Machine learning based prediction of image quality in prostate MRI using rapid localizer images

被引:1
作者
Al-Hayali, Abdullah [1 ]
Komeili, Amin [2 ]
Azad, Azar [3 ]
Sathiadoss, Paul [4 ]
Schieda, Nicola [4 ]
Ukwatta, Eranga [1 ]
机构
[1] Univ Guelph, Guelph Imaging Lab, Sch Engn, Guelph, ON, Canada
[2] Univ Calgary, Dept Biomed Engn, Calgary, AB, Canada
[3] AI Vali Inc, Toronto, ON, Canada
[4] Univ Ottawa, Dept Radiol, Ottawa, ON, Canada
关键词
prostate; magnetic resonance imaging MRI; quality; gas; artifact; TEXTURAL FEATURES; CANCER; RADIOMICS; SYSTEM; IMPACT; ENEMA; STAGE;
D O I
10.1117/1.JMI.11.2.026001
中图分类号
R8 [特种医学]; R445 [影像诊断学];
学科分类号
1002 ; 100207 ; 1009 ;
摘要
Purpose: Diagnostic performance of prostate MRI depends on high-quality imaging. Prostate MRI quality is inversely proportional to the amount of rectal gas and distention. Early detection of poor-quality MRI may enable intervention to remove gas or exam rescheduling, saving time. We developed a machine learning based quality prediction of yet-to-be acquired MRI images solely based on MRI rapid localizer sequence, which can be acquired in a few seconds. Approach: The dataset consists of 213 (147 for training and 64 for testing) prostate sagittal T2-weighted (T2W) MRI localizer images and rectal content, manually labeled by an expert radiologist. Each MRI localizer contains seven two-dimensional (2D) slices of the patient, accompanied by manual segmentations of rectum for each slice. Cascaded and end-to-end deep learning models were used to predict the quality of yet-to-be T2W, DWI, and apparent diffusion coefficient (ADC) MRI images. Predictions were compared to quality scores determined by the experts using area under the receiver operator characteristic curve and intra-class correlation coefficient. Results: In the test set of 64 patients, optimal versus suboptimal exams occurred in 95.3% (61/64) versus 4.7% (3/64) for T2W, 90.6% (58/64) versus 9.4% (6/64) for DWI, and 89.1% (57/64) versus 10.9% (7/64) for ADC. The best performing segmentation model was 2D U-Net with ResNet-34 encoder and ImageNet weights. The best performing classifier was the radiomics based classifier. Conclusions: A radiomics based classifier applied to localizer images achieves accurate diagnosis of subsequent image quality for T2W, DWI, and ADC prostate MRI sequences. (c) 2024 Society of Photo-Optical Instrumentation Engineers (SPIE)
引用
收藏
页数:14
相关论文
共 46 条
[1]   Decoding tumour phenotype by noninvasive imaging using a quantitative radiomics approach [J].
Aerts, Hugo J. W. L. ;
Velazquez, Emmanuel Rios ;
Leijenaar, Ralph T. H. ;
Parmar, Chintan ;
Grossmann, Patrick ;
Cavalho, Sara ;
Bussink, Johan ;
Monshouwer, Rene ;
Haibe-Kains, Benjamin ;
Rietveld, Derek ;
Hoebers, Frank ;
Rietbergen, Michelle M. ;
Leemans, C. Rene ;
Dekker, Andre ;
Quackenbush, John ;
Gillies, Robert J. ;
Lambin, Philippe .
NATURE COMMUNICATIONS, 2014, 5
[2]   Image Quality Assessment by Comparing CNN Features between Images [J].
Amirshahi, Seyed Ali ;
Pedersen, Marius ;
Yu, Stella X. .
JOURNAL OF IMAGING SCIENCE AND TECHNOLOGY, 2016, 60 (06)
[3]   Detection of Clinically Significant Prostate Cancer: Short Dual-Pulse Sequence versus Standard Multiparametric MR Imaging-A Multireader Study [J].
Barth, Borna K. ;
De Visschere, Pieter J. L. ;
Cornelius, Alexander ;
Nicolau, Carlos ;
Vargas, Hebert Alberto ;
Eberli, Daniel ;
Donati, Olivio F. .
RADIOLOGY, 2017, 284 (03) :725-736
[4]   Estimation of an inter-rater intra-class correlation coefficient that overcomes common assumption violations in the assessment of health measurement scales [J].
Bobak, Carly A. ;
Barr, Paul J. ;
O'Malley, A. James .
BMC MEDICAL RESEARCH METHODOLOGY, 2018, 18
[5]   Deep Neural Networks for No-Reference and Full-Reference Image Quality Assessment [J].
Bosse, Sebastian ;
Maniry, Dominique ;
Mueller, Klaus-Robert ;
Wiegand, Thomas ;
Samek, Wojciech .
IEEE TRANSACTIONS ON IMAGE PROCESSING, 2018, 27 (01) :206-219
[6]   Evaluating the effect of rectal distension on prostate multiparametric MRI image quality [J].
Caglic, Iztok ;
Hansen, Nienke L. ;
Slough, Rhys A. ;
Patterson, Andrew J. ;
Barrett, Tristan .
EUROPEAN JOURNAL OF RADIOLOGY, 2017, 90 :174-180
[7]   SMOTE: Synthetic minority over-sampling technique [J].
Chawla, Nitesh V. ;
Bowyer, Kevin W. ;
Hall, Lawrence O. ;
Kegelmeyer, W. Philip .
2002, American Association for Artificial Intelligence (16)
[8]   Prebiopsy Biparametric MRI for Clinically Significant Prostate Cancer Detection With PI-RADS Version 2: A Multicenter Study [J].
Choi, Moon Hyung ;
Kim, Chan Kyo ;
Lee, Young Joon ;
Jung, Seung Eun .
AMERICAN JOURNAL OF ROENTGENOLOGY, 2019, 212 (04) :839-846
[9]   Convolutional Neural Networks for Automated Classification of Prostate Multiparametric Magnetic Resonance Imaging Based on Image Quality [J].
Cipollari, Stefano ;
Guarrasi, Valerio ;
Pecoraro, Martina ;
Bicchetti, Marco ;
Messina, Emanuele ;
Farina, Lorenzo ;
Paci, Paola ;
Catalano, Carlo ;
Panebianco, Valeria .
JOURNAL OF MAGNETIC RESONANCE IMAGING, 2022, 55 (02) :480-490
[10]   MRI features predictive of negative surgical margins in patients with HER2 overexpressing breast cancer undergoing breast conservation [J].
Dashevsky, Brittany Z. ;
Oh, Jung Hun ;
Apte, Aditya P. ;
Bernard-Davila, Blanca ;
Morris, Elizabeth A. ;
Deasy, Joseph O. ;
Sutton, Elizabeth J. .
SCIENTIFIC REPORTS, 2018, 8