Jiannan Xu

Dr. Jiannan Xu


Research Focus & Metrics

Academic Career Length

7Years (2020-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
14 August 2025 in JMIR Medical Informatics
Verified
Journal Article
Open access
YHYuanheng Huang
Yuanheng Huang
GYGuozhen Yang
Guozhen Yang
YSYahui Shen
Yahui Shen
HCHuiguo Chen
Huiguo Chen
WWWeibin Wu
Weibin Wu
JZJian Zhang
Jian Zhang
Background: Large language models (LLMs) have made significant advancements in natural language processing (NLP) and are gradually showing potential for application in the medical field. However, LLMs still face challenges in medicine. Objective: This study aims to evaluate the efficiency, accuracy, and cost of LLMs in handling complex medical cases and to assess their potential and applicability as tools for clinical decision support. Methods: We selected cases from the database of the Department of Cardiothoracic Surgery, the Third Affiliated Hospital of Sun Yat-sen University (2021‐2024), and conducted a multidimensional preliminary evaluation of the latest LLMs in clinical decision-making for complex cases. The evaluation included measuring the time taken for the LLMs to generate decision recommendations, Likert scores, and calculating decision costs to assess the execution efficiency, accuracy, and cost-effectiveness of the models. Results: A total of 80 complex cases were included in this study, and the performance of multiple LLMs in clinical decision-making was evaluated. Experts required 33.60 minutes on average (95% CI 32.57‐34.63), far longer than any LLM. GPTo1 (0.71, 95% CI 0.67‐0.74), GPT4o (0.88, 95% CI 0.83‐0.92), and Deepseek (0.94, 95% CI 0.90‐0.96) all finished under a minute without statistical differences. Although Kimi, Gemini, LLaMa3-8B, and LLaMa3-70B took 1.02‐3.20 minutes, they were still faster than experts. In terms of decision accuracy, Deepseek-R1 had the highest accuracy (mean Likert score=4.19), with no significant difference compared to GPTo1 (P=.699), and both performed significantly better than GPT4o, Kimi, Gemini, LLaMa3-70B, and LLaMa3-8B (P<.001). Deepseek-R1 and GPTo1 demonstrated the lowest hallucination rates—6/80 (8%) and 5/80 (6%), respectively—significantly outperforming GPT-4o (7/80, 9%), Kimi (10/80, 12%), and the Gemini and LLaMa3 models, which exhibited substantially higher rates ranging from 13/80 (16%) to 25/80 (31%). Regarding decision costs, all LLMs showed significantly lower costs than the Multidisciplinary Team, with open-source models such as Deepseek-R1 offering a zero direct cost advantage. Conclusions: GPTo1 and Deepseek-R1 show strong clinical potential, boosting efficiency, maintaining accuracy, and reducing costs. GPT4o and Kimi performed moderately, indicating suitability for broader clinical tasks. Further research is needed to validate LLaMa3 series and Gemini in clinical decision.

11 January 2022 in Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering
Verified
Journal Article
Closed access
PDF file
FKFarong Kou
Farong Kou
XZXinqian Zhang
Xinqian Zhang
JXJiannan Xu
Jiannan Xu
Steering Angle is related to the design and optimization of steering mechanism and suspension, but it is not equal to the angle of knuckle around kingpin because of the existence of wheel alignment parameters. To calculate the steering angle, this paper derives based on homogeneous transformation its function expression by analyzing spatial geometric relation between the two angles and calculating coordinates related to steering trajectory of wheel center. Then, multi-body model of McPherson suspension with steering system is built and the calculation correctness is verified by comparing the function curve plotted by MATLAB software with the curve simulated by Adams/Car software. The calculation and simulation indicate that between the two angles, there is a ratio which is related to wheel alignment parameters and greater than 1.

5 October 2020 in Research Square
Verified
Preprint
Open access
PDF file
JXJiannan Xu
Jiannan Xu
GLGuangyao Liu
Guangyao Liu
XJXinghan Jin
Xinghan Jin
HWHuijin Wang
Huijin Wang
LZLihua Zhu
Lihua Zhu
YHYulong He
Yulong He
BackgroundThe incidence of gastric stump cancer (GSC) following resection of gastric cancer has increase. However, the definition between GSC and recurrent gastric cancer (RGC) is still being debated. This study was conducted to compare the clinicopathological characteristics and outcome for this two groups.Patients and MethodsThe investigators retrospectively reviewed patients in the Surveillance Epidemiology and End Results (SEER) databases from 1975 to 2016 to identify patients who underwent resection of gastric cancer subsequently experienced metachronous gastric cancer. GSC was defined as cancer occurring in remnant stomach of ≥10 years after gastrectomy, while RGC was defined as <10 years after gastrectomy. T-test, Pearson χ2-test and Kaplan-Meier estimator method was used in the present study.ResultsAmong the 167,747 gastric cancer patients, 1,006 (0.6%) patients were diagnosed multiple gastric cancer, 93 patients met the GSC criterion, and 282 patients met the RGC criterion. The mean time interval between the initial cancer and second cancer for GSC was 13.34 years and 4.24 years for RGC. Male gender (P=0.003), younger age (60.1 years old vs. 65.9 years old, P<0.001) developed more frequently in GSC group. The median OS in GSC group was 20.0 months compared with 16.0 months for RGC group (p=0.302). Surgery treatment or not, the median OS have statistically significant (in RGC group, 64.0 months vs. 10.0 months, P<0.001; 33 months vs. 13 months, P=0.014 in GSC group).ConclusionIn our definition of GSC and RGC, the OS were not statistically significant and gastrectomy may be the appropriate treatment for this two groups. It may demand for a more suitable definition and TNM staging system for GSC.

30 April 2026 in Communications Medicine
Verified
Journal Article
Open access
HYHaofan Yin
Haofan Yin
NZNingfang Zhang
Ningfang Zhang
LQLumei Qiu
Lumei Qiu
JZJiaju Zhang
Jiaju Zhang
JDJiahui Ding
Jiahui Ding
XYXiaopeng Yuan
Xiaopeng Yuan
Distinguishing malignant from benign pulmonary nodules remained a significant clinical challenge. Given the involvement of DNA methylation in anti-tumor immunity, we aimed to investigated whether DNA methylation patterns in peripheral blood mononuclear cells (PBMCs) could serve as non-invasive biomarkers for pulmonary nodule classification. Genome-wide DNA methylation profiling was performed using Methylome-seq on PBMCs from Discovery cohort, including patients with benign pulmonary nodules (BPN), minimally invasive adenocarcinoma (MIA), and early-stage invasive adenocarcinoma (eIAC). Differential analysis identified 56 candidates differentially methylated cytosines (DMCs) and regions (DMRs). Subsequently, 2 methylation features were validated using targeted bisulfite sequencing (TBS) in the Train cohort and Test cohort. We further developed 2 machine learning models integrating the DNA methylation features and routine blood biomarkers to enhance pulmonary nodule classification accuracy. SurgMalig Model was a surgical malignancy classifier used to differentiate between benign nodules and malignant nodules requiring surgical resection. And ResectGuide Model was a resection guide classifier used to subdivide malignant nodules into MIA and eIAC to inform surgical planning. The validated methylation features demonstrate significant discriminatory ability for pulmonary nodules properties compared to conventional blood biomarkers. SurgMalig Model achieves high performance in distinguishing benign from malignant nodules, with area under the curve (AUC) of 0.926 in the Train cohort and 0.806 in the Test cohort. ResectGuide Model effectively classifies MIA and eIAC subtypes, yielding AUC of 0.862 and 0.843 in the respective cohorts. PBMC methylation signatures provide clinically actionable tools for surgical triage in indeterminate pulmonary nodules.