Fig. 1

Study overview. A) The physiological digital twin presented in this work describes the dynamics of various ethanol markers: breath alcohol concentration (BrAC), blood alcohol concentration (BAC), urine alcohol concentration (UAC), blood ethyl glucuronide (EtG), blood ethyl sulphate (EtS), and phosphatidylethanol (PEth). B) The digital twin can offer predictions of the time profiles of the various ethanol markers, which could be valuable in several possible use-cases. For availability the framework is provided in a web application. (C) In this work we develop a physiological digital twin that incorporates existing knowledge and data of different drinking scenarios from ten studies. After the model development and validation, the model can make predictions of unknown drinking behaviour from available data samples. (D) The model consists of: (i) a stomach compartment that handles the interactions of alcoholic beverages, non-alcoholic beverages, and food, (ii) an intestine compartment from where the alcohol is absorbed into, (iii) a plasma compartment from which ethanol contributes to PEth formation and distributed into either, (iv) the bladder where UAC is present, (v) the tissue, or (vi) the liver where EtG and EtS are synthesized. BrAC is expressed from the plasma ethanol levels in the central compartment.