Jiawei Luo

Jiawei Luo

Auto-generated profile, based on publicly available information

Jiawei Luo

West China Biomedical Big Data Center, Sichuan University, Sichuan, China

Research Focus & Metrics

Academic Career Length

8Years (2019-Present)

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

Representative Publications

View All Publications
23 April 2025 in Journal of Medical Internet Research
Verified
Journal Article
Open access
ZZZhuo Zheng
Zhuo Zheng
JLJiawei Luo
Jiawei Luo
YZYingchao Zhu
Yingchao Zhu
LDLei Du
Lei Du
LLLan Lan
Lan Lan
SHShixin Huang
Shixin Huang
Background: Timely and accurate prediction of short-term mortality is critical in intensive care units (ICUs), where patients’ conditions change rapidly. Traditional scoring systems, such as the Simplified Acute Physiology Score and Acute Physiology and Chronic Health Evaluation, rely on static variables collected within the first 24 hours of admission and do not account for continuously evolving clinical states. These systems lack real-time adaptability, interpretability, and generalizability. With the increasing availability of high-frequency electronic medical record (EMR) data, machine learning (ML) approaches have emerged as powerful tools to model complex temporal patterns and support dynamic clinical decision-making. However, existing models are often limited by their inability to handle irregular sampling and missing values, and many lack rigorous external validation across institutions. Objective: We aimed to develop a real-time, interpretable risk prediction model that continuously assesses ICU patient mortality using irregular, longitudinal EMR data, with improved performance and generalizability over traditional static scoring systems. Methods: A time-aware bidirectional attention-based long short-term memory (TBAL) model was developed using EMR data from the MIMIC-IV (Medical Information Mart for Intensive Care) and eICU Collaborative Research Database (eICU-CRD) databases, comprising 176,344 ICU stays. The model incorporated dynamic variables, including vital signs, laboratory results, and medication data, updated hourly, to perform static and continuous mortality risk assessments. External cross-validation and subgroup sensitivity analyses were conducted to evaluate robustness and fairness. Model performance was assessed using the area under the receiver operating characteristic curve (AUROC), area under the precision-recall curve (AUPRC), accuracy, and F1-score. Interpretability was enhanced using integrated gradients to identify key predictors. Results: For the static 12-hour to 1-day mortality prediction task, the TBAL model achieved AUROCs of 95.9 (95% CI 94.2-97.5) and 93.3 (95% CI 91.5-95.3) and AUPRCs of 48.5 and 21.6 in MIMIC-IV and eICU-CRD, respectively. Accuracy and F1-scores reached 94.1 and 46.7 in MIMIC-IV and 92.2 and 28.1 in eICU-CRD. In dynamic prediction tasks, AUROCs reached 93.6 (95% CI 93.2-93.9) and 91.9 (95% CI 91.6-92.1), with AUPRCs of 41.3 and 50, respectively. The model maintained high recall for positive cases (82.6% and 79.1% in MIMIC-IV and eICU-CRD). Cross-database validation yielded AUROCs of 81.3 and 76.1, confirming generalizability. Subgroup analysis showed stable performance across age, sex, and severity strata, with top predictors including lactate, vasopressor use, and Glasgow Coma Scale score. Conclusions: The TBAL model offers a robust, interpretable, and generalizable solution for dynamic real-time mortality risk prediction in ICU patients. Its ability to adapt to irregular temporal patterns and to provide hourly updated predictions positions it as a promising decision-support tool. Future work should validate its utility in prospective clinical trials and investigate its integration into real-world ICU workflows to enhance patient outcomes.

1 January 2019 in Brazilian Journal of Medical and Biological Research
Verified
Journal Article
Open access
PDF file
FHFujuan He
Fujuan He
XLXiaoqiang Li
Xiaoqiang Li
STSuman Thapa
Suman Thapa
CLChi Li
Chi Li
JLJiawei Luo
Jiawei Luo
JLJin Liu
Jin Liu
This study investigated the effects of tidal volume (TV) on the diagnostic value of pulse pressure variation (PPV) and the inferior vena cava dispensability index (IVC-DI) for volume responsiveness during mechanical ventilation. In patients undergoing elective surgery with mechanical ventilation, different TVs of 6, 9, and 12 mL/kg were given for two min. The left ventricular outflow tract velocity-time integral (VTI) was measured by transthoracic echocardiography. The IVC-DI was measured at sub-xyphoid transabdominal long axis. The PPV was measured via the radial artery and served as baseline. Index measurements were repeated after fluid challenge. VTI increased by more than 15% after fluid challenge, which was considered as volume responsive. Seventy-nine patients were enrolled, 38 of whom were considered positive volume responsive. Baseline data between the response group and the non-response group were similar. Receiver operating characteristic curve confirmed PPV accuracy in diagnosing an increase in volume responsiveness with increased TV. When TV was 12 mL/kg, the PPV area under the curve (AUC) was 0.93 and the threshold value was 15.5%. IVC-DI had the highest diagnostic accuracy at a TV of 9 mL/kg and an AUC of 0.79, with a threshold value of 15.3%. When TV increased to 12 mL/kg, the IVC-DI value decreased. When the TV was 9 and 12 mL/kg, PPV showed improved performance in diagnosing volume responsiveness than did IVC-DI. PPV diagnostic accuracy in mechanically ventilated patients was higher than IVC-DI. PPV accuracy in predicting volume responsiveness was increased by increasing TV.

25 September 2025 in Catheterization and Cardiovascular Interventions
Verified
Journal Article
Open access
TZTingxi Zhu
Tingxi Zhu
JLJiawei Luo
Jiawei Luo
XHXuan Huang
Xuan Huang
LLLulu Liu
Lulu Liu
KLKehan Li
Kehan Li
XYXiaoyan Yang
Xiaoyan Yang
Background: Transcatheter aortic valve replacement (TAVR) has become a popular alternative to surgical aortic valve replacement (SAVR) for patients with valvular heart disease (VHD), particularly for those with aortic anomalies. Aims: The study aimed to compare the risks of post‐TAVR complications between patients with bicuspid and tricuspid aortic valves and to identify associated risk factors. Methods: The association between complications and valve type (bicuspid or tricuspid) was assessed. The study also explored various combinations of factors to understand their impact on complications. Separate analyses were conducted to identify specific risk factors for each complication. Results: Out of the 1154 eligible patients, 508 had bicuspid aortic valves (BAV) and 646 had tricuspid aortic valves (TAV). The study identified 40 cases of acute kidney injury, 134 instances of new‐onset permanent pacemaker implantations, 129 occurrences of paravalvular leak, and 30 device successes. The comprehensive logistic regression revealed no statistically significant association between complications and valve type (ORs: 0.52 (95% CI, 0.23−1.09) for acute kidney injury, 1.29 (95% CI, 0.87−1.92) for permanent pacemaker implantation, 1.29 (95% CI, 0.82−2.02) for paravalvular leak, and 0.54 (95% CI, 0.22−1.23) for device success). Total bilirubin (TBIL), albumin (ALB), age, and New York Heart Association (NYHA) scores, among other factors, were associated with specific post‐TAVR complications. Conclusions: The incidence of acute kidney injury, new permanent pacemaker implantations, paravalvular leaks, and device success did not differ significantly between patients with BAV and TAV following TAVR. Specific risk factors for these complications were identified, highlighting the importance of careful clinical monitoring in post‐TAVR management.

1 January 2026 in IEEE Transactions on Knowledge and Data Engineering
Verified
Journal Article
Closed access
Pengfei ZhangPengfei Zhang
Pengfei Zhang
ZYZhong Yuan
Zhong Yuan
SYSiyi Yu
Siyi Yu
JLJiawei Luo
Jiawei Luo
XMXin Min
Xin Min
ZYZheng Yu
Zheng Yu

19 January 2026 in Expert Systems with Applications
Verified
Journal Article
Closed access
SHShixin Huang
Shixin Huang
JLJiawei Luo
Jiawei Luo
XWXiaoyu Wan
Xiaoyu Wan
XNXixi Nie
Xixi Nie

26 November 2024 in JMIR Preprints
Verified
Preprint
Open access
ZZZhuo Zheng
Zhuo Zheng
JLJiawei Luo
Jiawei Luo
YZYingchao Zhu
Yingchao Zhu
LDLei Du
Lei Du
LLLan Lan
Lan Lan
SHShixin Huang
Shixin Huang
BACKGROUND Timely and accurate prediction of short-term mortality is critical in intensive care units (ICUs), where patients’ conditions change rapidly. Traditional scoring systems, such as the Simplified Acute Physiology Score and Acute Physiology and Chronic Health Evaluation, rely on static variables collected within the first 24 hours of admission and do not account for continuously evolving clinical states. These systems lack real-time adaptability, interpretability, and generalizability. With the increasing availability of high-frequency electronic medical record (EMR) data, machine learning (ML) approaches have emerged as powerful tools to model complex temporal patterns and support dynamic clinical decision-making. However, existing models are often limited by their inability to handle irregular sampling and missing values, and many lack rigorous external validation across institutions. OBJECTIVE We aimed to develop a real-time, interpretable risk prediction model that continuously assesses ICU patient mortality using irregular, longitudinal EMR data, with improved performance and generalizability over traditional static scoring systems. METHODS A time-aware bidirectional attention-based long short-term memory (TBAL) model was developed using EMR data from the MIMIC-IV (Medical Information Mart for Intensive Care) and eICU Collaborative Research Database (eICU-CRD) databases, comprising 176,344 ICU stays. The model incorporated dynamic variables, including vital signs, laboratory results, and medication data, updated hourly, to perform static and continuous mortality risk assessments. External cross-validation and subgroup sensitivity analyses were conducted to evaluate robustness and fairness. Model performance was assessed using the area under the receiver operating characteristic curve (AUROC), area under the precision-recall curve (AUPRC), accuracy, and F1-score. Interpretability was enhanced using integrated gradients to identify key predictors. RESULTS For the static 12-hour to 1-day mortality prediction task, the TBAL model achieved AUROCs of 95.9 (95% CI 94.2-97.5) and 93.3 (95% CI 91.5-95.3) and AUPRCs of 48.5 and 21.6 in MIMIC-IV and eICU-CRD, respectively. Accuracy and F1-scores reached 94.1 and 46.7 in MIMIC-IV and 92.2 and 28.1 in eICU-CRD. In dynamic prediction tasks, AUROCs reached 93.6 (95% CI 93.2-93.9) and 91.9 (95% CI 91.6-92.1), with AUPRCs of 41.3 and 50, respectively. The model maintained high recall for positive cases (82.6% and 79.1% in MIMIC-IV and eICU-CRD). Cross-database validation yielded AUROCs of 81.3 and 76.1, confirming generalizability. Subgroup analysis showed stable performance across age, sex, and severity strata, with top predictors including lactate, vasopressor use, and Glasgow Coma Scale score. CONCLUSIONS The TBAL model offers a robust, interpretable, and generalizable solution for dynamic real-time mortality risk prediction in ICU patients. Its ability to adapt to irregular temporal patterns and to provide hourly updated predictions positions it as a promising decision-support tool. Future work should validate its utility in prospective clinical trials and investigate its integration into real-world ICU workflows to enhance patient outcomes. CLINICALTRIAL