Warfarin maintenance dose prediction for Chinese after heart valve replacement by a feedforward neural network with equal stratified sampling

Abstract

Patients requiring low-dose warfarin are more likely to suffer bleeding due to overdose. The goal of this work is to improve the feedforward neural network model's precision in predicting the low maintenance dose for Chinese in the aspect of training data construction. We built the model from a resampled dataset created by equal stratified sampling (maintaining the same sample number in three dose-groups with a total of 3639) and performed internal and external validations. Comparing to the model trained from the raw dataset of 19,060 eligible cases, we improved the low-dose group's ideal prediction percentage from 0.7 to 9.6% and maintained the overall performance (76.4% vs. 75.6%) in external validation. We further built neural network models on single-dose subsets to invest whether the subsets samples were sufficient and whether the selected factors were appropriate. The training set sizes were 1340 and 1478 for the low and high dose subsets; the corresponding ideal prediction percentages were 70.2% and 75.1%. The training set size for the intermediate dose varied and was 1553, 6214, and 12,429; the corresponding ideal prediction percentages were 95.6, 95.1%, and 95.3%. Our conclusion is that equal stratified sampling can be a considerable alternative approach in training data construction to build drug dosing models in the clinic.

Clinical effect of short-term spinal cord stimulation in the treatment of patients with primary brainstem hemorrhage-induced disorders of consciousness

“…Furthermore, clinical data such as age, sex, and history of the disease are important for clinical treatment (42). There was no significant difference between the efficacy and inefficacy groups in terms of age, sex, hypertension, or MISPT history in our study; this result is similar to that reported in the literature (33).…”

Effects of Different Anesthetics on Perioperative Organ Protection and Postoperative Cognitive Function in Patients Undergoing Cardiac Valve Replacement with Cardiopulmonary Bypass

“…[ 2 ]. The clinical manifestations are valvular insufficiency, stenosis resulting in increased cardiac load, and cardiac insufficiency [ 3 ]. A common treatment method for this disease is heart valve replacement for both middle and advanced level rheumatic heart disease, mainly replacing the diseased valve with an artificial valve to improve clinical symptoms, and the operation is usually carried out under cardiopulmonary bypass [ 4 ].…”

Machine Learning-Based Predictive Models for Patients with Venous Thromboembolism: A Systematic Review

“…Theoretically, ML algorithms could be applied to develop predictive models to optimize warfarin dosing, using either regression or multi classification models. Nine studies were included [96][97][98][99][100][101][102][103][104] . Most of them derive from a multiethnic International Warfarin Pharmacogenetics Consortium (IWPC) dataset 105 using several ML regression models (Supplementary Table 5).…”

“…Nine studies were included. 96 97 98 99 100 101 102 103 104 Most of them derive from a multiethnic International Warfarin Pharmacogenetics Consortium (IWPC) dataset 105 using several ML regression models ( Supplementary Table S5 , available in the online version). Sharabiani et al 96 have incorporated patient-specific clinical and genetic data to predict a personalized warfarin dose in the IWPC dataset using relevance vector machines.…”