Abdomen, Liver, Oncology, MR, CT, Observer performance, Staging, Cancer, Cirrhosis, Workforce
M. Sini, M. Erta, G. Corrias, L. Saba; Cagliari/IT
Methods and materials
Retrospective study involving 42 patients belonging to the Duilio Casula Polyclinic of Monserrato in the reference period between 01.01.2016 and 31.07.2018.
Patients were not diagnosed with HCC before the radiological examination: almost all of these were initial examinations following the detection of a lesion of the ultrasound check performed under surveillance for HCC; some of them were second examination given the non-conclusiveness of the first imaging exam.
28 patients performed CT,
14 patients performed MR.
The evaluation of the examination was performed by 3 expert residents and 3 residents with little experience first with analytical reporting method,
based on the main guidelines currently in use,
and subsequently with a structured reporting method according to the LI-RADS 2018.
One of the residents never had experience with MRI and thus the MRI were not showed to them.
All residents had never read the LIRADS manual previously and few of them had heard of its existence.
They were asked to firstly analyze and report on a 1 to 5 scale (being 1 an certainly bening lesion and 5 a certainly HCC) the examinations and after having read at least twice the LIRADS v 2018 manual to analyze and report according to the LIRADS guidelines.
A three weeks period interval was set between the two readings.
Readers were divided in two groups: residents with less than 3 years of experience,
and readers with more than 3 years of experience (green rows in tables and yellow rows in tables).
Cohen's Kappa statistic was used to evaluate the degree of agreement between each pair of readers,
before and after the LIRADS application.