scholarly journals Intracranial Aneurysm Detection from 3D Vascular Mesh Models with Ensemble Deep Learning

2019 ◽  
pp. 243-252 ◽  
Author(s):  
Mingsong Zhou ◽  
Xingce Wang ◽  
Zhongke Wu ◽  
Jose M. Pozo ◽  
Alejandro F. Frangi
Keyword(s):  
Deep Learning ◽  
Intracranial Aneurysm ◽  
Aneurysm Detection ◽  
Mesh Models

Using convolutional neural network and long short time memory to automatically detect aneurysm on 2D DSA images (Preprint)

2021 ◽  
Author(s):  
JunHua Liao ◽  
LunXin Liu ◽  
HaiHan Duan ◽  
YunZhi Huang ◽  
LiangXue Zhou ◽  
...  
Keyword(s):  
Deep Learning ◽  
Intracranial Aneurysm ◽  
Spatial Information ◽  
Diagnostic System ◽  
Two Stage ◽  
Regional Localization ◽  
Detection Stage ◽  
Aneurysm Detection ◽  
Sensitivity Specificity ◽  
Computation Load

BACKGROUND It is hard to distinguish cerebral aneurysm from overlap vessels based on the 2D DSA images, for its lack the spatial information. OBJECTIVE The aim of this study is to construct a deep learning diagnostic system to improve the ability of detecting the PCoA aneurysm on 2D-DSA images and validate the efficiency of deep learning diagnostic system in 2D-DSA aneurysm detecting. METHODS We proposed a two stage detecting system. First, we established the regional localization stage (RLS) to automatically locate specific detection region of raw 2D-DSA sequences. And then, in the intracranial aneurysm detection stage (IADS) ,we build three different frames, RetinaNet, RetinaNet+LSTM, Bi-input+RetinaNet+LSTM, to detect the aneurysms. Each of the frame had fivefold cross-validation scheme. The area under curve (AUC), the receiver operating characteristic (ROC) curve, and mean average precision (mAP) were used to validate the efficiency of different frames. The sensitivity, specificity and accuracy were used to identify the ability of different frames. RESULTS 255 patients with PCoA aneurysms and 20 patients without aneurysm were included in this study. The best results of AUC of the RetinaNet, RetinaNet+LSTM, and Bi-input+RetinaNet+LSTM were 0.95, 0.96, and 0.97, respectively. The sensitivity of the RetinaNet, RetinaNet+LSTM, and Bi-input+RetinaNet+LSTM were 81.65% (59.40% to 94.76%), 87.91% (64.24% to 98.27%), 84.50% (69.57% to 93.97%), respectively. The specificity of the RetinaNet, RetinaNet+LSTM, and Bi-input+RetinaNet+LSTM were 88.89% (66.73% to 98.41%), 88.12% (66.06% to 98.08%), and 88.50% (74.44% to 96.39%), respectively. The accuracy of the RetinaNet, RetinaNet+LSTM, and Bi-input+RetinaNet+LSTM were 92.71% (71.29% to 99.54%), 89.42% (68.13% to 98.49%), and 91.00% (77.63% to 97.72%), respectively. CONCLUSIONS Two stage aneurysm detecting system can reduce time cost and the computation load. According to our results, more spatial and temporal information can help improve the performance of the frames, so that Bi-input+RetinaNet+LSTM has the best performance compared to other frames. And our study can demonstrate that our system was feasible to assist doctor to detect intracranial aneurysm on 2D-DSA images.

scholarly journals Clinically Applicable Deep Learning for Intracranial Aneurysm Detection in Computed Tomography Angiography Images: A Comprehensive Multicohort Study

2020 ◽  
Author(s):  
Zhao Shi ◽  
Chongchang Miao ◽  
Chengwei Pan ◽  
Xue Chai ◽  
Xiu Li Li ◽  
...  
Keyword(s):  
Deep Learning ◽  
Intracranial Aneurysm ◽  
High Confidence ◽  
Common Life ◽  
Bone Removal ◽  
Life Threatening ◽  
Diagnosis Tool ◽  
Aneurysm Detection ◽  
A Prospective Study ◽  
Human Burden

AbstractIntracranial aneurysm is a common life-threatening disease. CTA is recommended as a standard diagnosis tool, while the interpretation is time-consuming and challenging. We presented a novel deep-learning-based framework trained on 1,177 DSA verified bone-removal CTA cases. The framework had excellent tolerance to the influence of occult cases of CTA-negative but DSA-positive aneurysms, image quality, and manufacturers. Simulated real-world studies were conducted in consecutive internal and external cohorts, achieving improved sensitivity and negative predictive value than radiologists. A specific cohort of suspected acute ischemic stroke was employed and found 96.8% predicted-negative cases can be trusted with high confidence, leading to reducing in human burden. A prospective study is warranted to determine whether the algorithm could improve patients’ care in comparison to radiologists’ assessment.

scholarly journals A deep learning algorithm may automate intracranial aneurysm detection on MR angiography with high diagnostic performance

2020 ◽  
Vol 30 (11) ◽  
pp. 5785-5793 ◽  
Author(s):  
Bio Joo ◽  
Sung Soo Ahn ◽  
Pyeong Ho Yoon ◽  
Sohi Bae ◽  
Beomseok Sohn ◽  
...  
Keyword(s):  
Deep Learning ◽  
Intracranial Aneurysm ◽  
Mr Angiography ◽  
Diagnostic Performance ◽  
Learning Algorithm ◽  
Deep Learning Algorithm ◽  
Aneurysm Detection

Evaluation of diagnostic accuracy of intracranial aneurysm detection using medical-grade versus commercial consumer-grade displays and different image reconstructions against the background of process optimization for telemedicine

2019 ◽  
Vol 61 (7) ◽  
pp. 936-944
Author(s):  
Timo Alexander Auer ◽  
Hanna Münzfeld ◽  
Helena Posch ◽  
Juliane Stöckel ◽  
Anna Tietze ◽  
...  
Keyword(s):  
Image Quality ◽  
Intracranial Aneurysm ◽  
Diagnostic Accuracy ◽  
Process Optimization ◽  
Reading Speed ◽  
Diagnostic Confidence ◽  
Multiplanar Reconstructions ◽  
Aneurysm Detection ◽  
Maximum Intensity Projections ◽  
Medical Grade

Background Process optimization in computed tomography (CT) and telemedicine. Purpose To compare image quality and objective diagnostic accuracy of medical-grade and consumer-grade digital displays/computer terminals for detection of intracranial aneurysms. Material and Methods Four radiologists with different levels of experience retrospectively read a total of 60 patients including 30 cases of proven therapy-naïve intracranial aneurysm detectable on a medical-grade grayscale calibrated display. They had 5 min per case reading the first 20 datasets using only axial slices, the next 20 patients using axial slices and multiplanar reconstructions (MPRs), and the last 20 patients using axial slices, MPRs, and maximum intensity projections (MIPs). Three months after the first reading session on a medical-grade display, they read all datasets again under the same standardized conditions but on a consumer-grade display. Diagnostic performance, subjective diagnostic confidence, and reading speed were analyzed and compared. Readers rated image quality on a five-point Likert scale. Results Diagnostic accuracy did not differ significantly with areas under the curve of 0.717–0.809 for all readers on both display devices. Sensitivity and specificity did not increase significantly when adding MPRs and/or MIPs. Reading speed was similar with both devices. There were no significant differences in subjective image quality scores, and overall inter-reader variability of all subjective parameters correlated positively between the two devices ( P <0.001–0.011). Conclusion Diagnostic accuracy and readers’ diagnostic confidence in detecting and ruling out intracranial aneurysm were similar on commercial-grade and medical-grade displays. Additional reconstructions did not increase sensitivity/specificity or reduce the time needed for diagnosis.

scholarly journals Diagnostic fidelity of the Z-buffer segmentation algorithm: preliminary assessment based on intracranial aneurysm detection

2004 ◽  
Vol 37 (1) ◽  
pp. 19-29 ◽  
Author(s):  
Brian E Chapman ◽  
Dennis L Parker ◽  
Janet O Stapelton ◽  
Jay S Tsuruda ◽  
Claudia Mello-Thoms ◽  
...  
Keyword(s):  
Intracranial Aneurysm ◽  
Segmentation Algorithm ◽  
Preliminary Assessment ◽  
Aneurysm Detection

Intracranial aneurysm and coarctation of the aorta: prevalence in the current era

2020 ◽  
pp. 1-4
Author(s):  
Lauren Andrade ◽  
Arvind Hoskoppal ◽  
Mary Hunt Martin ◽  
Kevin Whitehead ◽  
Zhining Ou ◽  
...  
Keyword(s):  
Computed Tomography ◽  
Magnetic Resonance ◽  
Magnetic Resonance Angiography ◽  
Intracranial Aneurysm ◽  
Computed Tomography Angiography ◽  
Intracranial Aneurysms ◽  
Coarctation Of The Aorta ◽  
Former Smokers ◽  
Aneurysm Detection ◽  
Magnetic Resonance Imaging Mri

Abstract Background: A 10% prevalence of intracranial aneurysms in patients with coarctation of the aorta has been described in a few studies. Our objective is to describe the rate of intracranial aneurysm detection in patients with coarctation of the aorta in the current era. We hypothesise that, with earlier detection and coarctation of the aorta intervention, the rate of intracranial aneurysm is lower than previously reported and screening imaging may only be warranted in older patients or patients with certain risk factors. Methods: This is a retrospective study of 102 patients aged 13 years and older with coarctation who underwent brain computed tomography angiography, magnetic resonance imaging (MRI), or magnetic resonance angiography between January, 2000 and February, 2018. Results: The median age of coarctation repair was 4.4 months (2 days–47 years) and the initial repair was primarily surgical (90.2%). There were 11 former smokers, 4 current smokers, and 13 patients with ongoing hypertension. Imaging modalities included computed tomography angiography (13.7%), MRI (41.2%), and magnetic resonance angiography (46.1%), performed at a median age of 33.3 years, 22.4 years, and 25 years, respectively. There were 42 studies performed for screening, 48 studies performed for neurologic symptoms, and 12 studies performed for both screening and symptoms. There were no intracranial aneurysms detected in this study. Conclusions: These results suggest that the rate of intracranial aneurysms may be lower than previously reported and larger studies should explore the risk of intracranial aneurysms in coarctation of the aorta in the current era.

Sign in / Sign up

Export Citation Format

Share Document