Objective: High-resolution images of finger joints with chemical-shift elimination can be obtained using an interleaved water-fat (IWF) sequence. This study assessed IWF imaging of finger joints in the delineation of bone structures by comparing images of cadaver fingers with those of microcomputed tomography (CT) that served as a standard reference.
Materials and Methods: IWF images with spatial resolution of 176 µ × 176 µ × 300 µ were obtained from the distal and proximal interphalangeal joints of two cadaver finger specimens using a custom-built radiofrequency receive coil at 1.5T. Regular three-dimensional gradient-echo (GRE) images were also acquired with similar parameters and compared with the IWF images to evaluate the effects of chemical shift. Micro-CT scans were obtained and served as the standard reference. The image data were reviewed by two experienced musculoskeletal radiologists in consensus. The delineation of normal joint structures and abnormalities in the finger specimens as revealed by the magnetic resonance imaging (MRI) and micro-CT images were compared. The IWF and regular GRE images were assigned scores 0–3 for the depiction of apparent marginal bone defects, with zero being the same in appearance to the micro-CT image and three as having minimal resemblance to it. Statistical analysis of the scoring results was conducted to compare the two MRI techniques.
Results: The high-resolution IWF images provided accurate delineation of bone and calcified structures as seen in micro-CT. The thickness of subchondral bone was depicted similarly on the IWF water + fat and the micro-CT images but not on the regular GRE images. The regular GRE sequence showed false marginal bone defects not observed with IWF and micro-CT. In addition, the IWF water-only images facilitated the identification of bone cyst by revealing its water content.
Conclusion: High-resolution IWF imaging should be useful for the early diagnosis and treatment assessment of arthritis and should also benefit basic research in the pathophysiology of the disease.
Boosting Resolution and Recovering Texture of 2D and 3D Micro‐CT Images with Deep Learning