Predicting Toxic Effects of Tigecycline Using Deep Learning
Author Information
Author(s): Xiong Yin, Liu Guoxin, Tang Xin, Xia Boyang, Yu Yalian, Fan Guangjun
Primary Institution: The Second Affiliated Hospital of Dalian Medical University
Hypothesis
This study aims to explore the potential risk factors for hepatotoxicity in patients treated with tigecycline using artificial intelligence technology.
Conclusion
The study found that hospitalization days of infected patients can be predicted with low error using a deep learning model, which is related to clinical test parameters and hepatotoxicity.
Supporting Evidence
- The degree of abnormal liver function was significantly correlated with hospitalization days.
- The AUC of the liver function prediction model reached 0.90.
- Multiple clinical laboratory parameters were significantly correlated with hospitalization days.
Takeaway
Doctors can use a computer program to guess how long patients will stay in the hospital after taking a medicine called tigecycline, which can sometimes hurt the liver.
Methodology
The study collected clinical data from 263 patients and established a hepatotoxicity prediction model using deep learning based on various clinical indicators.
Limitations
The study had a small sample size of patients with hepatotoxicity, which limits the interpretation of drug hepatotoxicity prediction.
Participant Demographics
Patients included were those with pulmonary infections treated with tigecycline, aged between 14 and 98 years.
Statistical Information
P-Value
0.001
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website