Sign in Subscribe
Machine learning

Machine Learning Model Predicts Multidrug-Resistant UTIs in Brain and Spinal Cord Injury Patients

Strategic Insights

— 2 min read

Investigations into multidrug-resistant urinary tract infections (MDR UTIs) in patients with brain and spinal cord injuries have led to the development of a machine learning model that aims to predict the risk of developing such infections. Researchers from the Korea Advanced Institute of Science and Technology (KAIST) have analyzed data from 849 patients with brain or spinal cord injuries to develop the model. The study found that important predictors of MDR UTIs included recent antibiotic use, total Functional Independence Measure score, neutrophil-to-platelet ratio, and other factors. The machine learning model, which combines Naive Bayes and Random Forest, achieved an accuracy of 0.7688 and can potentially support early MDR risk stratification in patients with brain and spinal cord injuries.

Key Takeaways:

  • Researchers from KAIST have developed a machine learning model that predicts the risk of multidrug-resistant urinary tract infections (MDR UTIs) in patients with brain and spinal cord injuries.
  • The model, which combines Naive Bayes and Random Forest, achieved an accuracy of 0.7688 in predicting MDR UTIs.
  • Important predictors of MDR UTIs included recent antibiotic use, total Functional Independence Measure score, neutrophil-to-platelet ratio, and other factors.
  • The model was trained and evaluated using data from 849 patients with brain or spinal cord injuries and was externally validated across institutions.
  • The study suggests that the model can support early MDR risk stratification in patients with brain and spinal cord injuries.
  • The model may guide antibiotic stewardship in neurorehabilitation settings and help identify at-risk individuals.

Statistics:

  • The machine learning model achieved an accuracy of 0.7688 in predicting MDR UTIs.
  • The model achieved a sensitivity of 0.7647, specificity of 0.7697, and positive predictive value of 0.4262.
  • The model achieved a negative predictive value of 0.9360 and an area under the receiver operating characteristic curve of 0.8425.
  • The model was trained and evaluated using data from 849 patients with brain or spinal cord injuries.
  • The study suggests that the model can be used to support early MDR risk stratification in patients with brain and spinal cord injuries.

Sources:

  • [International Journal of Medical Informatics, "Machine learning prediction of multidrug-resistant urinary tract infections in brain and spinal cord injury patients: a dual-center validation study", Vol. 206, 2025, p 106143].
  • [Elsevier Ireland Ltd, "International Journal of Medical Informatics", Elsevier House, Brookvale Plaza, East Park Shannon, Co, Clare, 00000, Ireland].
  • [Hangyul Yoon, Kim Jaechul Graduate School of AI, Korea Advanced Institute of Science and Technology (KAIST), Daejeon, South Korea].

Read more