This study aims to develop an image-modality conversion technology based on CT images. A deep learning–based model is designed to generate information similar to other imaging modalities from CT images. Through this approach, the system learns the relationships between medical imaging modalities and generates high-quality images useful for diagnosis. Training strategies are applied to preserve structural consistency and anatomical information in the converted images. This technology is expected to support clinicians in diagnosis and treatment planning by enabling more efficient use of medical imaging data.
Project Period
2024
This study aims to develop an image-modality conversion technology based on CT images. A deep learning–based model is designed to generate information similar to other imaging modalities from CT images. Through this approach, the system learns the relationships between medical imaging modalities and generates high-quality images useful for diagnosis. Training strategies are applied to preserve structural consistency and anatomical information in the converted images. This technology is expected to support clinicians in diagnosis and treatment planning by enabling more efficient use of medical imaging data.