Sheela Agarwal MD, MS, Abstract Co-Author: Nothing to Disclose

Duangkamon Prapruttam MD, Abstract Co-Author: Nothing to Disclose

Sandeep Subhash Hedgire MD, Abstract Co-Author: Nothing to Disclose

Jennifer W. Uyeda MD, Presenter: Nothing to Disclose

Cinthia Cruz MD, Abstract Co-Author: Nothing to Disclose

Mukesh Gobind Harisinghani MD, Abstract Co-Author: Nothing to Disclose

Creation of MinIPs marginally improved detection of hepatic lesions irrespective of reader experience level. Given increasing utilization of surgical resection for colorectal metastases, this small improvement in detection may be clinically significant. There was no significant difference in the interpretation time for standard versus MinIPs. Qualitatively, readers found MinIPs to be particularly useful in detection of lesions less than 10mm.

MIP reconstruction algorithms have proven to increase detection of lung nodules on MDCT, due to increasing number of slices in conventional imaging and large amount of “anatomic noise” in the lung. Analogously, MinIPs may lead to improved sensitivity for detection of liver lesions on hepatobiliary phase (HBP) imaging. Accurate detection of all hepatic lesions is particularly relevant in the setting of colorectal cancer, where multi-lesion resections are being performed with increasing frequency given survival benefits.

Patients were screened using inclusion criteria of biopsoy proven colon cancer, gadoxetate MRI, and at least 1 lesion less than 10mm in maximal diameter. MinIPs were created on axial HBP (20 min) images, with slice thickness of 10mm and reconstruction increment of 2.5mm, using AGFA PACS post-processing software. HBP and MinIP images were independently evaluated by 3 readers with varied experience, in random order. Interpretation time was recorded for each sequence. An expert reader who reviewed all available sequences and pertinent clinical details served as the reference standard. The Wilcoxon matched-pairs signed rank test and confidence intervals for differences between methods were used to compare the sensitivities of the two methods.

All 24 exams, with a total of 243 liver lesions (range: 4-29 lesions), were considered of diagnostic quality and included in the study. Avg lesion size was 9.5mm, and 70% of lesions were less than 10mm. On standard HBP, 3 readers missed an average of 1.24 lesions per exam, with an overall detection rate of 88%. On MinIPs, the 3 readers missed an average of 0.79 lesions per exam with an overall detection rate of 92%. Average read time was 2:45 min for HBP and 2:57 min for MinIPs.

Agarwal, S, Prapruttam, D, Hedgire, S, Uyeda, J, Cruz, C, Harisinghani, M, Incremental Value of Minimum Intensity Projection Images on Hepatobiliary Phase Imaging for Detection of Hepatic Lesions.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14016519.html