Periodic Reporting for period 1 - InSilicoPlacenta (Image-Based In Silico Modelling of Feto-Placental Vasculature)
Periodo di rendicontazione: 2023-07-10 al 2026-02-09
Fetal growth restriction (FGR) is a major pregnancy complication in which a baby does not reach its biological growth potential. It affects millions of families worldwide and is responsible for a significant proportion of stillbirths. A key contributor to FGR is abnormal development or function of the placenta—the temporary organ that supplies the fetus with oxygen and nutrients. Yet, despite its importance, the placenta remains one of the least understood organs in human biology. Its complex internal vascular structure cannot be directly examined during pregnancy, and current clinical imaging tools provide only limited information about how blood and oxygen move through this vital organ.
Today, ultrasound is the standard method for monitoring placental health, but it mainly captures shape and size rather than function. Magnetic Resonance Imaging (MRI) can reveal more detailed tissue characteristics, but even advanced MRI techniques cannot resolve the fine scale vascular architecture where crucial exchanges of oxygen and nutrients occur. As a result, clinicians lack the tools needed to fully understand placental dysfunction or to detect early signs of FGR.
High resolution micro CT imaging allows researchers to visualise the placental vascular network with finely resolved structural information, but this technique can only be used after delivery because it relies on ionising radiation. At the same time, computational modelling, which uses mathematics and physics to simulate biological processes, has become a powerful approach for exploring how blood moves through complex tissues. Yet, despite these advances, current models of placental blood flow remain simplified and do not capture the full structure or dynamic behaviour of the feto placental circulation.
The InSilicoPlacenta project aims to overcome the current limitations in placental imaging and modelling by creating the first comprehensive computer based model of the human placental vascular system. By combining ex vivo micro CT imaging, advanced simulations of blood flow, and state of the art MRI modelling, the project will generate a detailed digital representation of how the placenta functions from the umbilical cord down to the smallest villous branches. This integrated framework will allow researchers to explore how structural abnormalities disrupt blood flow, oxygen transfer, and tissue microstructure in pregnancies affected by fetal growth restriction.
A key strength of the project is its commitment to experimental validation. The in silico models will be compared with real MRI data from healthy and FGR affected human pregnancies, as well as with histological examinations of placental tissue. Because direct measurement of oxygen transfer in human pregnancies is not ethically possible, the project will also use a well established sheep model of pregnancy to obtain gold standard invasive measurements of oxygenation. These data will help confirm whether the simulated blood flow and oxygen transport accurately reflect biological reality.
By the end of the project, InSilicoPlacenta will deliver a powerful new tool for understanding how placental structure and function interact, why these processes fail in FGR, and how early signs of pathology might be detected through non invasive imaging. The expected impact is substantial: improved diagnostic capabilities, better pregnancy monitoring, and ultimately, healthier outcomes for the more than 300,000 babies affected by FGR each year in the EU. The project also contributes to Europe’s strategic goals in digital health, maternal fetal medicine, and computational modelling, demonstrating how interdisciplinary research can address pressing public health challenges.
Today, ultrasound is the standard method for monitoring placental health, but it mainly captures shape and size rather than function. Magnetic Resonance Imaging (MRI) can reveal more detailed tissue characteristics, but even advanced MRI techniques cannot resolve the fine scale vascular architecture where crucial exchanges of oxygen and nutrients occur. As a result, clinicians lack the tools needed to fully understand placental dysfunction or to detect early signs of FGR.
High resolution micro CT imaging allows researchers to visualise the placental vascular network with finely resolved structural information, but this technique can only be used after delivery because it relies on ionising radiation. At the same time, computational modelling, which uses mathematics and physics to simulate biological processes, has become a powerful approach for exploring how blood moves through complex tissues. Yet, despite these advances, current models of placental blood flow remain simplified and do not capture the full structure or dynamic behaviour of the feto placental circulation.
The InSilicoPlacenta project aims to overcome the current limitations in placental imaging and modelling by creating the first comprehensive computer based model of the human placental vascular system. By combining ex vivo micro CT imaging, advanced simulations of blood flow, and state of the art MRI modelling, the project will generate a detailed digital representation of how the placenta functions from the umbilical cord down to the smallest villous branches. This integrated framework will allow researchers to explore how structural abnormalities disrupt blood flow, oxygen transfer, and tissue microstructure in pregnancies affected by fetal growth restriction.
A key strength of the project is its commitment to experimental validation. The in silico models will be compared with real MRI data from healthy and FGR affected human pregnancies, as well as with histological examinations of placental tissue. Because direct measurement of oxygen transfer in human pregnancies is not ethically possible, the project will also use a well established sheep model of pregnancy to obtain gold standard invasive measurements of oxygenation. These data will help confirm whether the simulated blood flow and oxygen transport accurately reflect biological reality.
By the end of the project, InSilicoPlacenta will deliver a powerful new tool for understanding how placental structure and function interact, why these processes fail in FGR, and how early signs of pathology might be detected through non invasive imaging. The expected impact is substantial: improved diagnostic capabilities, better pregnancy monitoring, and ultimately, healthier outcomes for the more than 300,000 babies affected by FGR each year in the EU. The project also contributes to Europe’s strategic goals in digital health, maternal fetal medicine, and computational modelling, demonstrating how interdisciplinary research can address pressing public health challenges.
During the project, substantial progress was made toward developing an integrated computational framework for understanding how placental structure and function interact in healthy and fetal growth restricted pregnancies. The work combined high resolution ex vivo imaging, advanced computational modelling, quantitative MRI analysis, and invasive physiological measurements in a large animal model.
A major achievement was the processing and reconstruction of high resolution micro CT datasets from human placentae. These data were used to generate detailed three dimensional vascular networks, providing the anatomical foundation for the haemodynamic simulations. Building on these reconstructions, a physiologically informed modelling framework was implemented to simulate blood flow, pressure, and hematocrit distribution across the placental vascular tree. This approach incorporated vessel specific viscosity and microcirculatory effects, enabling realistic network level simulations of placental haemodynamics.
In parallel, the project established a complete diffusion MRI processing pipeline for both retrospective and prospective human datasets. Diffusion tensor and intravoxel incoherent motion models were applied to derive quantitative microstructural parameters, forming a key analytical component for future integration with synthetic MRI simulations. Although the full Monte Carlo–based diffusion simulations were not completed within the reporting period, the necessary imaging datasets, vascular reconstructions, and haemodynamic outputs were prepared to support this next stage.
The experimental component of the project was successfully advanced through a secondment in University of South Australia in a specialised large animal research environment. Invasive blood gas measurements were obtained from pregnant ewes under normal and fetal growth restricted conditions, providing gold standard data on maternal and fetal oxygenation and hematocrit. These measurements offer essential physiological reference values for validating the computational and imaging models.
The project also generated a comprehensive human dataset combining prospective placental MRI, quantitative diffusion analysis, and post delivery histopathology. This dataset spans both healthy and growth restricted pregnancies and provides a robust basis for future validation of the simulation based imaging framework.
Together, these achievements demonstrate the feasibility of integrating micro CT derived vascular reconstructions, haemodynamic modelling, quantitative MRI, and invasive physiological measurements into a unified platform. The work completed during the project establishes the anatomical, computational, and experimental foundations required to advance in silico placental imaging and supports future efforts to improve early detection and understanding of fetal growth restriction.
A major achievement was the processing and reconstruction of high resolution micro CT datasets from human placentae. These data were used to generate detailed three dimensional vascular networks, providing the anatomical foundation for the haemodynamic simulations. Building on these reconstructions, a physiologically informed modelling framework was implemented to simulate blood flow, pressure, and hematocrit distribution across the placental vascular tree. This approach incorporated vessel specific viscosity and microcirculatory effects, enabling realistic network level simulations of placental haemodynamics.
In parallel, the project established a complete diffusion MRI processing pipeline for both retrospective and prospective human datasets. Diffusion tensor and intravoxel incoherent motion models were applied to derive quantitative microstructural parameters, forming a key analytical component for future integration with synthetic MRI simulations. Although the full Monte Carlo–based diffusion simulations were not completed within the reporting period, the necessary imaging datasets, vascular reconstructions, and haemodynamic outputs were prepared to support this next stage.
The experimental component of the project was successfully advanced through a secondment in University of South Australia in a specialised large animal research environment. Invasive blood gas measurements were obtained from pregnant ewes under normal and fetal growth restricted conditions, providing gold standard data on maternal and fetal oxygenation and hematocrit. These measurements offer essential physiological reference values for validating the computational and imaging models.
The project also generated a comprehensive human dataset combining prospective placental MRI, quantitative diffusion analysis, and post delivery histopathology. This dataset spans both healthy and growth restricted pregnancies and provides a robust basis for future validation of the simulation based imaging framework.
Together, these achievements demonstrate the feasibility of integrating micro CT derived vascular reconstructions, haemodynamic modelling, quantitative MRI, and invasive physiological measurements into a unified platform. The work completed during the project establishes the anatomical, computational, and experimental foundations required to advance in silico placental imaging and supports future efforts to improve early detection and understanding of fetal growth restriction.
The project delivered several advances that push the boundaries of current placental research by integrating high resolution imaging, computational haemodynamics, quantitative MRI, and experimental physiology into a unified framework. These results go beyond the state of the art in both methodology and scientific insight.
1. First integrated pipeline linking micro CT vascular reconstructions with haemodynamic modelling.
The project developed a complete workflow for reconstructing human placental vascular networks from micro CT data and modelling blood flow, pressure, and hematocrit across the entire tree. This represents a major step forward compared with existing models, which typically rely on simplified geometries or steady state assumptions. The ability to simulate spatially heterogeneous haemodynamics across anatomically realistic networks opens new avenues for understanding how structural abnormalities contribute to fetal growth restriction.
2. Novel quantitative MRI framework for human placental imaging
A full diffusion MRI processing pipeline was established for both prospective and retrospective datasets, enabling systematic extraction of microstructural parameters such as mean diffusivity, fractional anisotropy, and pseudo diffusion coefficients. This provides a robust foundation for future synthetic MRI simulations and for linking MRI derived biomarkers to underlying vascular structure and function.
3. Gold standard physiological measurements in a large animal model
The project successfully obtained invasive maternal–fetal oxygenation and hematocrit measurements in a validated sheep model of fetal growth restriction. These data are rarely available and provide essential ground truth for validating computational predictions of oxygen transport. This dataset significantly strengthens the translational relevance of the modelling framework.
4. New automatic segmentation framework for sheep placentomes
The development of an automatic, reproducible segmentation tool for placentomes in sheep MRI represents a methodological innovation. It addresses a major gap in large animal placental imaging, where manual segmentation is time consuming and prone to observer variability. This tool enhances scalability and supports future comparative studies between animal and human placental physiology.
1. First integrated pipeline linking micro CT vascular reconstructions with haemodynamic modelling.
The project developed a complete workflow for reconstructing human placental vascular networks from micro CT data and modelling blood flow, pressure, and hematocrit across the entire tree. This represents a major step forward compared with existing models, which typically rely on simplified geometries or steady state assumptions. The ability to simulate spatially heterogeneous haemodynamics across anatomically realistic networks opens new avenues for understanding how structural abnormalities contribute to fetal growth restriction.
2. Novel quantitative MRI framework for human placental imaging
A full diffusion MRI processing pipeline was established for both prospective and retrospective datasets, enabling systematic extraction of microstructural parameters such as mean diffusivity, fractional anisotropy, and pseudo diffusion coefficients. This provides a robust foundation for future synthetic MRI simulations and for linking MRI derived biomarkers to underlying vascular structure and function.
3. Gold standard physiological measurements in a large animal model
The project successfully obtained invasive maternal–fetal oxygenation and hematocrit measurements in a validated sheep model of fetal growth restriction. These data are rarely available and provide essential ground truth for validating computational predictions of oxygen transport. This dataset significantly strengthens the translational relevance of the modelling framework.
4. New automatic segmentation framework for sheep placentomes
The development of an automatic, reproducible segmentation tool for placentomes in sheep MRI represents a methodological innovation. It addresses a major gap in large animal placental imaging, where manual segmentation is time consuming and prone to observer variability. This tool enhances scalability and supports future comparative studies between animal and human placental physiology.