TIDAL (daTa-driven morphofunctIonal peDiAtric Lung mri)

• Applicants: Professor Dr. Simon Veldhoen; Professor Dr. Tobias Wech
• Subject Area: Radiology
• Term: since 2024
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 535829883

Respiratory diseases often significantly affect patients' quality of life and represent a major burden on healthcare systems. Based on their impact on different parts of the respiratory system, a classification can be made into diseases of the airways, which impede gas exchange, diseases of the lung tissue, often resulting in limited expansion capacity of the lungs and, ultimately, diseases of the pulmonary circulation, which affect the exchange of gases to the blood. All of these entities, however, ultimately impair lung function. Chronic lung diseases, particularly in children, require ongoing monitoring and personalized therapies. While conventional X-ray and computed tomography (CT) have been commonly used for diagnostic imaging, magnetic resonance imaging (MRI) is gaining importance also in lung applications. Morphofunctional MRI combines both structural and functional assessments of the lung in a single examination, allowing for a comprehensive evaluation. New techniques, such as Fourier decomposition and ventilation mapping, aim to enable functional lung MRI without the need for contrast agents. The use of lung MRI is especially desirable in pediatric radiology due to its potential to reduce radiation exposure. However, challenges arise in imaging children due to their lack of cooperation and the need for anesthesia. To overcome these obstacles, techniques that allow extremely rapid data acquisition, preferably during free breathing are required. Machine learning and neural networks have been employed to accelerate the MRI process and to improve image quality. However, the success of these techniques depends heavily on the quality of the training data, which may explain why their full potential has yet to be realized in pediatric lung MRI. The project aims to improve lung diagnostics for pediatric patients by accelerating MRI investigations of lung morphology and function. This will have positive effects on patient safety, comfort, image quality, and overall costs. By utilizing latest machine learning concepts in conjunction with a unique database consisting of images from more than 10 years of studies on lung MRI in our lab, the project intends to develop new MR approaches that can assess lung morphology and function in pediatric patients within shorter scan times and without the need for breath-holding. The project has four sub-objectives, including equipping different 3D UTE MR pulse sequences for self-gated lung imaging, developing new machine learning-driven MR reconstruction techniques for morphological and functional lung data, developing an MR-based method to image forced expiration for enhanced lung spirometry, and validating the developed approaches in a patient cohort with well-characterized clinical data.

• DFG Programme: Research Grants