Model Extraction From Clinical Data Subject to Large Uncertainties and Poor Identifiability

Abstract

This letter presents an extension to system theory as a novel approach to provide models from clinical data under large uncertainty and poor identifiability conditions. These difficult conditions are often present in medical systems due to ethical, safety and regulatory limitations regarding application of persistent drug-related excitation to human body. Furthermore, drug-dose effect relationship is of particular challenge due to large interand intra-patient variability. This is strengthened by the lack of suitable instrumentation to measure the necessary information, rather making available inferences and surrogate metrics. A notable advantage of the proposed approach is its robustness to uncertainty. The efficacy of our approach was examined in clinical data from patients monitored during induction phase of target controlled intravenous anesthesia. The proposed method delivered models with physiological explainable parameters and suitable for closed loop control of anesthesia.

Selecting Optimal Long Short-Term Memory (LSTM) Architectures for Online Estimation of Mean Arterial Pressure (MAP) in Patients Undergoing General Anesthesia

“…LSTM networks have been increasingly applied in the prediction of medical signals due to their ability to model time series data effectively. Particularly in the context of irregularly sampled medical time series, which is common in healthcare data [10], LSTMs demonstrate notable advantages. These networks adeptly handle intra-series irregularities (varying time intervals within a single data stream) and inter-series irregularities (different sampling rates across multiple data streams) [11,12].…”