Machine learning-based analysis of MR radiomics can help to improve the diagnostic performance of PI-RADS v2 in clinically relevant prostate cancer

• Published in: European radiology Vol. 27; no. 10; pp. 4082 - 4090
• Main Authors: Wang, Jing, Wu, Chen-Jiang, Bao, Mei-Ling, Zhang, Jing, Wang, Xiao-Ning, Zhang, Yu-Dong
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.10.2017; Springer Nature B.V
• Subjects: Aged; Aged, 80 and over; Artificial intelligence; Cancer; Cancer surgery; Correlation analysis; Diagnostic Radiology; Diagnostic systems; Humans; Imaging; Internal Medicine; Interventional Radiology; Learning algorithms; Machine Learning; Magnetic resonance imaging; Magnetic Resonance Imaging - methods; Male; Medical diagnosis; Medicine; Medicine & Public Health; Middle Aged; Neuroradiology; Oncology; Performance enhancement; Predictive Value of Tests; Prostate cancer; Prostatectomy; Prostatic Neoplasms - diagnostic imaging; Prostatic Neoplasms - pathology; Radiology; Radiology Information Systems - standards; Radiomics; ROC Curve; Sensitivity and Specificity; Support Vector Machine; Tumors; Ultrasound; Prostate Imaging Reporting and Data System v2; Multi-parametric MRI; Support vector machine; Prostate cancer; Machine learning
• ISSN: 0938-7994; 1432-1084; 1432-1084
• DOI: 10.1007/s00330-017-4800-5

Objective To investigate whether machine learning-based analysis of MR radiomics can help improve the performance PI-RADS v2 in clinically relevant prostate cancer (PCa). Methods This IRB-approved study included 54 patients with PCa undergoing multi-parametric (mp) MRI before prostatectomy. Imaging analysis was performed on 54 tumours, 47 normal peripheral (PZ) and 48 normal transitional (TZ) zone based on histological-radiological correlation. Mp-MRI was scored via PI-RADS, and quantified by measuring radiomic features. Predictive model was developed using a novel support vector machine trained with: (i) radiomics, (ii) PI-RADS scores, (iii) radiomics and PI-RADS scores. Paired comparison was made via ROC analysis. Results For PCa versus normal TZ, the model trained with radiomics had a significantly higher area under the ROC curve (Az) (0.955 [95% CI 0.923–0.976]) than PI-RADS (Az: 0.878 [0.834–0.914], p < 0.001). The Az between them was insignificant for PCa versus PZ (0.972 [0.945–0.988] vs. 0.940 [0.905–0.965], p = 0.097). When radiomics was added, performance of PI-RADS was significantly improved for PCa versus PZ (Az: 0.983 [0.960–0.995]) and PCa versus TZ (Az: 0.968 [0.940–0.985]). Conclusion Machine learning analysis of MR radiomics can help improve the performance of PI-RADS in clinically relevant PCa. Key Points • Machine - based analysis of MR radiomics outperformed in TZ cancer against PI - RADS . • Adding MR radiomics significantly improved the performance of PI - RADS . • DKI - derived Dapp and Kapp were two strong markers for the diagnosis of PCa .