Ex vivo lung perfusion (EVLP) presents a critical biotechnological platform in the evaluation of donated organs before transplantation and in the examination of biological processes. This technology enables the production of comprehensive data sets consisting of imaging, physiological monitoring, and molecular analyses by allowing the lung to maintain its physiological functions while outside the body. In biotechnological research, combining this data with the concept of the digital twin holds strategic importance for modeling complex biological systems at the cellular and clinical levels.
In clinical research, evaluating the efficacy of new treatments faces significant technical bottlenecks due to the high natural variation in human lung functions. The control groups used in traditional methods can fall short in detecting the true effects of drugs due to limited organ availability and the biological diversity among patients. This study aims to develop personalized digital lung models that can create a dedicated virtual control group for each individual organ and predict the future functions of that organ with high accuracy.
The researchers used a hybrid physics-informed machine learning architecture, drawing on an extensive data set obtained from more than 1000 clinical cases. This model combines physics-based mechanical equations derived from “mechanical” ventilator data with data-driven algorithms that analyze molecular and imaging data. The developed system consists of both static and dynamic components, and it has the capacity to simultaneously model more than 75 parameters spanning physiology, biochemistry, radiography, transcriptomics, metabolomics, and proteomics.
The results of the study reveal that the digital twins predict the physiological and biochemical parameters of the lung with low error margins ranging between 2% and 11%. In particular, in lungs treated with alteplase, a clot-dissolving tissue plasminogen activator, the effect of the treatment on vascular pressure was statistically significantly confirmed thanks to the personalized controls generated by the digital models. This method, which identifies treatment responses that traditional cohort analyses fail to detect, increases research efficiency by reducing the required sample size threefold.
This research shows that the creation of digital copies of human organs may establish a new approach in drug development and preclinical evaluation processes. Digital twin technology has the capacity to support treatment decisions and identify potential risks by simulating individual organ responses in advance. Furthermore, the fact that this architecture is adaptable to other ex vivo organ systems paves the way for safer and more effective personalized medicine applications in the future.
Translated by: Ziya Burak Ekinci
Editor: Elinsu Ak
Referans: Zhou, X., Wang, B., Wei, Y. et al. Digital twins of ex vivo human lungs enable accurate and personalized evaluation of therapeutic efficacy. Nat Biotechnol(2026). https://doi.org/10.1038/s41587-026-03121-4
-Bioinfocodes Scientific News Service-
News articles prepared by our team members, reviewing and compiling scientific research
published in journals with and impact factor greater than 20 (click here for the list)