Multitask prediction of organ dysfunction in the intensive care unit using sequential subnetwork routing

Abstract

Objective Multitask learning (MTL) using electronic health records allows concurrent prediction of multiple endpoints. MTL has shown promise in improving model performance and training efficiency; however, it often suffers from negative transfer – impaired learning if tasks are not appropriately selected. We introduce a sequential subnetwork routing (SeqSNR) architecture that uses soft parameter sharing to find related tasks and encourage cross-learning between them. Materials and Methods Using the MIMIC-III (Medical Information Mart for Intensive Care-III) dataset, we train deep neural network models to predict the onset of 6 endpoints including specific organ dysfunctions and general clinical outcomes: acute kidney injury, continuous renal replacement therapy, mechanical ventilation, vasoactive medications, mortality, and length of stay. We compare single-task (ST) models with naive multitask and SeqSNR in terms of discriminative performance and label efficiency. Results SeqSNR showed a modest yet statistically significant performance boost across 4 of 6 tasks compared with ST and naive multitasking. When the size of the training dataset was reduced for a given task (label efficiency), SeqSNR outperformed ST for all cases showing an average area under the precision-recall curve boost of 2.1%, 2.9%, and 2.1% for tasks using 1%, 5%, and 10% of labels, respectively. Conclusions The SeqSNR architecture shows superior label efficiency compared with ST and naive multitasking, suggesting utility in scenarios in which endpoint labels are difficult to ascertain.

Continual learning of longitudinal health records

“…Since class sizes are highly imbalanced in all experiments (mortality outcome averaging 10% across tasks, see Table IX), and the degree of class imbalance is not constant across domain splits, accuracy is an inappropriate measure of model performance [37]. In minority-event detection, metrics such as sensitivity and specificity (i.e.…”

Healthsheet: Development of a Transparency Artifact for Health Datasets

“…was due to the availablity of the dataset, wide range of tasks, and applications on the dataset, availability and depth of documentation and details available from the dataset. In addition, multiple publications, employed the dataset to suggest new sub-version, tasks and labels on the dataset [59]. We also studied MSOAC [35] and Floodlight [6], both on Multiple Sclerosis disease and are respectively on Electronic Health Records (EHR), and smartphone-based performance outcome measures.…”

Advances in artificial intelligence and deep learning systems in ICU-related acute kidney injury

“…Twenty studies evaluated at least one logistic regression model; fourteen evaluated logistic regression models exclusively. Nine articles [5 ▪▪ ,17 ▪▪ ,18,19 ▪▪ ,20 ▪▪ ,22 ▪▪ ,23 ▪ ,25 ▪▪ ] used deep learning models and five [7 ▪▪ ,10 ▪▪ ,16 ▪▪ ,21 ▪▪ ,26 ▪ ,32 ▪▪ ] used other machine learning models: gradient boosting trees (GBT), random forest, support vector machines (SVM), Naive Bayes, k-Nearest Neighbor (kNN) and decision trees. Model performance was most commonly assessed by area under the receiver operating characteristic curve (AUROC), ranging from 0.52 to 0.98 (Table 2 and Supplemental Table 1, http://links.lww.com/COCC/A37) [10 ▪▪ ].…”