Munich, Germany – April 15, 2026 — Ebenbuild GmbH, a Munich-based deep-tech healthcare company, today announced the publication of a peer-reviewed validation study in Nature Communications Medicine. The study , titled "In silico high-resolution whole lung model to predict the locally delivered dose of inhaled drugs", comprehensively validates the predictive performance of Ebenbuild's patient-specific lung digital twin technology for transport and regional deposition of inhaled drugs across the entire human lung with high spatial resolution, integrating subject-specific airway anatomy and lung mechanics. Model predictions demonstrated strong quantitative agreement with in vivo imaging data obtained using single-photon emission computed tomography combined with computed tomography (3D SPECT/CT), a widely accepted reference standard for assessing three-dimensional drug deposition patterns.
Making the locally delivered dose measurable
Understanding how much of an inhaled drug reaches its site of action within the lung remains a fundamental challenge in respiratory medicine and drug development. Local drug deposition is strongly influenced by individual lung anatomy, breathing patterns, disease-related structural changes and aerosol properties, yet it cannot be measured directly in routine clinical practice or during clinical stages of drug development. As a result, critical development and dosing decisions continue to rely on indirect measurements, population averages or late-stage clinical outcomes.
Ebenbuild addresses this gap by generating patient-specific digital twins of the lung from standard CT data and applying physics-based computational modeling to analyze airflow, tissue mechanics and aerosol transport throughout the entire respiratory system. Unlike existing approaches, this lung digital twin represents both the conducting airways and the alveolar region, enabling locally resolved predictions of aerosol deposition at the individual patient level.
Validation against clinical imaging data
In the Nature Communications Medicine study, the lung digital twin was rigorously validated against in vivo 3D SPECT/CT imaging data from a controlled clinical reference study. Predictions of aerosol deposition demonstrated excellent agreement with measured imaging data across multiple inhalation scenarios. Lobar deposition predictions, for example, achieved a correlation coefficient of 0.95 with experimental data. Overall, this confirms the model's predictive performance across individual cases and aggregated analyses.
The study further illustrates how the model handles disease-related heterogeneity by simulating altered mechanical properties in fibrotic lung regions for a patient with idiopathic pulmonary fibrosis (IPF). While disease modeling was not the focus, these analyses highlight the platform's ability to incorporate patient-specific structural and mechanical differences and to assess their impact on local dose levels throughout the human lung.
Crucially, the automated and validated workflow combines image processing, model generation and high-performance simulation, delivering results within hours rather than days or weeks. This scalability distinguishes the approach from purely academic proof-of-concept research and supports its application in translational research as well as in drug and device development settings.
From validation to platform applications
"This publication provides rigorous, independent, clinical validation of the core modeling engine that underpins our platform," said Dr. Kei W. Müller, CEO and Co-founder of Ebenbuild. "Being able to quantitatively predict regional aerosol deposition at patient level is essential for replacing assumptions with evidence. The validation fuels the further scale-up of our technology, which is already successfully used in revenue-generating projects, proving its value for our customers from pharma, biotech, and medtech who can make better-informed development decisions earlier."
While the study establishes the model's predictive performance, significant challenges remain in translating inhaled therapies from development to clinical use.
"From a modeling perspective, this study shows how combining detailed patient-specific lung geometry, airflow, mechanics, and aerosol transport into a single modeling framework enables quantitative insight into pulmonary drug delivery," added Dr. Jonas Biehler, CTO and Co-founder of Ebenbuild. "By resolving these processes across the entire lung and over the breathing cycle, we can link inhalation conditions to the locally delivered dose at the site of action."
The validated technology forms the foundation of Ebenbuild's platform applications. TWINHALE applies the lung digital twin approach to enable in silico trials for the development of inhaled drugs and inhaler devices, supporting patient-specific analysis of aerosol drug deposition for dose selection, formulation development as well as trial planning and monitoring. AEROGRAM, Ebenbuild's regulated clinical application currently in development, is designed to extend the same modeling principles into intensive care, supporting decision-making for mechanically ventilated patients.
Building evidence for in silico approaches in respiratory medicine
Beyond its immediate relevance for inhaled drug development, the study contributes to the growing evidence base for computational, patient-specific modeling approaches in respiratory medicine. By demonstrating that digital twins can be both physiologically accurate and computationally scalable, the publication positions this approach as a robust and scalable in silico complement to experimental and imaging-based techniques in respiratory care. It also supports the broader regulatory shift toward the use of computational modeling and simulation as well as model-informed evidence in the development and assessment of inhaled drug products.
About Ebenbuild
