Authors: Sergio Martínez-Agüero, Antonio G. Marques, Inmaculada Mora-Jiménez, Joaquín Alvárez-Rodríguez, Cristina Soguero-Ruiza
Published on: February 09, 2024
Impact Score: 8.35
Arxiv code: Arxiv:2402.06295
Summary
- What is new: An approach using interpretable multimodal data-driven models to predict the emergence of antimicrobial multidrug resistance (AMR) in the ICU.
- Why this is important: Need for improved interpretability in deep neural network models for clinical prediction, especially for understanding and predicting AMR germs in the ICU.
- What the research proposes: A collection of interpretable multimodal DNN models that integrate static patient data with multivariate time series (MTS) for enhanced prediction of AMR.
- Results: Effective prediction of AMR in the ICU and provision of an explainable prediction support system, with potential applications in other clinical EHR data problems.
Technical Details
Technological frameworks used: Deep Neural Networks (DNNs)
Models used: Interpretable multimodal DNN models
Data used: Electronic health records (EHR) including static data and MTS like mechanical ventilation and antibiotics intake
Potential Impact
Healthcare providers, especially ICU departments in hospitals, health technology companies specializing in EHR systems, and firms focusing on AI in healthcare.
Want to implement this idea in a business?
We have generated a startup concept here: IntelliHealth AMR Predictor.
Leave a Reply