Neural network predicts need for red blood cell transfusion for patients with acute gastrointestinal bleeding admitted to the intensive care unit

Abstract

Acute gastrointestinal bleeding is the most common gastrointestinal cause for hospitalization. For high-risk patients requiring intensive care unit stay, predicting transfusion needs during the first 24 h using dynamic risk assessment may improve resuscitation with red blood cell transfusion in admitted patients with severe acute gastrointestinal bleeding. A patient cohort admitted for acute gastrointestinal bleeding (N = 2,524) was identified from the Medical Information Mart for Intensive Care III (MIMIC-III) critical care database and separated into training (N = 2,032) and internal validation (N = 492) sets. The external validation patient cohort was identified from the eICU collaborative database of patients admitted for acute gastrointestinal bleeding presenting to large urban hospitals (N = 1,526). 62 demographic, clinical, and laboratory test features were consolidated into 4-h time intervals over the first 24 h from admission. The outcome measure was the transfusion of red blood cells during each 4-h time interval. A long short-term memory (LSTM) model, a type of Recurrent Neural Network, was compared to a regression-based models on time-updated data. The LSTM model performed better than discrete time regression-based models for both internal validation (AUROC 0.81 vs 0.75 vs 0.75; P < 0.001) and external validation (AUROC 0.65 vs 0.56 vs 0.56; P < 0.001). A LSTM model can be used to predict the need for transfusion of packed red blood cells over the first 24 h from admission to help personalize the care of high-risk patients with acute gastrointestinal bleeding.

Figure 1

Example of neural network prediction for the first 24 h of a 62 year old man with Hepatitis C cirrhosis presenting with 2 days of intermittent coffee ground emesis and lethargy. Initial Glasgow Blatchford Score = 14 (a) Continuous risk prediction of the neural network through the first 24 h with the threshold set above 0.5 for detecting need for transfusion. The arrows indicate need for transfusion during that time period. (b) Measurements of Heart Rate, Systolic Blood Pressure, and Hemoglobin occurring during the first 24 h.

Figure 2

Long-Short Term Memory (LSTM) Network Model Overview. Electronic Health Record data (vitals, laboratory values) is fed into the model, passed through the layers, transformed, and gives a probability of the outcome (transfusion of packed red blood cells). At the beginning of each 4-h interval the LSTM Network can generate a probability of needing transfusion. T represents the time in hours, X represents input data (vitals, laboratory values), Y represents the probability of needing transfusion, and FCN is a fully convolutional network that processes the information from the previous time period to generate the prediction.

Figure 3

Comparison on external validation of the overall Area Under the Receiver Operating Curve (AUROC) as a measure of performance of the Long-Short Term Memory (LSTM) Neural Network model, discrete time Logistic Regression (LR), and regression with elastic net penalty.