PI-RADS version 2: quantitative analysis aids reliable interpretation of diffusion-weighted imaging for prostate cancer

Sung Yoon Park, Su Jin Shin, Dae Chul Jung, Nam Hoon Cho, Young Deuk Choi, Koon Ho Rha, Sung Joon Hong, Young Taik Oh

Research output: Contribution to journalArticle

13 Citations (Scopus)

Abstract

Objectives: To determine whether apparent diffusion coefficient (ADC) ratio aids reliable interpretation of diffusion-weighted imaging (DWI) for prostate cancer (PCa). Methods: Seventy-six consecutive patients with PCa who underwent DWI and surgery were included. Based on pathologic tumour location, two readers independently performed DWI scoring according to the revised Prostate Imaging Reporting and Data System (PI-RADSv2). ADC ratios of benign to cancerous prostatic tissue were then measured independently and compared between cases showing concordant and discordant DWI scores ≥4. Area under the curve (AUC) and threshold of ADC ratio were analyzed for DWI scores ≥4. Results: The rate of inter-reader disagreement for DWI score ≥4 was 11.8% (9/76). ADC ratios were higher in concordant vs. discordant DWI scores ≥4 (median, 1.7 vs. 1.1–1.2; p < 0.001). For DWI scores ≥4, the AUCs of ADC ratios were 0.970 for reader 1 and 0.959 for reader 2. In patients with an ADC ratio >1.3, the rate of inter-reader disagreement for DWI score ≥4 decreased to 5.9–6.0%. An ADC ratio >1.3 yielded 100% (reader 1, 54/54; reader 2, 51/51) positive predictive value for clinically significant cancer. Conclusion: ADC ratios may be useful for reliable interpretation of DWI score ≥4 in PI-RADSv2. Key points: • The ADC ratio correlated positively with DWI score of PI-RADSv2. • ADC ratio >1.3 was associated with concordant interpretation of DWI score ≥4. • ADC ratio >1.3 was associated with high PPV for clinically significant cancer. • ADC ratio is useful for reliable interpretation of DWI scoring in PI-RADSv2.

Original languageEnglish
Pages (from-to)2776-2783
Number of pages8
JournalEuropean Radiology
Volume27
Issue number7
DOIs
Publication statusPublished - 2017 Jul 1

    Fingerprint

All Science Journal Classification (ASJC) codes

  • Radiology Nuclear Medicine and imaging

Cite this