Graduate School of Medicine, Division of Health Sciences
I am working on the development of an innovative X-ray phase-contrast image reconstruction method that integrates simulation and AI technologies.
Conventional X-ray imaging visualizes differences in X-ray absorption; however, materials with low atomic numbers, such as soft tissues, generally absorb X-rays weakly, making it difficult to obtain clear contrast.
X-ray phase-contrast imaging offers a promising solution to this challenge. This technique leverages X-ray refraction, diffraction, and interference in image reconstruction, and has the distinct advantage of enhancing edge visibility in the imaged subject.
However, acquiring phase-contrast X-ray images requires highly coherent, monochromatic X-rays with matched wavelength and phase. This requirement poses a major challenge for clinical application, where X-ray sources typically have broad energy distributions.
To address this, my research aims to develop a novel method for extracting phase-contrast information from conventional X-ray images. This is achieved by generating simulated phase-contrast images and training deep learning models to learn their characteristics.
Through my involvement in PQBA, I hope to cultivate a multifaceted research perspective by incorporating insights not only from my field of specialization but also from a wide range of disciplines.