Major lung diseases, such as lung cancer and chronic obstructive pulmonary disease, have always been a threat to human health and are becoming more serious with the increase of air pollution. MRI is an important clinical medical imaging technique, which has the advantages of non-radiation compared with chest X-ray, CT and PET. However, the lungs are mostly empty tissue, making them a blind spot on conventional MRI. It is well known that the physiological and pathophysiological characteristics of different lung diseases lead to different ventilation modes, and the observation of pulmonary ventilation modes can effectively explain the etiology of pulmonary ventilation defects. However, it is difficult to accurately describe the dynamic process of pulmonary ventilation with existing techniques, so it is urgent to develop new techniques for dynamic visualization of pulmonary ventilation.

A graduate student in Wuhan has independently developed a hyperpolarized 129Xe MRI instrument for human lungs, enabling visual observation of pulmonary gas exchange and "lighting up" the lungs. When the hyperpolarized xenon-129 gas continuously enters the human lungs, the gas inflow effect will have a great influence on the image quality. This effect interferes with the hyperpolarized gas signal in the sampling process, similar to high-pass filtering, while in the image reconstruction process, additional noise and artifacts are introduced to reduce the image SNR. This is an important factor that is difficult to accurately describe the dynamic process of pulmonary ventilation.

In order to develop fast and high-resolution MRI of human lung gas, the researchers designed an under-sampling strategy based on variable Angle excitation (VFA) and combined with low-rank, sparse and other dynamic image characteristics to significantly improve signal acquisition speed and quality. Compared with the existing hyperpolarized xenon-129 lung dynamic imaging methods in the world, this technology achieves the temporal resolution of human lung gas MRI of 445 ms/layer and spatial resolution of 3 mm, providing rapid and dynamic lung function imaging under free breathing for the early diagnosis of major lung diseases.