arXiv subjects

Jingsong Li

Publications and source records attributed to Jingsong Li.

8 recordsLinked to original sources

arXiv2026

Enhancing Trustworthy Clinical Diagnosis Decision-Making in Large Language Models via Etiology-Aware Attention Supervision

Objective: Large Language Models (LLMs) have demonstrated strong capabilities in medical text understanding and generation. However, their trustworthiness in diagnosis-oriented medical tasks remains constrained by the lack of structured guidance on how clinically relevant diagnostic evidence is internally attended to and utilized during model learning. Method: We propose an Etiology-Aware Attention Supervision framework that introduces structured etiological information as an external supervisory signal for training large language models. Specifically, we construct Clinical Etiology Schema (CES) derived from authoritative clinical guidelines for three acute abdominal conditions: acute appendicitis, acute pancreatitis, and acute cholecystitis. Based on CES annotations, we develop an Etiology-Aware Head Identification strategy to identify attention heads that consistently align with etiological evidence. Building on this analysis, we design a structure-guided parameter-efficient fine-tuning approach that steers attention distributions toward clinically relevant evidence through an additional supervision loss, without modifying the base model architecture. Result: Experiments conducted on a Consistent Diagnosis Cohort demonstrate that the proposed framework improves average diagnostic accuracy by 15.65% compared with baseline models. Attention-based metrics, including Inference Focus Score and Inference Attention Frequency, show more concentrated attention on etiologically relevant evidence. External evaluation on a Discrepant Diagnosis Cohort further confirms the robustness of diagnostic performance improvements under real-world clinical inconsistencies.

cs.CL↗

arXiv2026

Can LLM Agents Generate Real-World Evidence? Evaluating Observational Studies in Medical Databases

Observational studies can yield clinically actionable evidence at scale, but executing them on real-world databases is open-ended and requires coherent decisions across cohort construction, analysis, and reporting. Prior evaluations of LLM agents emphasize isolated steps or single answers, missing the integrity and internal structure of the resulting evidence bundle. To address this gap, we introduce RWE-bench, a benchmark grounded in MIMIC-IV and derived from peer-reviewed observational studies. Each task provides the corresponding study protocol as the reference standard, requiring agents to execute experiments in a real database and iteratively generate tree-structured evidence bundles. We evaluate six LLMs (three open-source, three closed-source) under three agent scaffolds using both question-level correctness and end-to-end task metrics. Across 162 tasks, task success is low: the best agent reaches 39.9%, and the best open-source model reaches 30.4%. Agent scaffolds also matter substantially, causing over 30% variation in performance metrics. Furthermore, we implement an automated cohort evaluation method to rapidly localize errors and identify agent failure modes. Overall, the results highlight persistent limitations in agents' ability to produce end-to-end evidence bundles, and efficient validation remains an important direction for future work. Code and data are available at https://github.com/somewordstoolate/RWE-bench.

cs.AI↗

arXiv2026

Streamlining evidence based clinical recommendations with large language models

Clinical evidence underpins informed healthcare decisions, yet integrating it into real-time practice remains challenging due to intensive workloads, complex procedures, and time constraints. This study presents Quicker, an LLM-powered system that automates evidence synthesis and generates clinical recommendations following standard guideline development workflows. Quicker delivers an end-to-end pipeline from clinical questions to recommendations and supports customized decision-making through integrated tools and interactive interfaces. To evaluate how closely Quicker can reproduce guideline development processes, we constructed Q2CRBench-3, a benchmark derived from guideline development records for three diseases. Experiments show that Quicker produces precise question decomposition, expert-aligned retrieval, and near-comprehensive screening. Quicker assistance improved the accuracy of extracted study data, and its recommendations were more comprehensive and coherent than clinician-written ones. In system-level testing, Quicker working with one participant reduced recommendation development to 20-40 min. Overall, the findings demonstrate Quicker's potential to enhance the speed and reliability of evidence-based clinical decision-making.

cs.CL↗

arXiv2023

Generating Synthetic Mixed-type Longitudinal Electronic Health Records for Artificial Intelligent Applications

The recent availability of electronic health records (EHRs) have provided enormous opportunities to develop artificial intelligence (AI) algorithms. However, patient privacy has become a major concern that limits data sharing across hospital settings and subsequently hinders the advances in AI. Synthetic data, which benefits from the development and proliferation of generative models, has served as a promising substitute for real patient EHR data. However, the current generative models are limited as they only generate single type of clinical data for a synthetic patient, i.e., either continuous-valued or discrete-valued. To mimic the nature of clinical decision-making which encompasses various data types/sources, in this study, we propose a generative adversarial network (GAN) entitled EHR-M-GAN which simultaneously synthesizes mixed-type timeseries EHR data. EHR-M-GAN is capable of capturing the multidimensional, heterogeneous, and correlated temporal dynamics in patient trajectories. We have validated EHR-M-GAN on three publicly-available intensive care unit databases with records from a total of 141,488 unique patients, and performed privacy risk evaluation of the proposed model. EHR-M-GAN has demonstrated its superiority over state-of-the-art benchmarks for synthesizing clinical timeseries with high fidelity, while addressing the limitations regarding data types and dimensionality in the current generative models. Notably, prediction models for outcomes of intensive care performed significantly better when training data was augmented with the addition of EHR-M-GAN-generated timeseries. EHR-M-GAN may have use in developing AI algorithms in resource-limited settings, lowering the barrier for data acquisition while preserving patient privacy.

cs.LG↗

arXiv2021

Multi-phase Liver Tumor Segmentation with Spatial Aggregation and Uncertain Region Inpainting

Multi-phase computed tomography (CT) images provide crucial complementary information for accurate liver tumor segmentation (LiTS). State-of-the-art multi-phase LiTS methods usually fused cross-phase features through phase-weighted summation or channel-attention based concatenation. However, these methods ignored the spatial (pixel-wise) relationships between different phases, hence leading to insufficient feature integration. In addition, the performance of existing methods remains subject to the uncertainty in segmentation, which is particularly acute in tumor boundary regions. In this work, we propose a novel LiTS method to adequately aggregate multi-phase information and refine uncertain region segmentation. To this end, we introduce a spatial aggregation module (SAM), which encourages per-pixel interactions between different phases, to make full use of cross-phase information. Moreover, we devise an uncertain region inpainting module (URIM) to refine uncertain pixels using neighboring discriminative features. Experiments on an in-house multi-phase CT dataset of focal liver lesions (MPCT-FLLs) demonstrate that our method achieves promising liver tumor segmentation and outperforms state-of-the-arts.

eess.IV↗

arXiv2020

Interactive Deep Refinement Network for Medical Image Segmentation

Deep learning techniques have successfully been employed in numerous computer vision tasks including image segmentation. The techniques have also been applied to medical image segmentation, one of the most critical tasks in computer-aided diagnosis. Compared with natural images, the medical image is a gray-scale image with low-contrast (even with some invisible parts). Because some organs have similar intensity and texture with neighboring organs, there is usually a need to refine automatic segmentation results. In this paper, we propose an interactive deep refinement framework to improve the traditional semantic segmentation networks such as U-Net and fully convolutional network. In the proposed framework, we added a refinement network to traditional segmentation network to refine the segmentation results.Experimental results with public dataset revealed that the proposed method could achieve higher accuracy than other state-of-the-art methods.

cs.CV↗

arXiv2017

A Circulating Biomarker-based Framework for Diagnosis of Hepatocellular Carcinoma in a Clinically Relevant Model of Non-alcoholic Steatohepatitis; An OAD to NASH

Although cirrhosis is a key risk factor for the development of hepatocellular carcinoma (HCC), mounting evidence indicates that in a subset of patients presenting with non-alcoholic steatohepatitis (NASH), HCC manifests in the absence of cirrhosis. Given the sheer size of the non-alcoholic fatty liver disease (NAFLD) epidemic, and the dismal prognosis associated with late-stage primary liver cancer, there is an urgent need for HCC surveillance in the NASH patient. In the present study, adult male mice randomized to control diet or a fast food diet (FFD) were followed for up to 14 mo and serum level of a panel of HCC-relevant biomarkers was compared with liver biopsies at 3 and 14 mo. Both NAFLD Activity Score (NAS) and hepatic hydroxyproline content were elevated at 3 and 14 mo on FFD. Picrosirius red staining of liver sections revealed a filigree pattern of fibrillar collagen deposition with no cirrhosis at 14 mo on FFD. Nevertheless, 46% of animals bore one or more tumors on their livers confirmed as HCC in hematoxylin-eosin-stained liver sections. Receiver operating characteristic (ROC) curves analysis for serum levels of the HCC biomarkers osteopontin (OPN), alpha-fetoprotein (AFP) and Dickkopf-1 (DKK1) returned concordance-statistic/area under ROC curve of > 0.89. These data suggest that serum levels of OPN (threshold, 218 ng/mL; sensitivity, 82%; specificity, 86%), AFP (136 ng/mL; 91%; 97%) and DKK1 (2.4 ng/mL; 82%; 81%) are diagnostic for HCC in a clinically relevant model of NASH

q-bio.TO↗

arXiv2016

Renal Parenchymal Area and Kidney Collagen Content

The extent of renal scarring in chronic kidney disease (CKD) can only be ascertained by highly invasive, painful and sometimes risky tissue biopsy. Interestingly, CKD-related abnormalities in kidney size can often be visualized using ultrasound. Nevertheless, not only does the ellipsoid formula used today underestimate true renal size but also the relation governing renal size and collagen content remains unclear. We used coronal kidney sections from healthy mice and mice with renal disease to develop a new technique for estimating the renal parenchymal area. While treating the kidney as an ellipse with the major axis the polar distance, this technique involves extending the minor axis into the renal pelvis. The calculated renal parenchymal area is remarkably similar to the measured area. Biochemically determined kidney collagen content revealed a strong and positive correlation with the calculated renal parenchymal area. The extent of renal scarring, i.e. kidney collagen content, can now be computed by making just two renal axial measurements which can easily be accomplished via noninvasive imaging of this organ.

q-bio.TO↗