Mianxin Liu

Mianxin Liu, Han Zhang, Feng Shi et al. · tsinghua

Functional connectivity network (FCN) data from functional magnetic resonance imaging (fMRI) is increasingly used for the diagnoses of brain disorders. However, state-of-the-art studies used to build the FCN using a single brain parcellation atlas at a certain spatial scale, which largely neglected functional interactions across different spatial scales in hierarchical manners. In this study, we propose a novel framework to perform multiscale FCN analysis for brain disorder diagnosis. We first use a set of well-defined multiscale atlases to compute multiscale FCNs. Then, we utilize biologically meaningful brain hierarchical relationships among the regions in multiscale atlases to perform nodal pooling across multiple spatial scales, namely "Atlas-guided Pooling". Accordingly, we propose a Multiscale-Atlases-based Hierarchical Graph Convolutional Network (MAHGCN), built on the stacked layers of graph convolution and the atlas-guided pooling, for a comprehensive extraction of diagnostic information from multiscale FCNs. Experiments on neuroimaging data from 1792 subjects demonstrate the effectiveness of our proposed method in the diagnoses of Alzheimer's disease (AD), the prodromal stage of AD (i.e., mild cognitive impairment [MCI]), as well as autism spectrum disorder (ASD), with accuracy of 88.9%, 78.6%, and 72.7% respectively. All results show significant advantages of our proposed method over other competing methods. This study not only demonstrates the feasibility of brain disorder diagnosis using resting-state fMRI empowered by deep learning, but also highlights that the functional interactions in the multiscale brain hierarchy are worth being explored and integrated into deep learning network architectures for better understanding the neuropathology of brain disorders.

Jingyang Zhang, Peng Xue, Ran Gu et al.

In clinical practice, a segmentation network is often required to continually learn on a sequential data stream from multiple sites rather than a consolidated set, due to the storage cost and privacy restriction. However, during the continual learning process, existing methods are usually restricted in either network memorizability on previous sites or generalizability on unseen sites. This paper aims to tackle the challenging problem of Synchronous Memorizability and Generalizability (SMG) and to simultaneously improve performance on both previous and unseen sites, with a novel proposed SMG-learning framework. First, we propose a Synchronous Gradient Alignment (SGA) objective, which not only promotes the network memorizability by enforcing coordinated optimization for a small exemplar set from previous sites (called replay buffer), but also enhances the generalizability by facilitating site-invariance under simulated domain shift. Second, to simplify the optimization of SGA objective, we design a Dual-Meta algorithm that approximates the SGA objective as dual meta-objectives for optimization without expensive computation overhead. Third, for efficient rehearsal, we configure the replay buffer comprehensively considering additional inter-site diversity to reduce redundancy. Experiments on prostate MRI data sequentially acquired from six institutes demonstrate that our method can simultaneously achieve higher memorizability and generalizability over state-of-the-art methods. Code is available at https://github.com/jingyzhang/SMG-Learning.

Zhongying Deng, Cheng Tang, Ziyan Huang et al. · pku

Foundation models have demonstrated remarkable success across diverse domains and tasks, primarily due to the thrive of large-scale, diverse, and high-quality datasets. However, in the field of medical imaging, the curation and assembling of such medical datasets are highly challenging due to the reliance on clinical expertise and strict ethical and privacy constraints, resulting in a scarcity of large-scale unified medical datasets and hindering the development of powerful medical foundation models. In this work, we present the largest survey to date of medical image datasets, covering over 1,000 open-access datasets with a systematic catalog of their modalities, tasks, anatomies, annotations, limitations, and potential for integration. Our analysis exposes a landscape that is modest in scale, fragmented across narrowly scoped tasks, and unevenly distributed across organs and modalities, which in turn limits the utility of existing medical image datasets for developing versatile and robust medical foundation models. To turn fragmentation into scale, we propose a metadata-driven fusion paradigm (MDFP) that integrates public datasets with shared modalities or tasks, thereby transforming multiple small data silos into larger, more coherent resources. Building on MDFP, we release an interactive discovery portal that enables end-to-end, automated medical image dataset integration, and compile all surveyed datasets into a unified, structured table that clearly summarizes their key characteristics and provides reference links, offering the community an accessible and comprehensive repository. By charting the current terrain and offering a principled path to dataset consolidation, our survey provides a practical roadmap for scaling medical imaging corpora, supporting faster data discovery, more principled dataset creation, and more capable medical foundation models.

Peng Xue, Jingyang Zhang, Lei Ma et al.

Image registration of liver dynamic contrast-enhanced computed tomography (DCE-CT) is crucial for diagnosis and image-guided surgical planning of liver cancer. However, intensity variations due to the flow of contrast agents combined with complex spatial motion induced by respiration brings great challenge to existing intensity-based registration methods. To address these problems, we propose a novel structure-aware registration method by incorporating structural information of related organs with segmentation-guided deep registration network. Existing segmentation-guided registration methods only focus on volumetric registration inside the paired organ segmentations, ignoring the inherent attributes of their anatomical structures. In addition, such paired organ segmentations are not always available in DCE-CT images due to the flow of contrast agents. Different from existing segmentation-guided registration methods, our proposed method extracts structural information in hierarchical geometric perspectives of line and surface. Then, according to the extracted structural information, structure-aware constraints are constructed and imposed on the forward and backward deformation field simultaneously. In this way, all available organ segmentations, including unpaired ones, can be fully utilized to avoid the side effect of contrast agent and preserve the topology of organs during registration. Extensive experiments on an in-house liver DCE-CT dataset and a public LiTS dataset show that our proposed method can achieve higher registration accuracy and preserve anatomical structure more effectively than state-of-the-art methods.

Mianxin Liu, Jingyang Zhang, Yao Wang et al.

Brain disorders in the early and late life of humans potentially share pathological alterations in brain functions. However, the key evidence from neuroimaging data for pathological commonness remains unrevealed. To explore this hypothesis, we build a deep learning model, using multi-site functional magnetic resonance imaging data (N=4,410, 6 sites), for classifying 5 different brain disorders from healthy controls, with a set of common features. Our model achieves 62.6(1.9)% overall classification accuracy on data from the 6 investigated sites and detects a set of commonly affected functional subnetworks at different spatial scales, including default mode, executive control, visual, and limbic networks. In the deep-layer feature representation for individual data, we observe young and aging patients with disorders are continuously distributed, which is in line with the clinical concept of the "spectrum of disorders". The revealed spectrum underlying early- and late-life brain disorders promotes the understanding of disorder comorbidities in the lifespan.

Yuxiao Liu, Mianxin Liu, Yuanwang Zhang et al.

The current clinical diagnosis framework of Alzheimer's disease (AD) involves multiple modalities acquired from multiple diagnosis stages, each with distinct usage and cost. Previous AD diagnosis research has predominantly focused on how to directly fuse multiple modalities for an end-to-end one-stage diagnosis, which practically requires a high cost in data acquisition. Moreover, a significant part of these methods diagnose AD without considering clinical guideline and cannot offer accurate sub-type diagnosis. In this paper, by exploring inter-correlation among multiple modalities, we propose a novel progressive AD sub-type diagnosis framework, aiming to give diagnosis results based on easier-to-access modalities in earlier low-cost stages, instead of modalities from all stages. Specifically, first, we design 1) a text disentanglement network for better processing tabular data collected in the initial stage, and 2) a modality fusion module for fusing multi-modality features separately. Second, we align features from modalities acquired in earlier low-cost stage(s) with later high-cost stage(s) to give accurate diagnosis without actual modality acquisition in later-stage(s) for saving cost. Furthermore, we follow the clinical guideline to align features at each stage for achieving sub-type diagnosis. Third, we leverage a progressive classifier that can progressively include additional acquired modalities (if needed) for diagnosis, to achieve the balance between diagnosis cost and diagnosis performance. We evaluate our proposed framework on large diverse public and in-home datasets (8280 in total) and achieve superior performance over state-of-the-art methods. Our codes will be released after the acceptance.

Caiwen Jiang, Xiaodan Xing, Zaixin Ou et al.

The progression of Idiopathic Pulmonary Fibrosis (IPF) significantly correlates with higher patient mortality rates. Early detection of IPF progression is critical for initiating timely treatment, which can effectively slow down the advancement of the disease. However, the current clinical criteria define disease progression requiring two CT scans with a one-year interval, presenting a dilemma: a disease progression is identified only after the disease has already progressed. To this end, in this paper, we develop a novel diffusion model to accurately predict the progression of IPF by generating patient's follow-up CT scan from the initial CT scan. Specifically, from the clinical prior knowledge, we tailor improvements to the traditional diffusion model and propose a Clinically-Informed Residual Diffusion model, called CIResDiff. The key innovations of CIResDiff include 1) performing the target region pre-registration to align the lung regions of two CT scans at different time points for reducing the generation difficulty, 2) adopting the residual diffusion instead of traditional diffusion to enable the model focus more on differences (i.e., lesions) between the two CT scans rather than the largely identical anatomical content, and 3) designing the clinically-informed process based on CLIP technology to integrate lung function information which is highly relevant to diagnosis into the reverse process for assisting generation. Extensive experiments on clinical data demonstrate that our approach can outperform state-of-the-art methods and effectively predict the progression of IPF.

Kaitao Chen, Mianxin Liu, Fang Yan et al.

The advent of vision-language models fosters the interactive conversations between AI-enabled models and humans. Yet applying these models into clinics must deal with daunting challenges around large-scale training data, financial, and computational resources. Here we propose a cost-effective instruction learning framework for conversational pathology named as CLOVER. CLOVER only trains a lightweight module and uses instruction tuning while freezing the parameters of the large language model. Instead of using costly GPT-4, we propose well-designed prompts on GPT-3.5 for building generation-based instructions, emphasizing the utility of pathological knowledge derived from the Internet source. To augment the use of instructions, we construct a high-quality set of template-based instructions in the context of digital pathology. From two benchmark datasets, our findings reveal the strength of hybrid-form instructions in the visual question-answer in pathology. Extensive results show the cost-effectiveness of CLOVER in answering both open-ended and closed-ended questions, where CLOVER outperforms strong baselines that possess 37 times more training parameters and use instruction data generated from GPT-4. Through the instruction tuning, CLOVER exhibits robustness of few-shot learning in the external clinical dataset. These findings demonstrate that cost-effective modeling of CLOVER could accelerate the adoption of rapid conversational applications in the landscape of digital pathology.

Shiyang Feng, Runmin Ma, Xiangchao Yan et al.

We introduce InternAgent-1.5, a unified system designed for end-to-end scientific discovery across computational and empirical domains. The system is built on a structured architecture composed of three coordinated subsystems for generation, verification, and evolution. These subsystems are supported by foundational capabilities for deep research, solution optimization, and long horizon memory. The architecture allows InternAgent-1.5 to operate continuously across extended discovery cycles while maintaining coherent and improving behavior. It also enables the system to coordinate computational modeling and laboratory experimentation within a single unified system. We evaluate InternAgent-1.5 on scientific reasoning benchmarks such as GAIA, HLE, GPQA, and FrontierScience, and the system achieves leading performance that demonstrates strong foundational capabilities. Beyond these benchmarks, we further assess two categories of discovery tasks. In algorithm discovery tasks, InternAgent-1.5 autonomously designs competitive methods for core machine learning problems. In empirical discovery tasks, it executes complete computational or wet lab experiments and produces scientific findings in earth, life, biological, and physical domains. Overall, these results show that InternAgent-1.5 provides a general and scalable framework for autonomous scientific discovery.

Caiwen Jiang, Tianyu Wang, Xiaodan Xing et al.

Ischemic stroke is a severe condition caused by the blockage of brain blood vessels, and can lead to the death of brain tissue due to oxygen deprivation. Thrombectomy has become a common treatment choice for ischemic stroke due to its immediate effectiveness. But, it carries the risk of postoperative cerebral hemorrhage. Clinically, multiple CT scans within 0-72 hours post-surgery are used to monitor for hemorrhage. However, this approach exposes radiation dose to patients, and may delay the detection of cerebral hemorrhage. To address this dilemma, we propose a novel prediction framework for measuring postoperative cerebral hemorrhage using only the patient's initial CT scan. Specifically, we introduce a dual-task mutual learning framework to takes the initial CT scan as input and simultaneously estimates both the follow-up CT scan and prognostic label to predict the occurrence of postoperative cerebral hemorrhage. Our proposed framework incorporates two attention mechanisms, i.e., self-attention and interactive attention. Specifically, the self-attention mechanism allows the model to focus more on high-density areas in the image, which are critical for diagnosis (i.e., potential hemorrhage areas). The interactive attention mechanism further models the dependencies between the interrelated generation and classification tasks, enabling both tasks to perform better than the case when conducted individually. Validated on clinical data, our method can generate follow-up CT scans better than state-of-the-art methods, and achieves an accuracy of 86.37% in predicting follow-up prognostic labels. Thus, our work thus contributes to the timely screening of post-thrombectomy cerebral hemorrhage, and could significantly reform the clinical process of thrombectomy and other similar operations related to stroke.

Zhengjie Zhang, Xiaoxie Mao, Qihao Guo et al.

Background: Alzheimer's disease (AD) diagnosis heavily relies on amyloid-beta positron emission tomography (Abeta-PET), which is limited by high cost and limited accessibility. This study explores whether Abeta-PET spatial patterns can be predicted from blood-based biomarkers (BBMs) and MRI scans. Methods: We collected Abeta-PET images, T1-weighted MRI scans, and BBMs from 566 participants. A language-enhanced generative model, driven by a large language model (LLM) and multimodal information fusion, was developed to synthesize PET images. Synthesized images were evaluated for image quality, diagnostic consistency, and clinical applicability within a fully automated diagnostic pipeline. Findings: The synthetic PET images closely resemble real PET scans in both structural details (SSIM = 0.920 +/- 0.003) and regional patterns (Pearson's r = 0.955 +/- 0.007). Diagnostic outcomes using synthetic PET show high agreement with real PET-based diagnoses (accuracy = 0.80). Using synthetic PET, we developed a fully automatic AD diagnostic pipeline integrating PET synthesis and classification. The synthetic PET-based model (AUC = 0.78) outperforms T1-based (AUC = 0.68) and BBM-based (AUC = 0.73) models, while combining synthetic PET and BBMs further improved performance (AUC = 0.79). Ablation analysis supports the advantages of LLM integration and prompt engineering. Interpretation: Our language-enhanced generative model synthesizes realistic PET images, enhancing the utility of MRI and BBMs for Abeta spatial pattern assessment and improving the diagnostic workflow for Alzheimer's disease.

Haoran Sun, Yankai Jiang, Wenjie Lou et al.

Multimodal large language models (MLLMs) have begun to demonstrate robust reasoning capabilities on general tasks, yet their application in the medical domain remains in its early stages. Constructing chain-of-thought (CoT) training data is essential for bolstering the reasoning abilities of medical MLLMs. However, existing approaches exhibit a deficiency in offering a comprehensive framework for searching and evaluating effective reasoning paths towards critical diagnosis. To address this challenge, we propose Mentor-Intern Collaborative Search (MICS), a novel reasoning-path searching scheme to generate rigorous and effective medical CoT data. MICS first leverages mentor models to initialize the reasoning, one step at a time, then prompts each intern model to continue the thinking along those initiated paths, and finally selects the optimal reasoning path according to the overall reasoning performance of multiple intern models. The reasoning performance is determined by an MICS-Score, which assesses the quality of generated reasoning paths. Eventually, we construct MMRP, a multi-task medical reasoning dataset with ranked difficulty, and Chiron-o1, a new medical MLLM devised via a curriculum learning strategy, with robust visual question-answering and generalizable reasoning capabilities. Extensive experiments demonstrate that Chiron-o1, trained on our CoT dataset constructed using MICS, achieves state-of-the-art performance across a list of medical visual question answering and reasoning benchmarks. Codes are available at https://github.com/manglu097/Chiron-o1

Kaitao Chen, Mianxin Liu, Daoming Zong et al.

Complex medical decision-making involves cooperative workflows operated by different clinicians. Designing AI multi-agent systems can expedite and augment human-level clinical decision-making. Existing multi-agent researches primarily focus on language-only tasks, yet their extension to multimodal scenarios remains challenging. A blind combination of diverse vision-language models (VLMs) can amplify an erroneous outcome interpretation. VLMs in general are less capable in instruction following and importantly self-reflection, compared to large language models (LLMs) of comparable sizes. This disparity largely constrains VLMs' ability in cooperative workflows. In this study, we propose MedOrch, a mediator-guided multi-agent collaboration framework for medical multimodal decision-making. MedOrch employs an LLM-based mediator agent that enables multiple VLM-based expert agents to exchange and reflect on their outputs towards collaboration. We utilize multiple open-source general-purpose and domain-specific VLMs instead of costly GPT-series models, revealing the strength of heterogeneous models. We show that the collaboration within distinct VLM-based agents can surpass the capabilities of any individual agent. We validate our approach on five medical vision question answering benchmarks, demonstrating superior collaboration performance without model training. Our findings underscore the value of mediator-guided multi-agent collaboration in advancing medical multimodal intelligence.

Jiamin Wu, Zichen Ren, Junyu Wang et al.

Non-invasive Brain-Computer Interfaces (BCI) offer a safe and accessible means of connecting the human brain to external devices, with broad applications in home and clinical settings to enhance human capabilities. However, the high noise level and limited task-specific data in non-invasive signals constrain decoding capabilities. Recently, the adoption of self-supervised pre-training is transforming the landscape of non-invasive BCI research, enabling the development of brain foundation models to capture generic neural representations from large-scale unlabeled electroencephalography (EEG) signals with substantial noises. However, despite these advances, the field currently lacks comprehensive, practical and extensible benchmarks to assess the utility of the public foundation models across diverse BCI tasks, hindering their widespread adoption. To address this challenge, we present AdaBrain-Bench, a large-scale standardized benchmark to systematically evaluate brain foundation models in widespread non-invasive BCI tasks. AdaBrain-Bench encompasses a diverse collection of representative BCI decoding datasets spanning 7 key applications. It introduces a streamlined task adaptation pipeline integrated with multi-dimensional evaluation metrics and a set of adaptation tools. The benchmark delivers an inclusive framework for assessing generalizability of brain foundation models across key transfer settings, including cross-subject, multi-subject, and few-shot scenarios. We leverage AdaBrain-Bench to evaluate a suite of publicly available brain foundation models and offer insights into practices for selecting appropriate models in various scenarios. We make our benchmark pipeline available to enable reproducible research and external use, offering a continuously evolving platform to foster progress toward robust and generalized neural decoding solutions.

Fang Yan, Jianfeng Wu, Jiawen Li et al.

The complexity and variability inherent in high-resolution pathological images present significant challenges in computational pathology. While pathology foundation models leveraging AI have catalyzed transformative advancements, their development demands large-scale datasets, considerable storage capacity, and substantial computational resources. Furthermore, ensuring their clinical applicability and generalizability requires rigorous validation across a broad spectrum of clinical tasks. Here, we present PathOrchestra, a versatile pathology foundation model trained via self-supervised learning on a dataset comprising 300K pathological slides from 20 tissue and organ types across multiple centers. The model was rigorously evaluated on 112 clinical tasks using a combination of 61 private and 51 public datasets. These tasks encompass digital slide preprocessing, pan-cancer classification, lesion identification, multi-cancer subtype classification, biomarker assessment, gene expression prediction, and the generation of structured reports. PathOrchestra demonstrated exceptional performance across 27,755 WSIs and 9,415,729 ROIs, achieving over 0.950 accuracy in 47 tasks, including pan-cancer classification across various organs, lymphoma subtype diagnosis, and bladder cancer screening. Notably, it is the first model to generate structured reports for high-incidence colorectal cancer and diagnostically complex lymphoma-areas that are infrequently addressed by foundational models but hold immense clinical potential. Overall, PathOrchestra exemplifies the feasibility and efficacy of a large-scale, self-supervised pathology foundation model, validated across a broad range of clinical-grade tasks. Its high accuracy and reduced reliance on extensive data annotation underline its potential for clinical integration, offering a pathway toward more efficient and high-quality medical services.

Shaohao Rui, Lingzhi Chen, Zhenyu Tang et al.

Self-supervised learning has greatly facilitated medical image analysis by suppressing the training data requirement for real-world applications. Current paradigms predominantly rely on self-supervision within uni-modal image data, thereby neglecting the inter-modal correlations essential for effective learning of cross-modal image representations. This limitation is particularly significant for naturally grouped multi-modal data, e.g., multi-parametric MRI scans for a patient undergoing various functional imaging protocols in the same study. To bridge this gap, we conduct a novel multi-modal image pre-training with three proxy tasks to facilitate the learning of cross-modality representations and correlations using multi-modal brain MRI scans (over 2.4 million images in 16,022 scans of 3,755 patients), i.e., cross-modal image reconstruction, modality-aware contrastive learning, and modality template distillation. To demonstrate the generalizability of our pre-trained model, we conduct extensive experiments on various benchmarks with ten downstream tasks. The superior performance of our method is reported in comparison to state-of-the-art pre-training methods, with Dice Score improvement of 0.28\%-14.47\% across six segmentation benchmarks and a consistent accuracy boost of 0.65\%-18.07\% in four individual image classification tasks.

Kaitao Chen, Shaohao Rui, Yankai Jiang et al.

Medical vision-language models (VLMs) excel at image-text understanding but typically rely on a single-pass reasoning that neglects localized visual cues. In clinical practice, however, human experts iteratively scan, focus, and refine the regions of interest before reaching a final diagnosis. To narrow this machine-human perception gap, we introduce ViTAR, a novel VLM framework that emulates the iterative reasoning process of human experts through a cognitive chain of "think-act-rethink-answer". ViTAR treats medical images as interactive objects, enabling models to engage multi-step visual reasoning. To support this approach, we curate a high-quality instruction dataset comprising 1K interactive examples that encode expert-like diagnostic behaviors. In addition, a 16K visual question answering training data has been curated towards fine-grained visual diagnosis. We introduce a two-stage training strategy that begins with supervised fine-tuning to guide cognitive trajectories, followed by the reinforcement learning to optimize decision-making. Extensive evaluations demonstrate that ViTAR outperforms strong state-of-the-art models. Visual attention analysis reveals that from the "think" to "rethink" rounds, ViTAR increasingly anchors visual grounding to clinically critical regions and maintains high attention allocation to visual tokens during reasoning, providing mechanistic insight into its improved performance. These findings demonstrate that embedding expert-style iterative thinking chains into VLMs enhances both performance and trustworthiness of medical AI.

Caiwen Jiang, Mianxin Liu, Kaicong Sun et al.

As a sensitive functional imaging technique, positron emission tomography (PET) plays a critical role in early disease diagnosis. However, obtaining a high-quality PET image requires injecting a sufficient dose (standard dose) of radionuclides into the body, which inevitably poses radiation hazards to patients. To mitigate radiation hazards, the reconstruction of standard-dose PET (SPET) from low-dose PET (LPET) is desired. According to imaging theory, PET reconstruction process involves multiple domains (e.g., projection domain and image domain), and a significant portion of the difference between SPET and LPET arises from variations in the noise levels introduced during the sampling of raw data as sinograms. In light of these two facts, we propose an end-to-end TriPle-domain LPET EnhancemenT (TriPLET) framework, by leveraging the advantages of a hybrid denoising-and-reconstruction process and a triple-domain representation (i.e., sinograms, frequency spectrum maps, and images) to reconstruct SPET images from LPET sinograms. Specifically, TriPLET consists of three sequentially coupled components including 1) a Transformer-assisted denoising network that denoises the inputted LPET sinograms in the projection domain, 2) a discrete-wavelet-transform-based reconstruction network that further reconstructs SPET from LPET in the wavelet domain, and 3) a pair-based adversarial network that evaluates the reconstructed SPET images in the image domain. Extensive experiments on the real PET dataset demonstrate that our proposed TriPLET can reconstruct SPET images with the highest similarity and signal-to-noise ratio to real data, compared with state-of-the-art methods.

Mianxin Liu, Jinru Ding, Jie Xu et al.

Ensuring the general efficacy and goodness for human beings from medical large language models (LLM) before real-world deployment is crucial. However, a widely accepted and accessible evaluation process for medical LLM, especially in the Chinese context, remains to be established. In this work, we introduce "MedBench", a comprehensive, standardized, and reliable benchmarking system for Chinese medical LLM. First, MedBench assembles the currently largest evaluation dataset (300,901 questions) to cover 43 clinical specialties and performs multi-facet evaluation on medical LLM. Second, MedBench provides a standardized and fully automatic cloud-based evaluation infrastructure, with physical separations for question and ground truth. Third, MedBench implements dynamic evaluation mechanisms to prevent shortcut learning and answer remembering. Applying MedBench to popular general and medical LLMs, we observe unbiased, reproducible evaluation results largely aligning with medical professionals' perspectives. This study establishes a significant foundation for preparing the practical applications of Chinese medical LLMs. MedBench is publicly accessible at https://medbench.opencompass.org.cn.