Next Article in Journal
Feasibility, Prediction and Association of Right Ventricular Free Wall Longitudinal Strain with 30-Day Mortality in Severe COVID-19 Pneumonia: A Prospective Study
Next Article in Special Issue
A Survey of Radiomics in Precision Diagnosis and Treatment of Adult Gliomas
Previous Article in Journal
Colour Duplex and/or Contrast-Enhanced Ultrasound Compared with Computed Tomography Angiography for Endoleak Detection after Endovascular Abdominal Aortic Aneurysm Repair: A Systematic Review and Meta-Analysis
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
JCM
Volume 11
Issue 13
10.3390/jcm11133623
jcm-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Differentiation of Cerebral Dissecting Aneurysm from Hemorrhagic Saccular Aneurysm by Machine-Learning Based on Vessel Wall MRI: A Multicenter Study

by
Xin Cao
1,2,†,
Yanwei Zeng
1,2,†,
Junying Wang
3,†,
Yunxi Cao
4,
Yifan Wu
1 and
Wei Xia
5,*
1
Department of Radiology, Huashan Hospital, Fudan University, Shanghai 200040, China
2
Greater Bay Area Institute of Precision Medicine (Guangzhou), Guangzhou 511466, China
3
Department of Medical Imaging, The First Affiliated Hospital of Shandong First Medical University & Shandong Province Qianfoshan Hospital, Jinan 250014, China
4
Radiology Academy, Shandong First Medical University & Shandong Academy of Medical Sciences, Taian 271016, China
5
Suzhou Institute of Biomedical Engineering and Technology, Chinese Academy of Sciences, Suzhou 215163, China
*
Author to whom correspondence should be addressed.
†
These authors contributed equally to this work.
J. Clin. Med. 2022, 11(13), 3623; https://doi.org/10.3390/jcm11133623
Submission received: 19 May 2022 / Revised: 17 June 2022 / Accepted: 20 June 2022 / Published: 23 June 2022
(This article belongs to the Special Issue Recent Advances in Clinical Neuroimaging)
Browse Figures
Versions Notes

Abstract

:
The differential diagnosis of a cerebral dissecting aneurysm (DA) and a hemorrhagic saccular aneurysm (SA) often depends on the intraoperative findings; thus, improved non-invasive imaging diagnosis before surgery is essential to distinguish between these two aneurysms, in order to provide the correct formulation of surgical procedure. We aimed to build a radiomic model based on high-resolution vessel wall magnetic resonance imaging (VW-MRI) and a machine-learning algorithm. In total, 851 radiomic features from 146 cases were analyzed retrospectively, and the ElasticNet algorithm was used to establish the radiomic model in a training set of 77 cases. A clinico-radiological model using clinical features and MRI features was also built. Then an integrated model was built by combining the radiomic model and clinico-radiological model. The area under the ROC curve (AUC) was used to quantify the performance of models. The models were evaluated using leave-one-out cross-validation in a training set, and further validated in an external test set of 69 cases. The diagnostic performance of experienced radiologists was also assessed for comparison. Eight features were used to establish the radiomic model, and the radiomic model performs better (AUC = 0.831) than the clinico-radiological model (AUC = 0.717), integrated model (AUC = 0.813), and even experienced radiologists (AUC = 0.801). Therefore, a radiomic model based on VW-MRI can reliably be used to distinguish DA and hemorrhagic SA, and, thus, be widely applied in clinical practice.

1. Introduction

A cerebral dissecting aneurysm (DA), as a clinical emergency, has different a pathophysiological mechanism and etiology from a common cerebral saccular aneurysm (SA) [1]. Under arterial pressure, the blood enters the arterial wall from the tear of intima to form intramural hematomas (IMH), which are usually located in the media layer [2]. A prior study shows that the annual rupture rate of DA is nearly 58%, increasing the risk of subarachnoid hemorrhage and ischemic stroke [3,4], while the annual rupture rate of SA is 0.95%. DA has a worse clinical outcome, with higher risk of rupture with than SA [5,6,7]. A diameter greater than 10 mm is an independent risk factor for re-bleeding of an aneurysm [8]. The preferred treatment options of DA are vessel occlusion, surgical trapping, or flow diverter [9], whereas clipping or stenting is more frequently used for SA. Some hemorrhagic SAs with hematoma and thrombus are difficult to differentiate from DA.
Vessel wall magnetic resonance imaging (VW-MRI) provides non-invasive, reliable measurements for the visualization of the vessel wall, lumen, and intimal tear [10,11], which is recommended as an optimal diagnostic technique for DA [12,13,14]. IMH usually shows high signal intensity in T1-weighted imaging (T1WI). Hemorrhagic SA with mural thrombus or blood swirls sometimes has similar MRI signs to DA. Thus, constructing a diagnostic model through machine-learning methods, which fully utilizes the 3D T1WI of VW-MRI data, could be very helpful. Radiomics has gradually been applied in the field of aneurysms. Podgorsak et al. study a convolutional neural network that automatically extracts blood flow-related radiomic features of intracranial aneurysms [15]. Some applications of artificial intelligence techniques for aneurysm detection, risk stratification, and prognosis prediction are reported [16,17]. However, the use of radiomic features to identify the classification of aneurysm has not been studied before. Therefore, this research aimed to build a VW-MRI-based radiomic model for differentiating DA from hemorrhagic SA, and to validate its generalizability.

2. Methods

This retrospective study was approved by the Institutional Review Board of Huashan Hospital, affiliated with Fudan University, and the First Affiliated Hospital of Shandong First Medical University, and the need for informed consent was waived by the ethical committees of both hospitals. We collected medical data retrospectively from January 2017 to January 2022 in both hospitals.

3. Participants

A total of 255 patients who underwent a high-resolution (HR) VW-MRI and had an aneurysm of the artery in the brain or neck detected were recruited. The subjects were included based on the following criteria: (i) final diagnosis based on findings during surgery; (ii) completed the imaging examination within 72 h after admission; (iii) large aneurysm (long diameter ≥ 7 mm) with mixed MRI signals; and iv) the MRI images show details with good or excellent quality. The exclusion criteria were as follows: (i) past neurosurgery history or any other neurological diseases, such as cerebral hemorrhage, brain tumors, ischemic stroke, brain trauma, arteriovenous malformations, etc.; (ii) poor image quality due to motion artifact; and (iii) small aneurysms with homogeneous low signal on MRI. Clinical information was recorded, including age, gender, clinical symptoms, lesion size, and lesion location. The international normalized ratio (INR) was recorded as an indicator of the patient′s coagulation status. Personal information was desensitized prior to analysis. To facilitate subsequent analysis, DA patients were defined as negative samples, and SA patients were marked as positive samples.

3.1. Image Data Acquisition

All MRIs were performed using two 3.0-T magnetic resonance systems of the same model in two hospitals (Discovery MR 750, GE Healthcare, Milwaukee, MI, USA). The patients in our hospital, as training sets, receive scanning parameters as follows: (i) 3D T1WI: repetition time (TR) = 600 ms, echo time (TE) = 14.8 ms, field of view (FOV) = 240 × 240 mm2, matrix = 256 × 256, signal noise ratio (SNR) = 1.00, slice thickness = 1.0 mm, and scanning time = 4′15′′; (ii) time-of-flight (TOF) magnetic resonance angiography (MRA): TR = 23 ms, TE = 3.5 ms, FOV = 160 × 160 mm3, flip angle = 20°, and scanning time = 5′20″. The patients in another hospital were treated as validation sets, and the scanning parameters were as follows: (i) 3D T1WI: TR = 600 ms, TE = 14.4 ms, FOV = 200 × 200 mm2, matrix = 288 × 288, SNR = 1.00, slice thickness = 1.0 mm, and scanning time = 4′30; (ii) TOF-MRA: TR = 25 ms, TE = 3.4 ms, FOV = 220 × 220 mm3, flip angle = 20°, and scanning time = 3′48″.

3.2. Image Processing and Radiomic Feature Extraction

ITK-SNAP software v.3.8.0 (the University of Pennsylvania and the University of Utah, open source, http://www.itksnap.org/pmwiki/pmwiki.php (accessed on 1 January 2022) was used for 3D manual segmentation. Region of interest (ROI) was drawn on all 3D T1WI, slice-by-slice, on the original sagittal images, reconstructed axial images, and reconstructed coronal images. The maximum range of ROI was drawn along the visible border to cover the entire volume of aneurysm. Reader 1, with 7 years of experience in neurovascular imaging, drew the ROI delineation twice within the first week, following the same procedure. In addition, Reader 2, with more than 11 years of experience, drew an ROI separately, to assess the consistency between observers by comparing with the results of Reader 1. To synchronize the acquired images, the mean value and standard deviation (SD) of the image intensity were calculated, and images were standardized by z-score method [18]. Radiomic features were extracted using the Pyradiomics v.2.2.0 in Python [19]. To further investigate the heterogeneity within ROI, wavelet filters were applied to the original images, transforming the original images into versions focused on information at different scales [19]. Details of these features are provided in the documentation for PyRadiomics (https://pyradiomics.readthedocs.io/en/2.2.0/index.html (accessed on 1 January 2022)).

3.3. Model Building

The intraclass correlation coefficients (ICC) of each feature were calculated; the feature with ICC < 0.75 was considered unreliable and discarded. Reader 1 accounted for the samples if good agreement (ICC > 0.75) was achieved. Spearman correlation analysis was performed to obtain the mutual correlation between features. Two features were considered highly correlated if the pairwise correlation coefficient reached 0.9, and the feature with the largest mean absolute correlation coefficient was redundant and eliminated.
To avoid the risk of overfitting, the ElasticNet [20] was used for model building. Due to the limited number of cases in the training set, too many features in the model may increase the probability of overfitting. The ratio of L1 and L2 penalties in ElasticNet were set to 1 and 0, respectively, to select the most relevant features, and reduce the feature number. The ElasticNet was trained using 5-fold cross-validation to select radiomic features and determine the corresponding weights, then a radiomic model was constructed as the weighted sum of the selected radiomic features. In addition, a clinico-radiological model was established using clinical features and MRI features. The features in the clinico-radiological model were determined by multivariate logistic regression with backward feature selection, and the Akaike information criterion was used as the quantitative indicator for feature selection. Finally, an integrated model, which incorporated the radiomic model and the clinico-radiological model, was built by multivariate logistic regression.

3.4. Model Evaluation and Statistical Analysis

Models were evaluated by the leave-one-out cross-validation (LOOCV) method in the training set [21]. The diagnostic performance of the models was quantified by the area under the receiver operating characteristic (ROC) curve (AUC). The accuracy (ACC), sensitivity (SEN), and specificity (SPE) of each model were calculated. The Mann–Whitney U test and chi-square test were used to compare the differences between DA and hemorrhagic SA, respectively. All statistical tests were two-sided; p < 0.05 was considered statistically significant.

3.5. Radiologists′ Diagnosis

Each diagnosis was made by three junior neuroradiologists, and confirmed by two more senior neuroradiologists with more than 10 years’ working experience, who were blind to clinical information and surgical outcomes. The diagnosis of an SA was based on the shape and the flowing void effect low signal in the aneurysm. Hemorrhagic SAs contain hematoma, thrombus, or intra-aneurysmal flow artifacts, which show high signal or mixed signal on T1WI. The diagnosis of a DA requires some specific MRI features, such as IMH, double lumen, or the intimal flap. When the diagnosis diverged, the opinion of two senior neuroradiologists was decisive.
Model building, evaluation, and statistical analysis using R programming language (version 3.4.3, R Core Team, Vienna, Austria).

4. Results

Of the 145 patients included in this study, 77 cases from our hospital constitute the training set, and 69 cases from another hospital constitute the external test set (Figure 1).
No significant differences are found in the age, symptoms, lesion size, INR, or sign resembling the intimal flap of MRI in the training set, and no differences are found in INR in the external test set (Table 1).
The formula of the clinico-radiological model:
Clinico − radiological   score = 2.737 + 19.737 × Sign . intimal   flap − 20.268 × HHT   − 0.901 × Long . diameter − 2.574 × location
The ROC curve of the clinico-radiological model is illustrated in Figure 2a.
The diagnostic performance of the clinico-radiological model is presented in Table 2.
In the radiomic model, eight radiomic features are included: one shape feature (elongation), which shows the relationship between the two largest principal components in the ROI shape, and its value ranges from 0 (line-like object) to 1 (circle-like object). The remaining seven features are wavelet features. These wavelet features quantify the inhomogeneity of aneurysm intensity from different aspects. The formula of the radiomic model:
Radiomic score = 13.490 + 0.628 × shape . Elongation + 1.980 × LLH . firstorder . Mean − 2.610 × LHL . gldm . DependenceEntropy − 0.012 × LHH . firstorder . Skewness + 0.002 × HLL . gldm . LargeDependenceEmphasis − 116.438 × HHL . firstorder . Mean − 8.674 × HHH . glcm . Imc 2 + 0.552 × LLL . glcm . ClusterShade
The radiomic score is transformed to the probability of DA or hemorrhagic SA by the sigmoid function in Equation (2). The extracted radiomic features are used as input, and the radiomic model yields the probability for each patient. The violin plots of the features in the ElasticNet-based radiomic model are illustrated, to show how different features are distributed in patients (Figure 3).
Figure 4 shows an example of representative MRI images.
In the training set, the integrated model improves the diagnostic performance in terms of AUC, ACC, and SPE (Table 2), which is superior to both the radiomic model and the clinico-radiological model (Figure 2a). When applying these models to the external validation set, the radiomic model performs the best, even better than the experienced radiologists (Figure 2b).

5. Discussion

HR VW-MRI provides useful information to distinguish different types of aneurysms [14,22,23], and this is the first time using the machine-learning method of multicenter external verification to distinguish DA from hemorrhagic SA [24]. Unlike several previous studies, that only draw ROI from a single image slice [25], the ROIs in this study are drawn in all slices with a thickness of only 1.0 mm, which contains sufficient data information. ElasticNet, as a compression estimation method for variable selection, is suitable for the selection of biomarkers in high-dimensional data [20,26,27]. In addition to surpassing the method of selecting predictors based on the strength of their univariable association with the outcome, it also enables the panel of selected features to be combined into a radiomic signature [20,28,29].
Eight potential predictors were selected from 851 candidate radiomic features for radiomic model construction. For the shape feature (elongation), our result is consistent with some studies in which the proportion of non-saccular morphology is higher in posterior circulation aneurysms than in anterior circulation aneurysms [5,30]. Shape-based features suggest a high correlation between aneurysm type and morphological properties. This phenomenon can perhaps be explained by vascular morphology and hemodynamics. The arteries in the posterior circulation are relatively straight, and the blood flow enters the middle layer through the intimal tear hole to form a dissection [31], making the long diameter of DA larger than the short diameter. The remaining seven features are wavelet features which quantify the heterogeneity of aneurysm MRI intensity from different aspects, indicating that wavelet features are more discriminative [32,33], and provide more supplementary information to the shape-based features [34]. Some of the wavelet features are mentioned in previous studies, which are mainly related to the heterogeneity of tumors, and used for tumor grading and prognosis evaluation [35,36,37,38,39]. We infer that they may have varying degrees of correlation with the mixed signals of hemorrhage or the intimal flap. The extraction and calculation of radiological features are based on mathematical formulas of machine vision, and it is difficult to explain the relationship between higher-order features and pathological manifestations [40].
The AUC of both the clinico-radiological model and integrated model in the training set are much higher than the external test set, which is most likely a hint of overfitting. Overfitting is the phenomenon of matching a particular dataset too closely or precisely to fit other data well, which shows that the generalization of these two models is limited. The diagnostic performance of the radiomic model is similar in both sets, which demonstrates that it has good repeatability and generalization. The radiomic model performs better than the other two models in the test set, which proves that it has good diagnostic efficiency and application value. In addition, the radiomic model exceeds that of experienced radiologists in this study. This means that the application of machine-learning in clinical work improves the accuracy of differential diagnosis, reduces the probability of misdiagnosis, reduces the workload of radiologists, and helps surgeons to choose the most suitable surgical plan. We believe that the problem of insufficient generalization ability of radiomics models based on MRI can be solved by further algorithm optimization of image data and processing.
There are still several limitations. First, the sample size was not large enough; however, we used LOOCV in the training set, and there was an external test set to evaluate the performance of the models. Second, the time-consuming and labor-intensive process of manual segmentation may cause bias; automatic segmentation should be applied in future radiomics studies. Third, only a 3D T1WI sequence was used, while models based on multimodal MRI may perform better. Fourth, in future research, we will unify the image acquisition protocols in multicenters in prospective studies to ensure the homogeneity of the image.
In conclusion, via multicenter external validation, the presented radiomic model is an effective tool for preoperative differentiation of DA from a hemorrhagic SA, with good accuracy and generalizability, which can provide supplementary diagnoses to radiologists, and assist the surgeon to make the most suitable operation plan.

Author Contributions

Conceptualization, X.C.; methodology, Y.Z.; software, J.W..; validation, J.W. and Y.C.; data curation, X.C.; writing—original draft preparation, X.C.; writing—review and editing, Y.W.; supervision, W.X.; funding acquisition, W.X. All authors have read and agreed to the published version of the manuscript.

Funding

This work was supported by research grants from Shanghai Sailing Project (grant numbers 22YF1405000), Clinical Research Plan of SHDC (grant numbers SHDC2020CR3020A), Research Startup Fund of Huashan Hospital Fudan University (grant numbers 2021QD035), and Greater Bay Area Institute of Precision Medicine (Guangzhou) (grant numbers KCH2310094).

Institutional Review Board Statement

The study was conducted in accordance with the Declaration of Helsinki, and approved by the Institutional Review Board of Huashan Hospital and First Affiliated Hospital of Shandong First Medical University (KY2021-965, 15 December 2021).

Informed Consent Statement

The need for informed consent was waived by the ethical committees of both hospitals.

Data Availability Statement

The data presented in this study are available on request from the corresponding author. The data are not publicly available due to protecting patient privacy.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Tian, Z.; Wang, Z.; Li, W.; Zhu, W.; Liu, J.; Zhang, Y.; Yang, X.; Zhang, Y. Dynamic contrast-enhanced MRI analysis for prognosis of intracranial dissecting aneurysm with intramural haematoma after endovascular treatment: An observational registry study. Stroke Vasc. Neurol. 2020, 6, 133–138. [Google Scholar] [CrossRef] [PubMed]
  2. Beletsky, V.; Norris, J.W. Spontaneous dissection of the carotid and vertebral arteries. N. Engl. J. Med. 2001, 345, 467. [Google Scholar] [PubMed]
  3. Kurata, A.; Ohmomo, T.; Miyasaka, Y.; Fujii, K.; Kan, S.; Kitahara, T. Coil embolization for the treatment of ruptured dissecting vertebral aneurysms. AJNR Am. J. Neuroradiol. 2001, 22, 11–18. [Google Scholar]
  4. Su, W.; Gou, S.; Ni, S.; Li, G.; Liu, Y.; Zhu, S.; Li, X. Management of ruptured and unruptured intracranial vertebral artery dissecting aneurysms. J. Clin. Neurosci. 2011, 18, 1639–1644. [Google Scholar] [CrossRef] [PubMed]
  5. Bender, M.T.; Colby, G.P.; Jiang, B.; Lin, L.M.; Campos, J.K.; Xu, R.; Westbroek, E.M.; Vo, C.D.; Zarrin, D.A.; Caplan, J.M.; et al. Flow diversion of posterior circulation cerebral aneurysms: A single-institution series of 59 cases. Neurosurgery 2019, 84, 206–216. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  6. International Study of Unruptured Intracranial Aneurysms Investigators. Unruptured intracranial aneurysms—Risk of rupture and risks of surgical intervention. N. Engl. J. Med. 1998, 339, 1725–1733. [Google Scholar] [CrossRef]
  7. Foster, M.T.; Herwadkar, A.; Patel, H.C. Posterior inferior cerebellar artery/vertebral artery subarachnoid hemorrhage: A comparison of saccular vs dissecting aneurysms. Neurosurgery 2018, 82, 93–98. [Google Scholar] [CrossRef]
  8. Alfotih, G.T.; Li, F.; Xu, X.; Zhang, S. Risk factors for re-bleeding of aneurysmal subarachnoid hemorrhage: Meta-analysis of observational studies. Neurol. Neurochir. Pol. 2014, 48, 346–355. [Google Scholar] [CrossRef]
  9. Zhao, X.; Wang, H.; Liu, J.; Zhang, Z.; Li, Z. Endovascular treatment of vertebral artery dissecting aneurysm: A single-center experience. Exp. Ther. Med. 2019, 18, 4838–4844. [Google Scholar] [CrossRef] [Green Version]
  10. Bachmann, R.; Nassenstein, I.; Kooijman, H.; Dittrich, R.; Stehling, C.; Kugel, H.; Niederstadt, T.; Kuhlenbaumer, G.; Ringelstein, E.B.; Kramer, S.; et al. High-resolution magnetic resonance imaging (MRI) at 3.0 Tesla in the short-term follow-up of patients with proven cervical artery dissection. Investig. Radiol. 2007, 42, 460–466. [Google Scholar] [CrossRef]
  11. Oppenheim, C.; Naggara, O.; Touze, E.; Lacour, J.C.; Schmitt, E.; Bonneville, F.; Crozier, S.; Guegan-Massardier, E.; Gerardin, E.; Leclerc, X.; et al. High-resolution MR imaging of the cervical arterial wall: What the radiologist needs to know. Radiographics 2009, 29, 1413–1431. [Google Scholar] [CrossRef] [PubMed]
  12. Horie, N.; Morikawa, M.; Fukuda, S.; Hayashi, K.; Suyama, K.; Nagata, I. Detection of blood blister-like aneurysm and intramural hematoma with high-resolution magnetic resonance imaging. J. Neurosurg. 2011, 115, 1206–1209. [Google Scholar] [CrossRef] [PubMed]
  13. Wang, Y.; Lou, X.; Li, Y.; Sui, B.; Sun, S.; Li, C.; Jiang, P.; Siddiqui, A.; Yang, X. Imaging investigation of intracranial arterial dissecting aneurysms by using 3 T high-resolution MRI and DSA: From the interventional neuroradiologists’ view. Acta Neurochir. 2014, 156, 515–525. [Google Scholar] [CrossRef] [PubMed]
  14. Tian, Z.; Chen, J.; Zhang, Y.; Liu, J.; Wang, Y.; Sui, B.; Yang, X. Quantitative analysis of intracranial vertebrobasilar dissecting aneurysm with intramural hematoma after endovascular treatment using 3-T high-resolution magnetic resonance imaging. World Neurosurg. 2017, 108, 236–243. [Google Scholar] [CrossRef]
  15. Podgorsak, A.R.; Rava, R.A.; Shiraz Bhurwani, M.M.; Chandra, A.R.; Davies, J.M.; Siddiqui, A.H.; Ionita, C.N. Automatic radiomic feature extraction using deep learning for angiographic parametric imaging of intracranial aneurysms. J. Neurointerv. Surg. 2020, 12, 417–421. [Google Scholar] [CrossRef]
  16. Nakao, T.; Hanaoka, S.; Nomura, Y.; Sato, I.; Nemoto, M.; Miki, S.; Maeda, E.; Yoshikawa, T.; Hayashi, N.; Abe, O. Deep neural network-based computer-assisted detection of cerebral aneurysms in MR angiography. J. Magn. Reson. Imaging 2018, 47, 948–953. [Google Scholar] [CrossRef]
  17. Liu, J.; Chen, Y.; Lan, L.; Lin, B.; Chen, W.; Wang, M.; Li, R.; Yang, Y.; Zhao, B.; Hu, Z.; et al. Prediction of rupture risk in anterior communicating artery aneurysms with a feed-forward artificial neural network. Eur. Radiol. 2018, 28, 3268–3275. [Google Scholar] [CrossRef]
  18. Weninger, L.; Haarburger, C.; Merhof, D. Robustness of radiomics for survival prediction of brain tumor patients depending on resection status. Front. Comput. Neurosci. 2019, 13, 73. [Google Scholar] [CrossRef]
  19. van Griethuysen, J.J.M.; Fedorov, A.; Parmar, C.; Hosny, A.; Aucoin, N.; Narayan, V.; Beets-Tan, R.G.H.; Fillion-Robin, J.C.; Pieper, S.; Aerts, H. Computational radiomics system to decode the radiographic phenotype. Cancer Res. 2017, 77, e104–e107. [Google Scholar] [CrossRef] [Green Version]
  20. Sun, R.; Limkin, E.J.; Vakalopoulou, M.; Dercle, L.; Champiat, S.; Han, S.R.; Verlingue, L.; Brandao, D.; Lancia, A.; Ammari, S.; et al. A radiomics approach to assess tumour-infiltrating CD8 cells and response to anti-PD-1 or anti-PD-L1 immunotherapy: An imaging biomarker, retrospective multicohort study. Lancet Oncol. 2018, 19, 1180–1191. [Google Scholar] [CrossRef]
  21. Cawley, G.C.; Talbot, N.L.C. On over-fitting in model selection and subsequent selection bias in performance evaluation. J. Mach. Learn. Res. 2010, 11, 2079–2107. [Google Scholar]
  22. Zhang, M.; Ye, G.; Liu, Y.; Wang, Q.; Li, S.; Wang, Y. Clinical application of high-resolution MRI in combination with digital subtraction angiography in the diagnosis of vertebrobasilar artery dissecting aneurysm: An observational study (STROBE compliant). Medicine 2019, 98, e14857. [Google Scholar] [CrossRef] [PubMed]
  23. Kim, S.; Chung, J.; Cha, J.; Kim, B.M.; Kim, D.J.; Kim, Y.B.; Lee, J.W.; Huh, S.K.; Park, K.Y. Usefulness of high-resolution three-dimensional proton density-weighted turbo spin-echo MRI in distinguishing a junctional dilatation from an intracranial aneurysm of the posterior communicating artery: A pilot study. J. NeuroInterv. Surg. 2019, 12, 315–319. [Google Scholar] [CrossRef] [PubMed]
  24. Cao, X.; Xia, W.; Tang, Y.; Zhang, B.; Yang, J.M.; Zeng, Y.W.; Geng, D.Y.; Zhang, J. Radiomic model for distinguishing dissecting aneurysms from complicated saccular aneurysms on high-resolution magnetic resonance imaging. J. Stroke Cerebrovasc. Dis. 2020, 29, 105268. [Google Scholar] [CrossRef] [PubMed]
  25. Yao, X.; Liao, L.; Han, Y.; Wei, T.; Wu, H.; Wang, Y.; Li, Y.; Zhang, X.; Ren, K. Computerized tomography radiomics features analysis for evaluation of perihematomal edema in basal ganglia hemorrhage. J. Craniofacial Surg. 2019, 30, e768–e771. [Google Scholar] [CrossRef] [PubMed]
  26. Gui, J.; Li, H. Penalized Cox regression analysis in the high-dimensional and low-sample size settings, with applications to microarray gene expression data. Bioinformatics 2005, 21, 3001–3008. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  27. Hui Zou, T.H. Regularization and variable selection via the elastic net. J. R. Stat. Soc. 2005, 67, 768. [Google Scholar] [CrossRef]
  28. Aerts, H.J.; Velazquez, E.R.; Leijenaar, R.T.; Parmar, C.; Grossmann, P.; Carvalho, S.; Bussink, J.; Monshouwer, R.; Haibe-Kains, B.; Rietveld, D.; et al. Decoding tumour phenotype by noninvasive imaging using a quantitative radiomics approach. Nat. Commun. 2014, 5, 4006. [Google Scholar] [CrossRef]
  29. Andreassen, C.N.; Schack, L.M.; Laursen, L.V.; Alsner, J. Radiogenomics—Current status, challenges and future directions. Cancer Lett. 2016, 382, 127–136. [Google Scholar] [CrossRef]
  30. Passero, S.G.; Rossi, S. Natural history of vertebrobasilar dolichoectasia. Neurology 2008, 70, 66–72. [Google Scholar] [CrossRef]
  31. Braunstein, H. Pathogenesis of dissecting aneurysm. Circulation 1963, 28, 1071–1080. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  32. Wang, X.; Zhao, X.; Li, Q.; Xia, W.; Peng, Z.; Zhang, R.; Li, Q.; Jian, J.; Wang, W.; Tang, Y.; et al. Can peritumoral radiomics increase the efficiency of the prediction for lymph node metastasis in clinical stage T1 lung adenocarcinoma on CT? Eur. Radiol. 2019, 29, 6049–6058. [Google Scholar] [CrossRef] [PubMed]
  33. Meng, X.; Xia, W.; Xie, P.; Zhang, R.; Li, W.; Wang, M.; Xiong, F.; Liu, Y.; Fan, X.; Xie, Y.; et al. Preoperative radiomic signature based on multiparametric magnetic resonance imaging for noninvasive evaluation of biological characteristics in rectal cancer. Eur. Radiol. 2019, 29, 3200–3209. [Google Scholar] [CrossRef] [PubMed]
  34. Martin-Carreras, T.; Li, H.; Cooper, K.; Fan, Y.; Sebro, R. Radiomic features from MRI distinguish myxomas from myxofibrosarcomas. BMC Med. Imaging 2019, 19, 67. [Google Scholar] [CrossRef]
  35. Yang, M.; Ren, Y.; She, Y.; Xie, D.; Sun, X.; Shi, J.; Zhao, G.; Chen, C. Imaging phenotype using radiomics to predict dry pleural dissemination in non-small cell lung cancer. Ann. Transl. Med. 2019, 7, 259. [Google Scholar] [CrossRef]
  36. Liao, X.; Cai, B.; Tian, B.; Luo, Y.; Song, W.; Li, Y. Machine-learning based radiogenomics analysis of MRI features and metagenes in glioblastoma multiforme patients with different survival time. J. Cell. Mol. Med. 2019, 23, 4375–4385. [Google Scholar] [CrossRef]
  37. Wang, H.; Hu, D.; Yao, H.; Chen, M.; Li, S.; Chen, H.; Luo, J.; Feng, Y.; Guo, Y. Radiomics analysis of multiparametric MRI for the preoperative evaluation of pathological grade in bladder cancer tumors. Eur. Radiol. 2019, 29, 6182–6190. [Google Scholar] [CrossRef]
  38. He, B.; Ji, T.; Zhang, H.; Zhu, Y.; Shu, R.; Zhao, W.; Wang, K. MRI-based radiomics signature for tumor grading of rectal carcinoma using random forest model. J. Cell. Physiol. 2019, 234, 20501–20509. [Google Scholar] [CrossRef]
  39. Yu, F.H.; Wang, J.X.; Ye, X.H.; Deng, J.; Hang, J.; Yang, B. Ultrasound-based radiomics nomogram: A potential biomarker to predict axillary lymph node metastasis in early-stage invasive breast cancer. Eur. J. Radiol. 2019, 119, 108658. [Google Scholar] [CrossRef]
  40. Tran, B.; Dancey, J.E.; Kamel-Reid, S.; McPherson, J.D.; Bedard, P.L.; Brown, A.M.; Zhang, T.; Shaw, P.; Onetto, N.; Stein, L.; et al. Cancer genomics: Technology, discovery, and translation. J. Clin. Oncol. 2012, 30, 647–660. [Google Scholar] [CrossRef]
Jcm 11 03623 g001 550
Figure 1. Flow chart of patient recruitment, inclusion, and exclusion criteria for the dataset.
Figure 1. Flow chart of patient recruitment, inclusion, and exclusion criteria for the dataset.
Jcm 11 03623 g001
Jcm 11 03623 g002 550
Figure 2. The ROC curve of models and radiologists. (a) training set, (b) external test set.
Figure 2. The ROC curve of models and radiologists. (a) training set, (b) external test set.
Jcm 11 03623 g002
Jcm 11 03623 g003 550
Figure 3. The violin plots of eight features (a–h) in the radiomic model.
Figure 3. The violin plots of eight features (a–h) in the radiomic model.
Jcm 11 03623 g003
Jcm 11 03623 g004 550
Figure 4. A 64 year old female who had headache and nausea for two weeks. The radiologist diagnosed a DA in the V4 segment of her left vertebral artery. There are double lumen and sign resembling the intimal flap in the aneurysm on 3D T1WI (a). TOF-MRA shows inhomogeneous signal within the aneurysm (b). DSA finds no bleeding site or intimal tear hole (c). Her 3D volumetric reconstruction of the ROI (d). The radiomic model diagnoses it as an SA, which is consistent with the surgical results.
Figure 4. A 64 year old female who had headache and nausea for two weeks. The radiologist diagnosed a DA in the V4 segment of her left vertebral artery. There are double lumen and sign resembling the intimal flap in the aneurysm on 3D T1WI (a). TOF-MRA shows inhomogeneous signal within the aneurysm (b). DSA finds no bleeding site or intimal tear hole (c). Her 3D volumetric reconstruction of the ROI (d). The radiomic model diagnoses it as an SA, which is consistent with the surgical results.
Jcm 11 03623 g004
Table
Table 1. Clinical characteristics and MRI features of patients enrolled.
Table 1. Clinical characteristics and MRI features of patients enrolled.
Training Set (n = 77) External Test Set (n = 69)
DA SA p DA SA p
No. of patients 43 34 28 41
Female (n, %) 13 (30.23%) 23 (67.65%) 0.001 9 (32.14%) 32 (78.05%) <0.001
Age (year) 49.79 ± 12.06 55.00 ± 13.55 0.094 54.29 ± 11.43 57.54 ± 14.91 0.035
Clinical symptoms (n, %) 37 (86.05%) 24 (70.59%) 0.097 11 (39.29%) 32 (78.05%) 0.001
MRI features (n, %)
Sign resembling the intimal flap 21 (48.84%) 23 (67.65%) 0.098 10 (35.71%) 26 (63.42%) 0.024
HHT 34 (79.07%) 12 (35.29%) <0.001 19 (67.86%) 14 (34.15%) 0.006
Size (cm)
Long diameter 1.91 ± 1.04 1.66 ± 0.92 0.543 1.47 ± 0.80 1.94 ± 0.95 0.004
Short diameter 1.15 ± 0.60 1.30 ± 0.73 0.203 1.22 ± 0.75 1.59 ± 0.78 0.005
Lesion location (n, %)
Anterior circulation 13 (30.23%) 26 (76.47) <0.001 5 (17.86%) 33 (80.49%) <0.001
ICA 11 (25.58%) 18 (52.94%) 0.014 1 (3.57%) 23 (56.10%) <0.001
MCA 2 (4.65%) 9 (26.47%) 0.007 4 (14.29%) 9 (21.95%) 0.424
Posterior circulation 30 (69.77%) 7 (20.59%) <0.001 23 (82.14%) 8 (19.51%) <0.001
BA 4 (9.30%) 2 (5.88%) 0.578 5 (14.86%) 1 (2.44%) 0.026
VA 24 (55.81%) 2 (5.88%) <0.001 18 (64.29%) 3 (7.32%) <0.001
PCA 2 (4.65%) 3 (8.82%) 0.461 0 4 (9.76%) 0.089
Coagulation examination
INR 0.97 ± 0.06 0.96 ± 0.10 0.372 0.96 ± 0.05 0.96 ± 0.13 0.214
MRI, magnetic resonance imaging; HHT, hemorrhage, hematoma, or thrombus; ICA, internal carotid artery; MCA, middle cerebral artery; BA, basilar artery; VA, vertebral artery; PCA, posterior cerebral artery; INR, international normalized ratio.
Table
Table 2. Diagnostic performance of models and radiologists.
Table 2. Diagnostic performance of models and radiologists.
Model or RadiologistsTraining SetExternal Test Set
AUCACCSENSPEAUCACCSENSPE
Clinico-radiological model0.8670.8310.8230.8370.7170.7530.7800.714
Radiomic model0.8530.8310.8820.7910.8310.8120.8780.714
Integrated model0.9770.9480.8821.0000.8130.7820.8290.714
Radiologists0.7870.7790.8520.7200.8010.7970.7800.821
AUC, area under the ROC curve; ACC, accuracy; SEN, sensitivity; SPE, specificity. The dissecting aneurysm is defined as negative and the saccular aneurysm is defined as positive.
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

Cao, X.; Zeng, Y.; Wang, J.; Cao, Y.; Wu, Y.; Xia, W. Differentiation of Cerebral Dissecting Aneurysm from Hemorrhagic Saccular Aneurysm by Machine-Learning Based on Vessel Wall MRI: A Multicenter Study. J. Clin. Med. 2022, 11, 3623. https://doi.org/10.3390/jcm11133623

AMA Style

Cao X, Zeng Y, Wang J, Cao Y, Wu Y, Xia W. Differentiation of Cerebral Dissecting Aneurysm from Hemorrhagic Saccular Aneurysm by Machine-Learning Based on Vessel Wall MRI: A Multicenter Study. Journal of Clinical Medicine. 2022; 11(13):3623. https://doi.org/10.3390/jcm11133623

Chicago/Turabian Style

Cao, Xin, Yanwei Zeng, Junying Wang, Yunxi Cao, Yifan Wu, and Wei Xia. 2022. "Differentiation of Cerebral Dissecting Aneurysm from Hemorrhagic Saccular Aneurysm by Machine-Learning Based on Vessel Wall MRI: A Multicenter Study" Journal of Clinical Medicine 11, no. 13: 3623. https://doi.org/10.3390/jcm11133623

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop