AI Medical Lab

Twitter @
Facebook

HI THERE

Our team is made up of data analysts and clinicians who are working hard to combine cutting-edge AI technologies with clinical knowledge and translate these results into online models to help clinical practice.

Bioinformatics

A lncRNA-mRNA regulatory module of lung adenocarcinoma (LUAD)

Public gene expression data of three cohorts were downloaded from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO). Differential expression analysis between LUAD and normal samples and survival analysis were performed. The protein-protein interaction (PPI) network and co-expression analysis were conducted as a further screening. The least absolute shrinkage and selection operator (LASSO) Cox regression model was developed to predict overall survival. Machine learning models were trained to recognize early-stage LUAD or EGFR-mutation.

Version:

Pyroptosis-Related Molecular Subtypes of Lung Adenocarcinoma

33 pyroptosis gene expression profiles and clinical information were collected from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases. By bioinformatics and machine learning analyses, we identified novel subtypes of LUAD based on 10 pyroptosis-related genes and further validated them in the GEO dataset, with machine learning models performing up to an AUC of 1 for classifying in GEO. LUAD patients were clustered into 3 subtypes (A, B, and C), and survival analysis showed that B had the best survival outcome and C had the worst survival outcome.

Version:

Blood Transfusion

Risk prediction of RBC transfusion during or after liver transplantation

This model was successfully developed by using a machine learning algorithm, for predicting the risk of RBC transfusion during or after liver transplantation. Using seven preoperative variables, the model is convenient to use and performs well in predicting the risk of RBC transfusion and can guide high-risk patients to take appropriate preventive measures, before liver transplantation.

Liu LP*, Zhao QY*, Wu J, et al. Machine learning for the prediction of red blood cell transfusion in patients during or after liver transplantation surgery. Front Med; 2021. DOI: 10.3389/fmed.2021.632210.

Version:

Critical Care

ALICE Score

The ALICE score (Acute myocardial ischemia, Lactate, Iliac arteries involved, and CreatininE are comprised) is a more reliable preoperative score for detecting fatality-prone patients than previously published preoperative risk scores for acute type A aortic dissection (aTAAD) in external validation, which might help bedside clinicians in early detection of the most severe aTAAD patients. 

Luo JC*, Zhong J*, Duan WX*, et al. Early risk stratification of acute type A aortic dissection: development and validation of a predictive score. Cardiovasc Diagn Ther; 2020. DOI: 10.21037/cdt-20-730.

Version:

Prediction of sepsis-induced coagulopathy (SIC) in critically ill patients with sepsis

This model was developed based on the MIMIC-IV database and externally validated on the eICU-CRD database. 15 features were selected according to feature importance and clinical availability. It could differentiate septic patients who would and would not develop SIC, with a AUC of more than 0.80 in validation.

Zhao QY*, Liu LP*, Luo JC*, et al. A machine-learning approach for dynamic prediction of sepsis-induced coagulopathy in critically ill patients with sepsis. Front Med; 2020. DOI: 10.3389/fmed.2020.637434.

Version:

v1

Prediction of extubation failure

Extubation failure (EF) can lead to an increased chance of ventilator-associated pneumonia, longer hospital stays, and a higher mortality rate. This study aimed to develop and prospectively validate an accurate machine-learning model to predict EF in intensive care units (ICUs).

Zhao QY*, Wang H*, Luo JC*, et al. Development and Validation of a Machine-Learning Model for Prediction of Extubation Failure in Intensive Care Units. Front Med . 2021 May 17;8:676343. doi: 10.3389/fmed.2021.676343.

Version:

Prediction of NIV failure

Noninvasive mechanic ventilation (NIV) has been used widely in critical ill patients after extubation. However, if NIV is ineffective, it may delay the initiation of invasive mechanical ventilation (IMV) for some patients, which is associated with poor outcome. This study aims to determine early predictors of NIV failure and construct an accurate machine-learning model to identify patients at risks of NIV failure after extubation in intensive care units (ICUs).

Version:

Health

Prediction of Diabetes Mellitus

Informal version

Version:

AI Medical Lab