Towards the clinical application of X-ray phase contrast imaging

• Published in: European journal of radiology 2008-12, Vol.68 (3), p.S73-S77
• Main Authors: Williams, I.M, Siu, K.K.W, Runxuan, G, He, X, Hart, S.A, Styles, C.B, Lewis, R.A
• Format: Article
• Language: English
• Subjects: Algorithms; Biological and medical sciences; Breast Neoplasms - diagnostic imaging; Clinical Trials as Topic; Female; Genital system. Mammary gland; Humans; Imaging, Three-Dimensional - methods; Investigative techniques, diagnostic techniques (general aspects); Mammography; Mammography - methods; Medical sciences; Phase contrast imaging; Propagation-based imaging; Radiodiagnosis. Nmr imagery. Nmr spectrometry; Radiographic Image Enhancement - methods; Radiographic Image Interpretation, Computer-Assisted - methods; Radiology; Refractometry - methods; Reproducibility of Results; Sensitivity and Specificity; Tomography, X-Ray - methods; X-ray
• ISSN: 0720-048X; 1872-7727
• DOI: 10.1016/j.ejrad.2008.04.042

Abstract Synchrotron-based propagation-based imaging, a type of phase contrast imaging, produces better soft tissue image contrast than conventional radiography. To determine whether the technique is directly transferable to the clinical environment for routine diagnostic or screening imaging, a micro-focus (100 μm spot-size) Molybdenum X-ray source with 0.03 mm molybdenum filtration was installed at a local hospital. Breast tissue samples, excised masses and mastectomies, were obtained directly from surgery and imaged at three geometries. The first geometry was optimised for visualizing phase contrast effects using a ray-line argument, the second was the same as that employed by Konica-Minolta in their commercial phase contrast system, and the third was the conventional contact arrangement. The three images taken of each tissue sample were comparatively scored in a pair-wise fashion. Scoring was performed by radiologist expert in mammography, general radiologists, associated clinicians and radiographers on high-resolution mammography rated monitors at two separate locations. Scoring indicated that the optimised and Konica geometries both outperformed the conventional mammographic geometry. An unexpected complication within the trial was the effect that the scoring platform and the associated display tools had on some of the scorer's responses. Additionally, the trial revealed that none of the conventional descriptors for image quality were adequate in the presence of phase contrast enhancements.