PurposeTo compare the content and communication between the radiologist and the clinicians for treatment planning of structured reports (SRs) and narrative reports (NRs) for reporting CT/MRI findings in patients with perihilar cholangiocarcinoma.Materials and methodThis retrospective study included 54 patients with perihilar cholangiocarcinoma who underwent CT/MRI before ERCP or surgery. For all patients, we generated both NRs and SRs for perihilar cholangiocarcinoma and compared the number of key features between NRs and SRs. In addition, three clinicians performed a questionnaire evaluation that included three questions regarding assessment of the sufficiency of information for surgical or procedural planning, the effort required for information extraction, and the report quality rated on a Likert scale.ResultsSRs included significantly more predefined key features (6.89 ± 0.31) than NRs (5.87 ± 0.70) (p < 0.001). SRs provided greater sufficiency of information for clinical planning than NRs (89.9% vs. 18.5% of the cases, respectively; p < 0.001). Extraction of information was easier from SRs than NRs (94.4% vs. 9.3%, respectively) (p < 0.001). SRs received a higher overall report quality rating (5.96 ± 0.19) than NRs (4.31 ± 0.77) (p < 0.001).ConclusionsSRs of CT/MRI findings for hilar cholangiocarcinoma can reveal more predefined key features, provide more sufficiency of information, and yield higher satisfaction levels, in comparison with NRs.