All over the map: An interobserver agreement study of tumor location based on the PI‐RADSv2 sector map

Abstract

3 Technical Efficacy: Stage 2 J. MAGN. RESON. IMAGING 2018;48:482-490.

Intra‐ and interreader reproducibility of PI‐RADSv2: A multireader study

“…Despite working with PI‐RADSv2 since 2015, readers of all experience levels continue to have difficulty agreeing with each other on category assignment, especially Category 3 . Readers have been shown to have poor agreement when attempting to localize identified lesions on a PI‐RADSv2 sector map . In addition, readers have also been shown to have difficulty in applying the PI‐RADSv2 lexicon, as expert readers only show moderate agreement when using it .…”

“…13 Readers have been shown to have poor agreement when attempting to localize identified lesions on a PI-RADSv2 sector map. 33 In addition, readers have also been shown to have difficulty in applying the PI-RADSv2 lexicon, as expert readers only show moderate agreement when using it. 14 It is also possible that the current PI-RADSv2 lexicon does not yet encompass the diverse manifestations of prostate cancer on mpMRI.…”

re: Interobserver Agreement of PI‐RADS v. 2 Lexicon Among Radiologists With Different Levels of Experience

“…The published institutional Likert‐based scoring systems such as UCLA (2011) or NIH (2012), which predated PI‐RADS, are usually more quantitative and have some common features with PI‐RADSv2, which might be the reason for the better agreement among the higher‐experienced readers in the current study as well. Although some previous studies have selected the readers from different centers to eliminate experience bias, they have shown similar results . Finally, the most important role of MRI is to maximize detection of clinically significant PCa lesions (csPCa), as these high‐grade tumors behave more aggressively and show more aggressive features on T 2 , DWI, and DCE.…”

“…The qualitative nature of the components of the PI‐RADSv2, which are mostly the same in PI‐RADSv2.1, has resulted in a range of reported sensitivity and specificity . To address this variation, multiple studies have evaluated the interreader agreement for PI‐RADSv2, mostly showing moderate agreement for lesion level final scoring . However, we have to ask the questions: Which parts of the system are most challenging for readers?…”

“…3 To address this variation, multiple studies have evaluated the interreader agreement for PI-RADSv2, mostly showing moderate agreement for lesion level final scoring. [4][5][6] However, we have to ask the questions: Which parts of the system are most challenging for readers? And what should be done to have a more standardized interpretation?…”