| Autor: |
Trottet, Cécile, Vogels, Thijs, Keitel, Kristina, Kulinkina, Alexandra V, Tan, Rainer, Cobuccio, Ludovico, Jaggi, Martin, Hartley, Mary-Anne |
| Rok vydání: |
2022 |
| Předmět: |
|
| Zdroj: |
Trottet, Cécile; Vogels, Thijs; Keitel, Kristina; Kulinkina, Alexandra V; Tan, Rainer; Cobuccio, Ludovico; Jaggi, Martin; Hartley, Mary-Anne (2023). Modular Clinical Decision Support Networks (MoDN)-Updatable, interpretable, and portable predictions for evolving clinical environments. PLOS digital health, 2(7), e0000108. Public Library of Science 10.1371/journal.pdig.0000108 |
| DOI: |
10.1101/2022.08.17.22278908 |
| Popis: |
BackgroundClinical Decision Support Systems (CDSS) have the potential to improve and standardise care with probabilistic guidance. However, many CDSS deploy static, generic rule-based logic, resulting in inequitably distributed accuracy and inconsistent performance in evolving clinical environments. Data-driven models could resolve this issue by updating predictions according to the data collected. However, the size of data required necessitates collaborative learning from analogous CDSS’s, which are often imperfectly interoperable (IIO) or unshareable. We propose Modular Clinical Decision Support Networks (MoDN) which allow flexible, privacy-preserving learning across IIO datasets as well as being robust to the systematic missingness common to CDSS-derived data, while providing interpretable, continuous predictive feedback to the clinician.Methods & FindingsMoDN is a novel decision tree composed of feature-specific neural network modules. It creates dynamic personalised representations of patients, and can make multiple predictions of diagnoses and features, updatable at each step of a consultation. The model is validated on a real-world CDSS-derived dataset, comprising 3,192 paediatric outpatients in Tanzania.MoDN significantly outperforms ‘monolithic’ baseline models (which take all features at once at the end of a consultation) with a mean macro F1 score across all diagnoses of 0.749 vs 0.651 for logistic regression and 0.620 for multilayer perceptron (p < 0.001).To test collaborative learning between IIO datasets, we create subsets with various percentages of feature overlap and port a MoDN model trained on one subset to another. Even with only 60% common features, fine-tuning a MoDN model on the new dataset or just making a composite model with MoDN modules matched the ideal scenario of sharing data in a perfectly interoperable setting.InterpretationMoDN integrates into consultation logic by providing interpretable continuous feedback on the predictive potential of each question in a CDSS questionnaire. The modular design allows it to compartmentalise training updates to specific features and collaboratively learn between IIO datasets without sharing any data.FundingBotnar Foundation (grant n°6278)Author summaryClinical Decision Support Systems (CDSS) are emerging as a standard-of-care, offering probabilistic guidance at the bedside. Many deploy static, generic rule-based logic, resulting in inconsistent performance in evolving environments. Machine learning (ML) models could resolve this by updating predictions according to the collected data. However, traditional methods are often criticised as uninterpretable “black-boxes” and are also inflexible to fluctuations in resources: requiring retraining (and costly re-validation) each time a question is altered or added.We propose MoDN: a novel, interpretable-by-design, modular decision tree network comprising a flexible composition of question-specific neural network modules, which can be assembled in real-time to build tailored decision networks at the point-of-care, as well as enabling collaborative model learning between CDSS with differing questionnaire structures without sharing any data. |
| Databáze: |
OpenAIRE |
| Externí odkaz: |
|
