Jerrold H. Mink

Before surgery, 277 menisci in 144 knees were examined with magnetic resonance (MR) imaging. They were then examined directly with arthroscopy or arthrotomy. Menisci were graded on a scale of 1-3 according to the character of the intrameniscal MR imaging signal. At surgery, 137 of 154 (89%) menisci exhibiting only grade 1 or grade 2 signal were found to be normal. One hundred sixteen of 123 (94%) menisci exhibiting intrameniscal signal communicating with a meniscal articular surface (grade 3 signal) had tears. If only a grade 3 signal is considered consistent with meniscal tears, then MR findings and surgical findings agreed in 91.3% of menisci. MR imaging can separate surgically significant from nonsignificant meniscal lesions and is useful in the noninvasive preoperative screening of suspected meniscal tears.

Scant attention has been paid to the role of magnetic resonance (MR) imaging in the assessment of posttraumatic disorders of bone and cartilage at the knee. The authors reviewed 66 cases and identified four types of injuries that were not clearly evident on plain radiographs. A bone bruise was the most common lesion, having a high association with anterior cruciate ligament tears. Osteochondral fractures, stress fractures, and supracondylar femoral and tibial plateau fractures were also identified. MR imaging appears extremely useful in the detection and assessment of bone and cartilage disorders around the knee.

Menisci from 12 autopsies and above-knee amputations were imaged with magnetic resonance (MR) at 1.5 T and then sectioned for gross and histologic examination. A histologic staging system was developed and showed a one-to-one correlation with corresponding grades of MR signal intensities. Histologic stages 1 and 2 represented a continuum of degeneration culminating in stage 3 fibrocartilaginous tears, seen most frequently in posterior-horn segments of the medial meniscus. Correlation of histologic stages with MR signal intensity allows for an improved diagnostic reading of MR images.

In 242 of 3,000 patients who underwent magnetic resonance (MR) imaging of the knee between September 1986 and August 1987, original MR imaging reports were compared with subsequent arthroscopic reports to determine the value of MR imaging in the evaluation of suspected meniscal and complete tears of the anterior cruciate ligament. The overall accuracy for the menisci was 93% (sensitivity, 95%; specificity, 91%) with a false-negative rate of 4.8%. For the anterior cruciate ligament the overall accuracy was 95%. T2-weighted sequences were associated with greater sensitivity, specificity, and accuracy than were T1 sequences; the false-negative rate was 0% in the T2-weighted group. MR imaging of the knee is an extremely accurate means for noninvasive assessment of the integrity of the menisci and anterior cruciate ligament, and the accuracy exceeds that usually reported for arthrography.

Coronal T1-weighted magnetic resonance (MR) images were obtained on a 1.5-T system in 23 patients in whom there was a high clinical suspicion of hip fracture although initial radiographs were normal. MR imaging correctly demonstrated fracture in nine of nine patients; the precise configuration of the fracture line was delineated in eight patients. MR imaging was useful in excluding fractures in 14 of 14 patients, who were subsequently followed up clinically for a minimum of 3 months. Radionuclide scans were positive in four of four patients with fractures and equivocal in one patient who was subsequently demonstrated to have no fracture. Additional imaging studies were requested by clinicians in all cases in which the bone scan was positive. The results of this study suggest that MR imaging can provide a rapid, cost-effective, and anatomically precise diagnosis of hip fracture in patients with normal or equivocal initial radiographs. The specificity of the diagnosis achieved can obviate supplemental imaging examinations, with their attendant additional expense and radiation exposure.