph neural networks, which contain a backbone, an action-recognition head, and a motion-prediction head. Two heads are trained jointly and enhance each other. For the backbone, we propose multi-branch multiscale graph convolution networks to extract spatial and temporal features. The multiscale graph convolution networks are based on joint-scale and part-scale graphs. The joint-scale graphs contain actional graphs, capturing action-based relations, and structural graphs, capturing physical constraints. The part-scale graphs integrate body-joints to form specific parts, representing high-level relations. Moreover, dual bone-based graphs and networks are proposed to learn complementary features. We conduct extensive experiments for skeleton-based action recognition and motion prediction with four datasets, NTU-RGB+D, Kinetics, Human3.6M, and CMU Mocap. Experiments show that our symbiotic graph neural networks achieve better performances on both tasks compared to the state-of-the-art methods.