Abstract
Objectives
To develop an automated model to detect brain metastases using a convolutional neural network (CNN) and volume isotropic simultaneous interleaved bright-blood and black-blood examination (VISIBLE) and to compare its diagnostic performance with the observer test.
Methods
This retrospective study included patients with clinical suspicion of brain metastases imaged with VISIBLE from March 2016 to July 2019 to create a model. Images with and without blood vessel suppression were used for training an existing CNN (DeepMedic). Diagnostic performance was evaluated using sensitivity and false-positive results per case (FPs/case). We compared the diagnostic performance of the CNN model with that of the twelve radiologists.
Results
Fifty patients (30 males and 20 females; age range 29–86 years; mean 63.3 ± 12.8 years; a total of 165 metastases) who were clinically diagnosed with brain metastasis on follow-up were used for the training. The sensitivity of our model was 91.7%, which was higher than that of the observer test (mean ± standard deviation; 88.7 ± 3.7%). The number of FPs/case in our model was 1.5, which was greater than that by the observer test (0.17 ± 0.09).
Conclusions
Compared to radiologists, our model created by VISIBLE and CNN to diagnose brain metastases showed higher sensitivity. The number of FPs/case by our model was greater than that by the observer test of radiologists; however, it was less than that in most of the previous studies with deep learning.
Key Points
• Our convolutional neural network based on bright-blood and black-blood examination to diagnose brain metastases showed a higher sensitivity than that by the observer test.
• The number of false-positives/case by our model was greater than that by the previous observer test; however, it was less than those from most previous studies.
• In our model, false-positives were found in the vessels, choroid plexus, and image noise or unknown causes.
Similar content being viewed by others
Abbreviations
- 3D :
-
Three-dimensional
- CNN :
-
Convolution neural network
- FP :
-
False-positive
- MSDE :
-
Motion-sensitized driven equilibrium
- VISIBLE:
-
Volume isotropic simultaneous interleaved bright and black-blood examination
References
Posner J, Chernik N (1978) Intracranial metastases from systemic cancer. Adv Neurol 19:579–592
Rosner D, Nemoto T, Pickren J, Lane W (1983) Management of brain metastases from breast cancer by combination chemotherapy. J Neurooncol 1:131–137
Linskey ME, Andrews DW, Asher AL et al (2010) The role of stereotactic radiosurgery in the management of patients with newly diagnosed brain metastases: a systematic review and evidence-based clinical practice guideline. J Neurooncol 96:45–68
Lippitz B, Lindquist C, Paddick I, Peterson D, O’Neill K, Beaney R (2014) Stereotactic radiosurgery in the treatment of brain metastases: the current evidence. Cancer Treat Rev 40:48–59
Monaco EA 3rd, Faraji AH, Berkowitz O et al (2013) Leukoencephalopathy after whole-brain radiation therapy plus radiosurgery versus radiosurgery alone for metastatic lung cancer. Cancer 119:226–232
Yamamoto M, Serizawa T, Higuchi Y et al (2017) A Multi-institutional prospective observational study of stereotactic radiosurgery for patients with multiple brain metastases (JLGK0901 Study Update): irradiation-related complications and long-term maintenance of Mini-Mental State Examination Scores. Int J Radiat Oncol Biol Phys 99:31–40
Sunwoo L, Kim YJ, Choi SH et al (2017) Computer-aided detection of brain metastasis on 3D MR imaging: observer performance study. PloS One 12:e0178265
Losch M (2015) Detection and segmentation of brain metastases with deep convolutional networks. Master’s Thesis, KTH Royal Institute of Technology in Stockholm.
Charron O, Lallement A, Jarnet D, Noblet V, Clavier JB, Meyer P (2018) Automatic detection and segmentation of brain metastases on multimodal MR images with a deep convolutional neural network. Comput Biol Med 95:43–54
Grovik E, Yi D, Iv M, Tong E, Rubin D, Zaharchuk G (2020) Deep learning enables automatic detection and segmentation of brain metastases on multisequence MRI. J Magn Reson Imaging 51:175–182
Dikici E, Ryu JL, Demirer M et al (2020) Automated brain metastases detection framework for T1-weighted contrast-enhanced 3D MRI. IEEE J Biomed Health Inform 24:2883–2893
Bousabarah K, Ruge M, Brand JS et al (2020) Deep convolutional neural networks for automated segmentation of brain metastases trained on clinical data. Radiat Oncol 15:87
Zhou Z, Sanders JW, Johnson JM et al (2020) Computer-aided detection of brain metastases in T1-weighted MRI for stereotactic radiosurgery using deep learning single-shot detectors. Radiology 295:407–415
Zhang M, Young GS, Chen H et al (2020) Deep-learning detection of cancer metastases to the brain on MRI. J Magn Reson Imaging 52:1227–1236
Park YW, Jun Y, Lee Y et al (2021) Robust performance of deep learning for automatic detection and segmentation of brain metastases using three-dimensional black-blood and three-dimensional gradient echo imaging. Eur Radiol. https://doi.org/10.1007/s00330-021-07783-3
Schellinger PD, Meinck HM, Thron A (1999) Diagnostic accuracy of MRI compared to CCT in patients with brain metastases. J Neurooncol 44:275–281
Kakeda S, Korogi Y, Hiai Y et al (2007) Detection of brain metastasis at 3T: comparison among SE, IR-FSE and 3D-GRE sequences. Eur Radiol 17:2345–2351
Takeda T, Takeda A, Nagaoka T et al (2008) Gadolinium-enhanced three-dimensional magnetization-prepared rapid gradient-echo (3D MP-RAGE) imaging is superior to spin-echo imaging in delineating brain metastases. Acta Radiol 49:1167–1173
Komada T, Naganawa S, Ogawa H (2008) Contrast-enhanced MR imaging of metastatic brain tumor at 3 tesla: utility of T (1)-weighted SPACE compared with 2D spin echo and 3D gradient echo sequence. Magn Reson Med Sci 7:13–21
Kato Y, Higano S, Tamura H et al (2009) Usefulness of contrast-enhanced T1-weighted sampling perfection with application-optimized contrasts by using different flip angle evolutions in detection of small brain metastasis at 3T MR imaging: comparison with magnetization-prepared rapid acquisition of gradient echo imaging. AJNR Am J Neuroradiol 30:923–929
Nagao E, Yoshiura T, Hiwatashi A et al (2011) 3D turbo spin-echo sequence with motion-sensitized driven-equilibrium preparation for detection of brain metastases on 3T MR imaging. AJNR Am J Neuroradiol 32:664–670
Park J, Kim EY (2010) Contrast-enhanced, three-dimensional, whole-brain, black-blood imaging: application to small brain metastases. Magn Reson Med 63:553–561
Yoneyama M, Nakamura M, Tabuchi T et al (2013) Whole-brain black-blood imaging with magnetization-transfer prepared spin echo-like contrast: a novel sequence for contrast-enhanced brain metastasis screening at 3T. Radiol Phys Technol 6:431–436
Kikuchi K, Hiwatashi A, Togao O et al (2015) 3D MR sequence capable of simultaneous image acquisitions with and without blood vessel suppression: utility in diagnosing brain metastases. Eur Radiol 25:901–910
Yoneyama M, Obara M, Takahara T et al (2014) Volume isotropic simultaneous interleaved black- and bright-blood imaging: a novel sequence for contrast-enhanced screening of brain metastasis. Magn Reson Med Sci 13:277–284
Kamnitsas K, Ledig C, Newcombe VFJ et al (2017) Efficient multi-scale 3D CNN with fully connected CRF for accurate brain lesion segmentation. Med Image Anal 36:61–78
Funding
This work was supported by JSPS KAKENHI Grant Number JP21K07645.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Guarantor
The scientific guarantor of this publication is Associate Professor Akio Hiwatashi, MD, PhD, from the Department of Clinical Radiology, Graduate School of Medical Sciences, Kyushu University (hiwatashi.akio.428@m.kyushu-u.ac.jp).
Conflict of interest
The authors of this manuscript declare relationships with the following companies: Philips Japan and Philips Healthcare. The authors of this manuscript declare that MO is an employee of Philips Japan and MVC and AF are employee of Philips Healthcare. They were not involved in data analysis in this study.
Statistics and biometry
No complex statistical methods were necessary for this paper.
Informed consent
Written informed consent was waived by the Institutional Review Board.
Ethical approval
Institutional Review Board approval was obtained.
Study subjects or cohorts overlap
There is an overlap with previous publication (Kikuchi K, Hiwatashi A, Togao O, et al 3D MR Sequence Capable of Simultaneous Image Acquisitions with and without Blood Vessel Suppression: Utility in Diagnosing Brain Metastases. Eur Radiol 2015;25:901–910.) We compared the diagnostic ability of the newly developed brain metastasis diagnosing system with deep learning and that of the previous observer test in this publication.
Methodology
• retrospective
• diagnostic or prognostic study
• performed at one institution
Additional information
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Kikuchi, Y., Togao, O., Kikuchi, K. et al. A deep convolutional neural network-based automatic detection of brain metastases with and without blood vessel suppression. Eur Radiol 32, 2998–3005 (2022). https://doi.org/10.1007/s00330-021-08427-2
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00330-021-08427-2