MR diagnosis of meniscal tears: analysis of causes of errors
MR imaging of the knee is an accurate method for diagnosing meniscal tears. However, MR findings do not always agree with surgical findings. In a retrospective study, we analyzed the various causes of incorrect MR diagnoses. We reviewed a series of 400 MR examinations for suspected meniscal tears in...
Gespeichert in:
Veröffentlicht in: | American journal of roentgenology (1976) 1994-12, Vol.163 (6), p.1419-1423 |
---|---|
Hauptverfasser: | , , , |
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
Zusammenfassung: | MR imaging of the knee is an accurate method for diagnosing meniscal tears. However, MR findings do not always agree with surgical findings. In a retrospective study, we analyzed the various causes of incorrect MR diagnoses.
We reviewed a series of 400 MR examinations for suspected meniscal tears in which the diagnostic accuracy was 90%. In this group, we found 70 patients in whom the original MR diagnosis did not agree with the surgical findings. Three musculoskeletal radiologists independently reviewed each of the 70 MR examinations without knowledge of the original interpretation or the surgical findings. Their interpretations and the MR images then were correlated with the surgical findings. The original incorrect diagnoses were categorized as being due to unavoidable errors, errors in interpretation, or errors made because of equivocal MR findings of a tear. Unavoidable errors were defined as false-positive and false-negative diagnoses that could not be avoided, even in retrospect.
Of the 83 original diagnostic errors made in the MR evaluation of 800 menisci, 33 (40%) were unavoidable errors, 32 (39%) were due to equivocal MR findings, and 18 (21%) were due to interpretation errors. The unavoidable errors consisted of 21 missed meniscal tears and 12 false-positive MR diagnoses. In the false-positive cases, the menisci were interpreted as torn on MR images by all three observers, but no tear was found at arthroscopy. Subtle MR findings that were equivocal for a tear caused both false-positive and false-negative diagnoses. Seven of the 18 interpretation errors occurred when normal variants were mistaken for a tear.
Using conventional coronal and sagittal spin-echo MR imaging, we could not identify 21 (6%) of the 333 meniscal tears, even in retrospect. In addition, subtle findings that are equivocal for a tear may still make MR diagnosis of every torn meniscus difficult even for experienced radiologists. Unavoidable false-positive diagnoses due to healed tears or tears missed at arthroscopy are an infrequent problem occurring in only 1.5% of the original 800 menisci evaluated with MR imaging. |
---|---|
ISSN: | 0361-803X 1546-3141 |
DOI: | 10.2214/ajr.163.6.7992739 |