Magnetic resonance diffusion-weighted imaging: extraneurological applications.

Abstract

Diffusion-weighted (Dw) imaging has for a number of years been a diagnostic tool in the field of neuroradiology, yet only since the end of the 1990s, with the introduction of echoplanar imaging (EPI) and the use of sequences capable of performing diffusion studies during a single breath hold, has it found diagnostic applications at the level of the abdomen. The inherent sensitivity to motion and the magnetic susceptibility of Dw sequences nonetheless still create problems in the study of the abdomen due to artefacts caused by the heartbeat and intestinal peristalsis, as well as the presence of various parenchymal-gas interfaces. With regard to focal liver lesions, a review of the literature reveals that Dw imaging is able to differentiate lesions with high water content (cysts and angiomas) from solid lesions. With regard to the latter, although there are differences between benign forms [focal nodular hyperplasia (FNH), adenoma] and malignant forms [metastasis, hepatocellular carcinoma (HCC)] in their apparent diffusion coefficient (ADC) in the average values for histological type, there is a significant overlap in values when lesions are assessed individually, with the consequent problem of their correct identification. One promising aspect is the possibility of quantifying the degree of fibrosis in patients with chronic liver disease and cirrhosis given that the deposit of collagen fibres "restricts" the motion of water molecules and therefore reduces ADC values. However, even in this field, studies can only be considered preliminary and far from real clinical applications. The retroperitoneum is less affected by motion artefacts and similarly deserves the attention of Dw imaging. Here it is possible to differentiate mucin-producing tumours of the pancreas from pseudocystic forms on the basis of ADC values even though the limited spatial resolution of Dw imaging does not enable the identification of small lesions. Dw imaging may be applied to the study of the kidney to differentiate hydronephrosis from pyonephrosis and with regard to tumours, solid from pseudocystic forms. In addition, given that renal parenchyma has significantly variable ADC values on the basis of the anatomic section and physiological conditions, the possibility of assessing functional alterations is currently being studied. Indeed, a good correlation has been found between ADC values and glomerular filtration rate. With regard to musculoskeletal applications, the absence of motion artefacts in the regions studied has enabled the development of sequences less sensitive to magnetic susceptibility and with greater spatial resolution than EPI. Attempts have therefore been made to use Dw imaging in the characterization of soft-tissue tumours although the findings so far have been disputed. Greater agreement has been found regarding sensitivity of the technique in assessing response of these tumours to chemotherapy: tumour necrosis is thought to increase ADC whereas the persistence of vital neoplastic tissue tends to lower it. One of the most promising applications of Dw imaging is without doubt the assessment of vertebral collapse where a high ADC has been shown to be associated with an osteoporotic cause and a low ADC with a neoplastic cause. Nonetheless, even here, a moderate overlap between ADC values of the two types has been encountered. Dw imaging has also been used in the assessment of bone marrow cellularity: areas of tightly packed cells show a higher ADC value than hypocellular areas. In particular, no significant difference in ADC is noted between normal hypercellular bone marrow and hypercellular bone marrow secondary to lymphomatous infiltration whereas this difference is significant between hypocellular, normocellular and haematopoietic hypercellular bone marrow. With regard to the study of joints, the limited structure dimensions, particularly cartilage, creates technical difficulties related to spatial resolution and an adequate signal-to-noise ratio, problems that can only be solved by further technological developments. Lastly, a significant difference in ADC values between degenerative and inflammatory effusion has been found, a fact that may be explained as the result of the activity of hyaluronidase present in inflammatory forms, which causes a reduction in the concentration of hyaluronic acid with a consequent decrease in viscosity.