Deep Learning Applications in Perfusion MRI: Recent Advances and Current Challenges
2
조회
0
다운로드
1
인용
0
참고문헌
키워드
초록
Deep learning has shown its feasibility for applications in medical imaging. Deep learning- based methods are also rapidly being applied in a wide range of areas to replace traditional model-based methods, showing remarkable improvements in several MR image processing areas such as image reconstruction, image contrast conversion, and image quality improvement. With improvement of perfusion MRI techniques, various clinical applications have been also researched, which have improved tracer-kinetic modeling, and vice versa. Representatively, increased vascularity due to tumor angiogenesis, altered permeability due to blood-brain barrier breakdown, and quantifiable absolute cerebral blood perfusion can be imaged via perfusion MRI techniques. As a result, a large number of retrospective and prospective studies have proven that perfusion MRI can be used clinically to investigate various diseases (ranging from brain tumor, stroke, and migraine to neurodegenerative diseases such as Alzheimer’s disease) and various organs including brain, breast, prostate, and pancreas. However, tracer-kinetic model-based processing of parametric maps requires many physical assumptions, which is time-consuming with limitations in clinical applications. With current rapid development of deep learning technologies, the importance of the role of deep learning in MR imaging research appears to be increasing. Specifically for processing perfusion MRI, deep learning has its merit as an end-to-end model-free approach, resulting in markedly reduced processing time compared to conventional iterative methods thanks to its nearly instantaneous inference despite long training time. In this review, first, basic principles of MR physics in perfusion MRI are described. Recent progress and current challenges in both technical improvement and clinical applications of perfusion MRI using deep learning techniques are then summarized.
Abstract
Deep learning has shown its feasibility for applications in medical imaging. Deep learning- based methods are also rapidly being applied in a wide range of areas to replace traditional model-based methods, showing remarkable improvements in several MR image processing areas such as image reconstruction, image contrast conversion, and image quality improvement. With improvement of perfusion MRI techniques, various clinical applications have been also researched, which have improved tracer-kinetic modeling, and vice versa. Representatively, increased vascularity due to tumor angiogenesis, altered permeability due to blood-brain barrier breakdown, and quantifiable absolute cerebral blood perfusion can be imaged via perfusion MRI techniques. As a result, a large number of retrospective and prospective studies have proven that perfusion MRI can be used clinically to investigate various diseases (ranging from brain tumor, stroke, and migraine to neurodegenerative diseases such as Alzheimer’s disease) and various organs including brain, breast, prostate, and pancreas. However, tracer-kinetic model-based processing of parametric maps requires many physical assumptions, which is time-consuming with limitations in clinical applications. With current rapid development of deep learning technologies, the importance of the role of deep learning in MR imaging research appears to be increasing. Specifically for processing perfusion MRI, deep learning has its merit as an end-to-end model-free approach, resulting in markedly reduced processing time compared to conventional iterative methods thanks to its nearly instantaneous inference despite long training time. In this review, first, basic principles of MR physics in perfusion MRI are described. Recent progress and current challenges in both technical improvement and clinical applications of perfusion MRI using deep learning techniques are then summarized.
참고문헌 (57)
- 이두희, "Deep Learning in MR Image Processing", Investigative Magnetic Resonance Imaging, 23(2), pp.81~99 (2019), 대한자기공명의과학회
- Kyu Sung Choi, "Artificial Intelligence in Neuroimaging: Clinical Applications", Investigative Magnetic Resonance Imaging, 26(1), pp.1~9 (2022), 대한자기공명의과학회
- Xiang L, "Deep embedding convolutional neural network for synthesizing CT image from T1-Weighted MR image", Med Image Anal, 47, pp.31~44 (2018)
- Han X, "MR-based synthetic CT generation using a deep convolutional neural network method", Med Phys, 44, pp.1408~1419 (2017)
- Liu F, "Deep learning MR imaging-based attenuation correction for PET", MR imaging, 2018, pp.286, 676~684
- Leu K, "Improved leakage correction for single-echo dynamic susceptibility contrast perfusion MRI estimates of relative cerebral blood volume in highgrade gliomas by accounting for bidirectional contrast agent exchange", AJNR Am J Neuroradiol, 37, pp.1440~1446 (2016)
- Li X, "A novel AIF tracking method and comparison of DCE-MRI parameters using individual and population-based AIFs in human breast cancer", Phys Med Biol, 56, pp.5753~5769 (2011)
- Kim KH, "Improving arterial spin labeling by using deep learning", Radiology, 287, pp.658~666 (2018)
- Pollock JM, "Migraine associated cerebral hyperperfusion with arterial spin-labeled MR imaging", AJNR Am J Neuroradiol, 29, pp.1494~1497 (2008)
- Wang K, "Deep learning detection of penumbral tissue on arterial spin labeling in stroke", Stroke, 51, pp.489~497 (2020)
- van Herten RLM, "Physics-informed neural networks for myocardial perfusion MRI quantification", Med Image Anal, 78, pp.102399~ (2022)
- Ho KC, "Predicting ischemic stroke tissue fate using a deep convolutional neural network on source magnetic resonance perfusion images", J Med Imaging(Bellingham), 6, pp.026001~ (2019)
- Yu Y, "Prediction of hemorrhagic transformation severity in acute stroke from source perfusion MRI", IEEE Trans Biomed Eng, 65, pp.2058~2065 (2018)
- Choi KS, "Prediction of IDH genotype in gliomas with dynamic susceptibility contrast perfusion MR imaging using an explainable recurrent neural network", Neuro Oncol, 21, pp.1197~1209 (2019)
- Shim KY, "Radiomics-based neural network predicts recurrence patterns in glioblastoma using dynamic susceptibility contrast-enhanced MRI", Sci Rep, 11, pp.9974~ (2021)
- Sanders JW, "Synthetic generation of DSC-MRI-derived relative CBV maps from DCE MRI of brain tumors", Magn Reson Med, 85, pp.469~479 (2021)
- Yu Y, "Use of deep learning to predict final ischemic stroke lesions from initial magnetic resonance imaging", JAMA Netw Open, 3, pp.e200772~ (2020)
- Zhang L, "Improving sensitivity of arterial spin labeling perfusion MRI in Alzheimer’s disease using transfer learning of deep learning-based ASL denoising", J Magn Reson Imaging, 55, pp.1710~1722 (2022)
- Ottens T, "Deep learning DCE-MRI parameter estimation : application in pancreatic cancer", Med Image Anal, 80, pp.102512~ (2022)
- Fashandi H, "An investigation of the effect of fat suppression and dimensionality on the accuracy of breast MRI segmentation using U-nets", Med Phys, 46, pp.1230~1244 (2019)
- Bitencourt AGV, "MRI-based machine learning radiomics can predict HER2 expression level and pathologic response after neoadjuvant therapy in HER2overexpressing breast cancer", EBioMedicine, 61, pp.103042~ (2020)
- 장건호, "Perfusion Magnetic Resonance Imaging: A Comprehensive Update on Principles and Techniques", Korean Journal of Radiology, 15(5), pp.554~577 (2014), 대한영상의학회[DOI]
- McGehee BE, "Brain perfusion imaging : how does it work and what should I use?", J Magn Reson Imaging, 36, pp.1257~1272 (2012)
- Choi KS, "Improving the reliability of pharmacokinetic parameters at dynamic contrastenhanced MRI in astrocytomas : a deep learning approach", Radiology, 297, pp.178~188 (2020)
- Buxton RB, "A general kinetic model for quantitative perfusion imaging with arterial spin labeling", Magn Reson Med, 40, pp.383~396 (1998)
- McKinley R, "A machine learning approach to perfusion imaging with dynamic susceptibility contrast MR", Front Neurol, 9, pp.717~ (2018)
- Kudo K, "Differences in CT perfusion maps generated by different commercial software : quantitative analysis by using identical source data of acute stroke patients", Radiology, 254, pp.200~209 (2010)
- Fang K, "Convolutional neural network for accelerating the computation of the extended Tofts model in dynamic contrast-enhanced magnetic resonance imaging", J Magn Reson Imaging, 53, pp.1898~1910 (2021)
- Park JE, "Diffusion and perfusion MRI radiomics obtained from deep learning segmentation provides reproducible and comparable diagnostic model to human in posttreatment glioblastoma", Eur Radiol, 31, pp.3127~3137 (2021)
- Kim M, "Diffusion-and perfusion-weighted MRI radiomics model may predict isocitrate dehydrogenase(IDH)mutation and tumor aggressiveness in diffuse lower grade glioma", Eur Radiol, 30, pp.2142~2151 (2020)
- Stupp R, "Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma", N Engl J Med, 352, pp.987~996 (2005)
- Lee J, "Discriminating pseudoprogression and true progression in diffuse infiltrating glioma using multiparametric MRI data through deep learning", Sci Rep, 10, pp.20331~ (2020)
- Li M, "DC-AL GAN : pseudoprogression and true tumor progression of glioblastoma multiform image classification based on DCGAN and AlexNet", Med Phys, 47, pp.1139~1150 (2020)
- Kim JY, "Incorporating diffusion-and perfusion-weighted MRI into a radiomics model improves diagnostic performance for pseudoprogression in glioblastoma patients", Neuro Oncol, 21, pp.404~414 (2019)
- Cha J, "Differentiation of tumor progression from pseudoprogression in patients with posttreatment glioblastoma using multiparametric histogram analysis", AJNR Am J Neuroradiol, 35, pp.1309~1317 (2014)
- Pak Elena, "Prediction of Prognosis in Glioblastoma Using Radiomics Features of Dynamic Contrast-Enhanced MRI", Korean Journal of Radiology, 22(9), pp.1514~1524 (2021), 대한영상의학회[DOI]
- Park JE, "Deep-learned time-signal intensity pattern analysis using an autoencoder captures magnetic resonance perfusion heterogeneity for brain tumor differentiation", Sci Rep, 10, pp.21485~ (2020)
- Kim AR, "Absolute quantification of tumor-infiltrating immune cells in high-grade glioma identifies prognostic and radiomics values", Cancer Immunol Immunother, 70, pp.1995~2008 (2021)
- Park YW, "Differentiating patients with schizophrenia from healthy controls by hippocampal subfields using radiomics", Schizophr Res, 223, pp.337~344 (2020)
- Ho KC, "A machine learning approach for classifying ischemic stroke onset time from imaging", IEEE Trans Med Imaging, 38, pp.1666~1676 (2019)
- Chen DYT, "Predicting PET cerebrovascular reserve with deep learning by using baseline MRI : a pilot investigation of a drug-free brain stress test", Radiology, 296, pp.627~637 (2020)
- Collij LE, "Application of machine learning to arterial spin labeling in mild cognitive impairment and Alzheimer disease", Radiology, 281, pp.865~875 (2016)
- Raissi M, "Physics-informed neural networks : a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations", J Comput Phys, 378, pp.686~707 (2019)
- Dey R, "Gate-variants of gated recurrent unit (GRU)neural networks", Proceedings of the 2017 IEEE 60th International Midwest Symposium on Circuits and Systems (MWSCAS), pp.1597~1600 (2017), IEEE
- Hochreiter S, "Long short-term memory", Neural Comput, 9, pp.1735~1780 (1997)
- Cuocolo R, "Machine learning applications in prostate cancer magnetic resonance imaging", Eur Radiol Exp, 3, pp.35~ (2019)
- Reig B, "Machine learning in breast MRI", J Magn Reson Imaging, 52, pp.998~1018 (2020)
- Hu Q, "A deep learning methodology for improved breast cancer diagnosis using multiparametric MRI", Sci Rep, 10, pp.10536~ (2020)
- Sutton EJ, "A machine learning model that classifies breast cancer pathologic complete response on MRI post-neoadjuvant chemotherapy", Breast Cancer Res, 22, pp.57~ (2020)
- Wu J, "Unsupervised clustering of quantitative image phenotypes reveals breast cancer subtypes with distinct prognoses and molecular pathways", Clin Cancer Res, 23, pp.3334~3342 (2017)
- Gulani V, "Gadolinium deposition in the brain : summary of evidence and recommendations", Lancet Neurol, 16, pp.564~570 (2017)
- Ribeiro MT, "“Why should I trust you?” Explaining the predictions of any classifier", Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2016, pp.1135~1144 (2016), KDD
- Lundberg SM, "A unified approach to interpreting model predictions", Proceedings of the 31st International Conference on Neural Information Processing Systems, pp.30~ (2017)
- Selvaraju RR, "Grad-cam: visual explanations from deep networks via gradient-based localization", Proceedings of the IEEE International Conference on Computer Vision (ICCV), pp.618~626 (2017), IEEE
- Carré A, "Standardization of brain MR images across machines and protocols : bridging the gap for MRIbased radiomics", Sci Rep, 10, pp.12340~ (2020)
- Sellergren AB, "Simplified transfer learning for chest radiography models using less data", Radiology, 305, pp.454~465 (2022)
- Candemir S, "Training strategies for radiology deep learning models in data-limited scenarios", Radiol Artif Intell, 3, pp.e210014~ (2021)