Apparent diffusion coefficient measurement using DW-HASTE imaging is helpful in differentiating invasive ductal carcinoma from fibroadenoma. Nevertheless, DW-HASTE imaging has low detectability at present; therefore, more technologic improvement is desired.

No abstract

Materials and MethodsDynamic CT studies of 42 patients with HCC having angiographically proven arterioportal shunt were reviewed. Thirty-eight were men and four were women with an average age of 58 years. The proximal site of the portal vein opacified on a hepatic angiogram was as follows: portal trunk or more proximal in 10, first-order (main) branch in five, second-order branches in two, and third-order and smaller branches in 25 cases. CT was performed using a GE CT/T 8800 or 9000 with a scanning time of 1 1 or 5 sec, respectively. All the CT examinations were done before the angiograms. Dynamic study with various intervals of scan and numbers of slices was performed with a manual bolus injection of 60 ml of Conray 400 (sodium iothalamate). Except for two patients, the dynamic scan sequence was performed only once at the level of the tumor detected on plain CT or near the level of the hilum when tumor was not demonstrated on plain CT.One hundred cases of HCC lacking an arterioportal shunt on angiography and examined by dynamic CT study were reviewed as control group.

MRI using T(2)*-WI is a sensitive, specific, and accurate method to evaluate EP. T(2)*-WI is highly accurate for detecting and diagnosing EP because of its sensitivity to fresh hematoma.

Fifty-three patients who had hepatic tumors (24 hepatomas, ten metastases, and 19 cavernous hemangiomas) underwent MR imaging using a 0.35-T superconducting imager. The transverse relaxation time (T2) was calculated from a pair of spin echo images (repetition time [TR] of 1600 msec) with echo delay times (TE) of 35 and 70 msec. The computed T2 value was obtained in a fashion similar to that used to obtain CT numbers with region-of-interest cursors. The mean T2 was 59 +/- 9 msec in hepatomas, 64 +/- 15 msec in metastases, and 100 +/- 30 msec in hemangiomas. The difference between the T2 of hemangioma and that of liver malignancies was statistically significant (P less than .001); however, differentiation between hepatoma and metastases was not possible. The T2 was shorter than 80 msec in all 24 hepatomas and in nine of ten metastases, and was longer than 80 msec in 16 of 19 hemangiomas. Forty-nine of 53 cases (92%) were correctly classified when the borderline of T2 between hemangioma and hepatic malignancies was set at 80 msec. MR with T2 calculation was valuable in differentiating between hemangioma and hepatic malignancies.