Building an Ethical and Trustworthy Biomedical AI Ecosystem for the Translational and Clinical Integration of Foundational Models

Foundational Models (FMs) are emerging as the cornerstone of the biomedical AI ecosystem due to their ability to represent and contextualize multimodal biomedical data. These capabilities allow FMs to be adapted for various tasks, including biomedical reasoning, hypothesis generation, and clinical decision-making. This review paper examines the foundational components of an ethical and trustworthy AI (ETAI) biomedical ecosystem centered on FMs, highlighting key challenges and solutions. The ETAI biomedical ecosystem is defined by seven key components which collectively integrate FMs into clinical settings: Data Lifecycle Management, Data Processing, Model Development, Model Evaluation, Clinical Translation, AI Governance and Regulation, and Stakeholder Engagement. While the potential of biomedical AI is immense, it requires heightened ethical vigilance and responsibility. For instance, biases can arise from data, algorithms, and user interactions, necessitating techniques to assess and mitigate bias prior to, during, and after model development. Moreover, interpretability, explainability, and accountability are key to ensuring the trustworthiness of AI systems, while workflow transparency in training, testing, and evaluation is crucial for reproducibility. Safeguarding patient privacy and security involves addressing challenges in data access, cloud data privacy, patient re-identification, membership inference attacks, and data memorization. Additionally, AI governance and regulation are essential for ethical AI use in biomedicine, guided by global standards. Furthermore, stakeholder engagement is essential at every stage of the AI pipeline and lifecycle for clinical translation. By adhering to these principles, we can harness the transformative potential of AI and develop an ETAI ecosystem.

Explainable Biomedical Hypothesis Generation via Retrieval Augmented Generation enabled Large Language Models

The vast amount of biomedical information available today presents a significant challenge for investigators seeking to digest, process, and understand these findings effectively. Large Language Models (LLMs) have emerged as powerful tools to navigate this complex and challenging data landscape. However, LLMs may lead to hallucinatory responses, making Retrieval Augmented Generation (RAG) crucial for achieving accurate information. In this protocol, we present RUGGED (Retrieval Under Graph-Guided Explainable disease Distinction), a comprehensive workflow designed to support investigators with knowledge integration and hypothesis generation, identifying validated paths forward. Relevant biomedical information from publications and knowledge bases are reviewed, integrated, and extracted via text-mining association analysis and explainable graph prediction models on disease nodes, forecasting potential links among drugs and diseases. These analyses, along with biomedical texts, are integrated into a framework that facilitates user-directed mechanism elucidation as well as hypothesis exploration through RAG-enabled LLMs. A clinical use-case demonstrates RUGGED's ability to evaluate and recommend therapeutics for Arrhythmogenic Cardiomyopathy (ACM) and Dilated Cardiomyopathy (DCM), analyzing prescribed drugs for molecular interactions and unexplored uses. The platform minimizes LLM hallucinations, offers actionable insights, and improves the investigation of novel therapeutics.

CliBench: Multifaceted Evaluation of Large Language Models in Clinical Decisions on Diagnoses, Procedures, Lab Tests Orders and Prescriptions

The integration of Artificial Intelligence (AI), especially Large Language Models (LLMs), into the clinical diagnosis process offers significant potential to improve the efficiency and accessibility of medical care. While LLMs have shown some promise in the medical domain, their application in clinical diagnosis remains underexplored, especially in real-world clinical practice, where highly sophisticated, patient-specific decisions need to be made. Current evaluations of LLMs in this field are often narrow in scope, focusing on specific diseases or specialties and employing simplified diagnostic tasks. To bridge this gap, we introduce CliBench, a novel benchmark developed from the MIMIC IV dataset, offering a comprehensive and realistic assessment of LLMs' capabilities in clinical diagnosis. This benchmark not only covers diagnoses from a diverse range of medical cases across various specialties but also incorporates tasks of clinical significance: treatment procedure identification, lab test ordering and medication prescriptions. Supported by structured output ontologies, CliBench enables a precise and multi-granular evaluation, offering an in-depth understanding of LLM's capability on diverse clinical tasks of desired granularity. We conduct a zero-shot evaluation of leading LLMs to assess their proficiency in clinical decision-making. Our preliminary results shed light on the potential and limitations of current LLMs in clinical settings, providing valuable insights for future advancements in LLM-powered healthcare.

Xmodel-LM Technical Report

We introduce Xmodel-LM, a compact and efficient 1.1B language model pre-trained on over 2 trillion tokens. Trained on our self-built dataset (Xdata), which balances Chinese and English corpora based on downstream task optimization, Xmodel-LM exhibits remarkable performance despite its smaller size. It notably surpasses existing open-source language models of similar scale. Our model checkpoints and code are publicly accessible on GitHub at https://github.com/XiaoduoAILab/XmodelLM.

Harmonic and Interharmonic Detection in Power Systems Based on Fractal-Optimized Variational Mode Decomposition

The proposed method introduces a parameter determination approach based on the minimum Fractal box dimension (FBD) of Variational Mode Decomposition (VMD) components, aiming to address the issue of manual determination of VMD decomposition layers in advance. Initially, VMD is applied to the original power signal, and the layer number for VMD decomposition is determined by selecting the K value associated with the smallest fractal box dimension among its components. Subsequently, several Intrinsic Mode Functions (IMFs) are obtained as fundamental, harmonic, and interharmonic signals representing different aspects of the power system. Furthermore, Hilbert transform(HT) is employed to extract instantaneous amplitude and frequency information from these harmonic signals. Experimental evaluation using simulation data and real-world power system data demonstrates that compared to Empirical Mode Decomposition (EMD) and Ensemble Empirical Mode Decomposition (EEMD), our proposed method achieves more accurate identification and effective extraction of harmonic signals.

Scene Summarization: Clustering Scene Videos into Spatially Diverse Frames

We propose scene summarization as a new video-based scene understanding task. It aims to summarize a long video walkthrough of a scene into a small set of frames that are spatially diverse in the scene, which has many impotant applications, such as in surveillance, real estate, and robotics. It stems from video summarization but focuses on long and continuous videos from moving cameras, instead of user-edited fragmented video clips that are more commonly studied in existing video summarization works. Our solution to this task is a two-stage self-supervised pipeline named SceneSum. Its first stage uses clustering to segment the video sequence. Our key idea is to combine visual place recognition (VPR) into this clustering process to promote spatial diversity. Its second stage needs to select a representative keyframe from each cluster as the summary while respecting resource constraints such as memory and disk space limits. Additionally, if the ground truth image trajectory is available, our method can be easily augmented with a supervised loss to enhance the clustering and keyframe selection. Extensive experiments on both real-world and simulated datasets show our method outperforms common video summarization baselines by 50%

Enhancing Mobile Face Anti-Spoofing: A Robust Framework for Diverse Attack Types under Screen Flash

Face anti-spoofing (FAS) is crucial for securing face recognition systems. However, existing FAS methods with handcrafted binary or pixel-wise labels have limitations due to diverse presentation attacks (PAs). In this paper, we propose an attack type robust face anti-spoofing framework under light flash, called ATR-FAS. Due to imaging differences caused by various attack types, traditional FAS methods based on single binary classification network may result in excessive intra-class distance of spoof faces, leading to a challenge of decision boundary learning. Therefore, we employed multiple networks to reconstruct multi-frame depth maps as auxiliary supervision, and each network experts in one type of attack. A dual gate module (DGM) consisting of a type gate and a frame-attention gate is introduced, which perform attack type recognition and multi-frame attention generation, respectively. The outputs of DGM are utilized as weight to mix the result of multiple expert networks. The multi-experts mixture enables ATR-FAS to generate spoof-differentiated depth maps, and stably detects spoof faces without being affected by different types of PAs. Moreover, we design a differential normalization procedure to convert original flash frames into differential frames. This simple but effective processing enhances the details in flash frames, aiding in the generation of depth maps. To verify the effectiveness of our framework, we collected a large-scale dataset containing 12,660 live and spoof videos with diverse PAs under dynamic flash from the smartphone screen. Extensive experiments illustrate that the proposed ATR-FAS significantly outperforms existing state-of-the-art methods. The code and dataset will be available at https://github.com/Chaochao-Lin/ATR-FAS.

Duet: efficient and scalable hybriD neUral rElation undersTanding

Learned cardinality estimation methods have achieved high precision compared to traditional methods. Among learned methods, query-driven approaches face the data and workload drift problem for a long time. Although both query-driven and hybrid methods are proposed to avoid this problem, even the state-of-the-art of them suffer from high training and estimation costs, limited scalability, instability, and long-tailed distribution problem on high cardinality and high-dimensional tables, which seriously affects the practical application of learned cardinality estimators. In this paper, we prove that most of these problems are directly caused by the widely used progressive sampling. We solve this problem by introducing predicates information into the autoregressive model and propose Duet, a stable, efficient, and scalable hybrid method to estimate cardinality directly without sampling or any non-differentiable process, which can not only reduces the inference complexity from O(n) to O(1) compared to Naru and UAE but also achieve higher accuracy on high cardinality and high-dimensional tables. Experimental results show that Duet can achieve all the design goals above and be much more practical and even has a lower inference cost on CPU than that of most learned methods on GPU.

A Self-Paced Mixed Distillation Method for Non-Autoregressive Generation

Non-Autoregressive generation is a sequence generation paradigm, which removes the dependency between target tokens. It could efficiently reduce the text generation latency with parallel decoding in place of token-by-token sequential decoding. However, due to the known multi-modality problem, Non-Autoregressive (NAR) models significantly under-perform Auto-regressive (AR) models on various language generation tasks. Among the NAR models, BANG is the first large-scale pre-training model on English un-labeled raw text corpus. It considers different generation paradigms as its pre-training tasks including Auto-regressive (AR), Non-Autoregressive (NAR), and semi-Non-Autoregressive (semi-NAR) information flow with multi-stream strategy. It achieves state-of-the-art performance without any distillation techniques. However, AR distillation has been shown to be a very effective solution for improving NAR performance. In this paper, we propose a novel self-paced mixed distillation method to further improve the generation quality of BANG. Firstly, we propose the mixed distillation strategy based on the AR stream knowledge. Secondly, we encourage the model to focus on the samples with the same modality by self-paced learning. The proposed self-paced mixed distillation algorithm improves the generation quality and has no influence on the inference latency. We carry out extensive experiments on summarization and question generation tasks to validate the effectiveness. To further illustrate the commercial value of our approach, we conduct experiments on three generation tasks in real-world advertisements applications. Experimental results on commercial data show the effectiveness of the proposed model. Compared with BANG, it achieves significant BLEU score improvement. On the other hand, compared with auto-regressive generation method, it achieves more than 7x speedup.

Factorisation-based Image Labelling

Segmentation of brain magnetic resonance images (MRI) into anatomical regions is a useful task in neuroimaging. Manual annotation is time consuming and expensive, so having a fully automated and general purpose brain segmentation algorithm is highly desirable. To this end, we propose a patched-based label propagation approach based on a generative model with latent variables. Once trained, our Factorisation-based Image Labelling (FIL) model is able to label target images with a variety of image contrasts. We compare the effectiveness of our proposed model against the state-of-the-art using data from the MICCAI 2012 Grand Challenge and Workshop on Multi-Atlas Labeling. As our approach is intended to be general purpose, we also assess how well it can handle domain shift by labelling images of the same subjects acquired with different MR contrasts.