Details

Autor(en) / Beteiligte

• Choi, Seung Joon
• Kim, Jonghoon
• Seo, Jongbum
• Kim, Hyung Sik
• Lee, Jong-min
• Park, Hyunjin

Titel

Parametric response mapping of dynamic CT for predicting intrahepatic recurrence of hepatocellular carcinoma after conventional transcatheter arterial chemoembolization

Ist Teil von

• European radiology, 2016-01, Vol.26 (1), p.225-234

Ort / Verlag

Berlin/Heidelberg: Springer Berlin Heidelberg

Erscheinungsjahr

2016

Link zum Volltext

Quelle

MEDLINE

Beschreibungen/Notizen

• Objectives The aim of our study was to determine the diagnostic value of a novel image analysis method called parametric response mapping (PRM) for prediction of intrahepatic recurrence of hepatocellular carcinoma (HCC) treated with conventional transcatheter arterial chemoembolization (TACE). Methods This retrospective study was approved by the IRB. We recruited 55 HCC patients who achieved complete remission (CR) after TACE and received longitudinal multiphasic liver computed tomography (CT). The patients fell into two groups: the recurrent tumour group (n = 29) and the non-recurrent tumour group (n = 26). We applied the PRM analysis to see if this technique could distinguish between the two groups. The results of the PRM analysis were incorporated into a prediction algorithm. We retrospectively removed data from the last time point and attempted to predict the response to therapy of the removed data. Results The PRM analysis was able to distinguish between the non-recurrent and recurrent groups successfully. The prediction algorithm detected response to therapy with an area under the curve (AUC) of 0.76, while the manual approach had AUC 0.64. Conclusions Adopting PRM analysis can potentially distinguish between recurrent and non-recurrent HCCs and allow for prediction of response to therapy after TACE. Key Points • Parametric response mapping (PRM) could help assess patients with recurrent HCCs after TACE. • Parametric response mapping could direct patients to individualized therapy. • Longitudinal CT images were analyzed with advanced image analysis method.