Monte Carlo Optimization of Liver Machine Perfusion Temperature Policies

Document Type

Conference Proceeding

Date of Original Version

1-1-2023

Abstract

In this work, a constrained multi-objective function formulation of liver machine perfusion (MP) based on widely accepted viability criteria and network metabolic efficiency is described. A novel Monte Carlo method is used to improve machine perfusion (MP) performance by finding optimal temperature policies for hypothermic machine perfusion (HMP), mid-thermic machine perfusion (MMP), and subnormothermic machine perfusion (SNMP). It is shown that the multi-objective function formulation can exhibit multiple maxima, that greedy optimization can get stuck at a local optimum, and that Monte Carlo optimization finds the best temperature policy in each case.

Publication Title, e.g., Journal

Lecture Notes in Computer Science Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics

Volume

13811 LNCS

Share

COinS