Non-invasive assessment of diffuse liver disease by in vivo measurement of proton nuclear magnetic resonance relaxation times at 0.08 T.

44 patients with a range of parenchymal liver diseases diagnosed by biopsy or laboratory investigations underwent proton nuclear magnetic resonance (NMR) relaxometry of the liver at 0.08 T. T1 maps were produced using an interleaved saturation recovery and inversion recovery sequence and T2 maps using a four echo Carr-Purcell-Meiboom-Gill sequence. Significantly raised relaxation times compared with a previously studied group of 42 normal volunteers were found in groups of patients with alcoholic cirrhosis (p < 0.001 for T1 and T2), chronic active hepatitis (CAH) (p < 0.01 for T1 and T2) and minor liver abnormalities (p < 0.01, T2 only). T1 was significantly higher in cirrhotics than in patients with CAH (p < 0.002) and minor abnormalities (p < 0.001). This suggests a role for relaxometry in the confirmation of the presence of cirrhosis (sensitivity = 75%, specificity approximately 97%, taking T1 > 266 ms as a positivity criterion). Reduced T2 values were found in patients with liver iron overload prior to venesection (p < 0.001 versus normals, p < 0.02 versus venesected patients). Although this latter test has relatively low sensitivity and specificity, it may have a role in the monitoring of treatment for iron overload.

Magnetic resonance imaging of urinary calculi.

Accurate prediction of the response of an individual patient to lithotripsy remains impossible. Certain factors such as the chemical composition, size, and position of the calculus are known to be important in determining the success rate. This paper reports the use of magnetic resonance imaging (MRI) to evaluate 141 urinary calculi in vitro. A wide range of signals for each chemical type of calculus was found on each of the three imaging sequences used (T1-weighted, T2-weighted, and proton density). None of the chemical groups examined showed a typical MRI profile allowing it to be distinguished from the other groups. Analysis of variance showed a statistical difference between signals for apatite and struvite on the T1-weighted sequence, and between struvite and uric acid on the proton density sequence (both, P < 0.05). These results show for the first time that MRI is capable of distinguishing between different chemical types of stones. This is particularly important for the comparison of struvite and apatite which appear to be similar in conventional investigations but have quite different hardness values. Further work is in progress correlating the results of this study with stone microhardness and extracorporeal shockwave lithotripsy fragility tests to determine whether MRI accurately predicts the success of lithotripsy.

Proton NMR relaxation times in the normal human liver at 0.08 T.

The spin-lattice (T1) and spin-spin (T2) relaxation times of liver in 42 normal volunteers (21 male and 21 female) were measured using a calibrated 0.08 T resistive imager capable of accurate and reproducible relaxometry. T1 was determined using an interleaved gradient echo saturation recovery and inversion recovery technique and T2 using a four-echo Carr-Purcell-Meiboom-Gill sequence. The ranges obtained were T1 = 213 +/- 14 ms and T2 = 66 +/- 5 ms. More specific ranges were obtained for each sex and for younger and older subjects. A small variation in T1 was found between older (greater than 40 years) and younger (less than 40 years) subjects, but no such effect was observed in the case of T2. No significant variations were found when female volunteers were imaged at weekly intervals through the menstrual cycle, when a male volunteer was imaged repeatedly over the course of several months or when male volunteers consumed small quantities of alcohol.

A study comparing SPECT with CT and MRI after closed head injury.

After closed head injury nineteen patients had single photon emission tomography (SPECT) using the lipophilic tracer 99m-Technetium hexamethyl-propylene-amineoxime (HMPAO) to compare the defects shown by CT and MRI. SPECT showed more focal cerebral lesions than either CT or MRI alone or in combination. Most lesions shown by SPECT were not shown by CT or MRI in the corresponding anatomical regions. The most severely disabled patients showed the highest number of SPECT lesions (average four per patient) and the lowest (mean, SE) cerebral blood flow (718, 69 ml/min) compared with the less disabled patients (two per patient and 1058, 51 ml/min, p less than 0.05). There was a correlation between the Glasgow Outcome Scale grade and the global cerebral blood flow (r 0.74, p less than 0.05). The perfusion defects may correlate with clinical signs that were not explained by CT or MRI findings. SPECT may complement the clinical evaluation in the assessment of outcome after head injury.