A clinically applicable deep-learning model for detecting intracranial aneurysm in computed tomography angiography images.
Ontology highlight
ABSTRACT:
SUBMITTER: Shi Z
PROVIDER: S-EPMC7705757 | biostudies-literature | 2020 Nov
REPOSITORIES: biostudies-literature
Publications
A clinically applicable deep-learning model for detecting intracranial aneurysm in computed tomography angiography images.
Shi Zhao Z Miao Chongchang C Schoepf U Joseph UJ Savage Rock H RH Dargis Danielle M DM Pan Chengwei C Chai Xue X Li Xiu Li XL Xia Shuang S Zhang Xin X Gu Yan Y Zhang Yonggang Y Hu Bin B Xu Wenda W Zhou Changsheng C Luo Song S Wang Hao H Mao Li L Liang Kongming K Wen Lili L Zhou Longjiang L Yu Yizhou Y Lu Guang Ming GM Zhang Long Jiang LJ
Nature communications 20201130 1
Intracranial aneurysm is a common life-threatening disease. Computed tomography angiography is recommended as the standard diagnosis tool; yet, interpretation can be time-consuming and challenging. We present a specific deep-learning-based model trained on 1,177 digital subtraction angiography verified bone-removal computed tomography angiography cases. The model has good tolerance to image quality and is tested with different manufacturers. Simulated real-world studies are conducted in consecut ...[more]
Similar Datasets
Project description:The accuracy of computed tomography angiography (CTA) image interpretation depends on the radiologist. This study aims to develop a new method for automatically detecting intracranial aneurysms from CTA images using deep learning, based on a convolutional neural network (CNN) implemented on the DeepMedic platform. Ninety CTA scans of patients with intracranial aneurysms are collected and divided into two datasets: training (80 subjects) and test (10 subjects) datasets. Subsequently, a deep learning architecture with a three-dimensional (3D) CNN model is implemented on the DeepMedic platform for the automatic segmentation and detection of intracranial aneurysms from the CTA images. The samples in the training dataset are used to train the CNN model, and those in the test dataset are used to assess the performance of the established system. Sensitivity, positive predictive value (PPV), and false positives are evaluated. The overall sensitivity and PPV of this system for detecting intracranial aneurysms from CTA images are 92.3% and 100%, respectively, and the segmentation sensitivity is 92.3%. The performance of the system in the detection of intracranial aneurysms is closely related to their size. The detection sensitivity for small intracranial aneurysms (≤ 3 mm) is 66.7%, whereas the sensitivity of detection for large (> 10 mm) and medium-sized (3-10 mm) intracranial aneurysms is 100%. The deep learning architecture with a 3D CNN model on the DeepMedic platform can reliably segment and detect intracranial aneurysms from CTA images with high sensitivity.
| S-EPMC9984635 | biostudies-literature
Project description:BackgroundAs infundibular dilation (ID) is less likely to cause hemorrhage or other clinical sequelae than an intracranial aneurysm (IA) and treating infundibulum itself may put the patient at unnecessary risk for stroke, it is important to distinguish between the ID and IA. Given the limitations of conventional single-phase computed tomography angiography (sCTA) to show small branches of intracranial arteries, the application of multiphase computed tomography angiography (mCTA) for identification seems promising. Our main objective was to evaluate whether using mCTA derived from computed tomography perfusion (CTP) data can improve distinction between IA and ID.MethodsA total of 35 patients diagnosed with IA or ID of the posterior communicating artery at its junction with the internal carotid artery junction (ICA-PComA) by sCTA at the 8th Medical Center of Chinese PLA General Hospital between January 2019 and May 2022 were retrospectively selected. All patients underwent CTP. The simulated mCTA was reconstructed from 0.75 mm CTP data for assessment of vascular branches. All data were processed separately by 2 CTA post-processors; 2 observers diagnosed IA and ID by source and volume rendering (VR) images of sCTA and VR images of mCTA, and compared the diagnostic efficacy of source and VR images of sCTA with VR images of mCTA.ResultsThe quality of the reconstructed images was more consistent between the 2 post-processors mCTA (K=0.856) than sCTA (K=0.648). The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy of the source image for ID identification were 78.9%, 86.7%, 84.2%, 81.3%, and 80.0% for sCTA, 73.7%, 81.2%, 82.3%, 72.2%, and 77.2% for the VR image of sCTA, and 94.7%, 87.5%, 90.0%, 93.3%, and 91.4% for the VR image based on mCTA, respectively. The net reclassification index (NRI) of mCTA for VR and the source image of sCTA was 0.273 and 0.220, respectively. VR base on mCTA was on average better than VR and the source image of sCTA at differentiating ID from IA (P=0.005 and P=0.001, respectively).ConclusionsCompared to sCTA, mCTA is more helpful in improving the distinction of ID and IA.
| S-EPMC10498264 | biostudies-literature
Project description:Background and purpose: Anatomical labeling of the cerebral vasculature is a crucial topic in determining the morphological nature and characterizing the vital variations of vessels, yet precise labeling of the intracranial arteries is time-consuming and challenging, given anatomical structural variability and surging imaging data. We present a U-Net-based deep learning (DL) model to automatically label detailed anatomical segments in computed tomography angiography (CTA) for the first time. The trained DL algorithm was further tested on a clinically relevant set for the localization of intracranial aneurysms (IAs). Methods: 457 examinations with varying degrees of arterial stenosis were used to train, validate, and test the model, aiming to automatically label 42 segments of the intracranial arteries [e.g., 7 segments of the internal carotid artery (ICA)]. Evaluation metrics included Dice similarity coefficient (DSC), mean surface distance (MSD), and Hausdorff distance (HD). Additionally, 96 examinations containing at least one IA were enrolled to assess the model's potential in enhancing clinicians' precision in IA localization. A total of 5 clinicians with different experience levels participated as readers in the clinical experiment and identified the precise location of IA without and with algorithm assistance, where there was a washout period of 14 days between two interpretations. The diagnostic accuracy, time, and mean interrater agreement (Fleiss' Kappa) were calculated to assess the differences in clinical performance of clinicians. Results: The proposed model exhibited notable labeling performance on 42 segments that included 7 anatomical segments of ICA, with the mean DSC of 0.88, MSD of 0.82 mm and HD of 6.59 mm. Furthermore, the model demonstrated superior labeling performance in healthy subjects compared to patients with stenosis (DSC: 0.91 vs. 0.89, p < 0.05; HD: 4.75 vs. 6.19, p < 0.05). Concurrently, clinicians with model predictions achieved significant improvements when interpreting the precise location of IA. The clinicians' mean accuracy increased by 0.04 (p = 0.003), mean time to diagnosis reduced by 9.76 s (p < 0.001), and mean interrater agreement (Fleiss' Kappa) increased by 0.07 (p = 0.029). Conclusion: Our model stands proficient for labeling intracranial arteries using the largest CTA dataset. Crucially, it demonstrates clinical utility, helping prioritize the patients with high risks and ease clinical workload.
| S-EPMC10794642 | biostudies-literature
Project description:Study designRetrospective diagnostic study.ObjectiveTo automatically detect osteolytic bone metastasis lesions in the thoracolumbar region using conventional computed tomography (CT) scans, we developed a new deep learning (DL)-based computer-aided detection model.Summary of background dataRadiographic detection of bone metastasis is often difficult, even for orthopedic surgeons and diagnostic radiologists, with a consequent risk for pathologic fracture or spinal cord injury. If we can improve detection rates, we will be able to prevent the deterioration of patients' quality of life at the end stage of cancer.Materials and methodsThis study included CT scans acquired at Tokyo Medical and Dental University (TMDU) Hospital between 2016 and 2022. A total of 263 positive CT scans that included at least one osteolytic bone metastasis lesion in the thoracolumbar spine and 172 negative CT scans without bone metastasis were collected for the datasets to train and validate the DL algorithm. As a test data set, 20 positive and 20 negative CT scans were separately collected from the training and validation datasets. To evaluate the performance of the established artificial intelligence (AI) model, sensitivity, precision, F1-score, and specificity were calculated. The clinical utility of our AI model was also evaluated through observer studies involving six orthopaedic surgeons and six radiologists.ResultsOur AI model showed a sensitivity, precision, and F1-score of 0.78, 0.68, and 0.72 (per slice) and 0.75, 0.36, and 0.48 (per lesion), respectively. The observer studies revealed that our AI model had comparable sensitivity to orthopaedic or radiology experts and improved the sensitivity and F1-score of residents.ConclusionWe developed a novel DL-based AI model for detecting osteolytic bone metastases in the thoracolumbar spine. Although further improvement in accuracy is needed, the current AI model may be applied to current clinical practice.Level of evidenceLevel III.
| S-EPMC10898548 | biostudies-literature
Project description:Background and purposeMultiple attempts at intracranial hemorrhage (ICH) detection using deep-learning techniques have been plagued by clinical failures. We aimed to compare the performance of a deep-learning algorithm for ICH detection trained on strongly and weakly annotated datasets, and to assess whether a weighted ensemble model that integrates separate models trained using datasets with different ICH improves performance.MethodsWe used brain CT scans from the Radiological Society of North America (27,861 CT scans, 3,528 ICHs) and AI-Hub (53,045 CT scans, 7,013 ICHs) for training. DenseNet121, InceptionResNetV2, MobileNetV2, and VGG19 were trained on strongly and weakly annotated datasets and compared using independent external test datasets. We then developed a weighted ensemble model combining separate models trained on all ICH, subdural hemorrhage (SDH), subarachnoid hemorrhage (SAH), and small-lesion ICH cases. The final weighted ensemble model was compared to four well-known deep-learning models. After external testing, six neurologists reviewed 91 ICH cases difficult for AI and humans.ResultsInceptionResNetV2, MobileNetV2, and VGG19 models outperformed when trained on strongly annotated datasets. A weighted ensemble model combining models trained on SDH, SAH, and small-lesion ICH had a higher AUC, compared with a model trained on all ICH cases only. This model outperformed four deep-learning models (AUC [95% C.I.]: Ensemble model, 0.953[0.938-0.965]; InceptionResNetV2, 0.852[0.828-0.873]; DenseNet121, 0.875[0.852-0.895]; VGG19, 0.796[0.770-0.821]; MobileNetV2, 0.650[0.620-0.680]; p < 0.0001). In addition, the case review showed that a better understanding and management of difficult cases may facilitate clinical use of ICH detection algorithms.ConclusionWe propose a weighted ensemble model for ICH detection, trained on large-scale, strongly annotated CT scans, as no model can capture all aspects of complex tasks.
| S-EPMC10784380 | biostudies-literature
Project description:BackgroundHighly accurate detection of intracranial hemorrhages (ICH) on head computed tomography (HCT) scans can prove challenging at high-volume centers. This study aimed to determine the number of additional ICHs detected by an artificial intelligence (AI) algorithm and to evaluate reasons for erroneous results at a level I trauma center with teleradiology services.MethodsIn a retrospective multi-center cohort study, consecutive emergency non-contrast HCT scans were analyzed by a commercially available ICH detection software (AIDOC, Tel Aviv, Israel). Discrepancies between AI analysis and initial radiology report (RR) were reviewed by a blinded neuroradiologist to determine the number of additional ICHs detected and evaluate reasons leading to errors.Results4946 HCT (05/2020-09/2020) from 18 hospitals were included in the analysis. 205 reports (4.1%) were classified as hemorrhages by both radiology report and AI. Out of a total of 162 (3.3%) discrepant reports, 62 were confirmed as hemorrhages by the reference neuroradiologist. 33 ICHs were identified exclusively via RRs. The AI algorithm detected an additional 29 instances of ICH, missed 12.4% of ICH and overcalled 1.9%; RRs missed 10.9% of ICHs and overcalled 0.2%. Many of the ICHs missed by the AI algorithm were located in the subarachnoid space (42.4%) and under the calvaria (48.5%). 85% of ICHs missed by RRs occurred outside of regular working-hours. Calcifications (39.3%), beam-hardening artifacts (18%), tumors (15.7%), and blood vessels (7.9%) were the most common reasons for AI overcalls. ICH size, image quality, and primary examiner experience were not found to be significantly associated with likelihood of incorrect AI results.ConclusionComplementing human expertise with AI resulted in a 12.2% increase in ICH detection. The AI algorithm overcalled 1.9% HCT.Trial registrationGerman Clinical Trials Register (DRKS-ID: DRKS00023593).
| S-EPMC8629230 | biostudies-literature
Project description:BackgroundIt is crucial to distinguish unstable from stable intracranial aneurysms (IAs) as early as possible to derive optimal clinical decision-making for further treatment or follow-up. The aim of this study was to investigate the value of a deep learning model (DLM) in identifying unstable IAs from computed tomography angiography (CTA) images and to compare its discriminatory ability with that of a conventional logistic regression model (LRM).MethodsFrom August 2011 to May 2021, a total of 1,049 patients with 681 unstable IAs and 556 stable IAs were retrospectively analyzed. IAs were randomly divided into training (64%), internal validation (16%), and test sets (20%). Convolutional neural network (CNN) analysis and conventional logistic regression (LR) were used to predict which IAs were unstable. The area under the curve (AUC), sensitivity, specificity and accuracy were calculated to evaluate the discriminating ability of the models. One hundred and ninety-seven patients with 229 IAs from Banan Hospital were used for external validation sets.ResultsThe conventional LRM showed 11 unstable risk factors, including clinical and IA characteristics. The LRM had an AUC of 0.963 [95% confidence interval (CI): 0.941-0.986], a sensitivity, specificity and accuracy on the external validation set of 0.922, 0.906, and 0.913, respectively, in predicting unstable IAs. In predicting unstable IAs, the DLM had an AUC of 0.771 (95% CI: 0.582-0.960), a sensitivity, specificity and accuracy on the external validation set of 0.694, 0.929, and 0.782, respectively.ConclusionsThe CNN-based DLM applied to CTA images did not outperform the conventional LRM in predicting unstable IAs. The patient clinical and IA morphological parameters remain critical factors for ensuring IA stability. Further studies are needed to enhance the diagnostic accuracy.
| S-EPMC11007515 | biostudies-literature
Project description:PurposeTo generate the first open dataset of retinal parafoveal optical coherence tomography angiography (OCTA) images with associated ground truth manual segmentations, and to establish a standard for OCTA image segmentation by surveying a broad range of state-of-the-art vessel enhancement and binarization procedures.MethodsHandcrafted filters and neural network architectures were used to perform vessel enhancement. Thresholding methods and machine learning approaches were applied to obtain the final binarization. Evaluation was performed by using pixelwise metrics and newly proposed topological metrics. Finally, we compare the error in the computation of clinically relevant vascular network metrics (e.g., foveal avascular zone area and vessel density) across segmentation methods.ResultsOur results show that, for the set of images considered, deep learning architectures (U-Net and CS-Net) achieve the best performance (Dice = 0.89). For applications where manually segmented data are not available to retrain these approaches, our findings suggest that optimally oriented flux (OOF) is the best handcrafted filter (Dice = 0.86). Moreover, our results show up to 25% differences in vessel density accuracy depending on the segmentation method used.ConclusionsIn this study, we derive and validate the first open dataset of retinal parafoveal OCTA images with associated ground truth manual segmentations. Our findings should be taken into account when comparing the results of clinical studies and performing meta-analyses. Finally, we release our data and source code to support standardization efforts in OCTA image segmentation.Translational relevanceThis work establishes a standard for OCTA retinal image segmentation and introduces the importance of evaluating segmentation performance in terms of clinically relevant metrics.
| S-EPMC7718823 | biostudies-literature
Project description:The lack of publicly available datasets of computed-tomography angiography (CTA) images for pulmonary embolism (PE) is a problem felt by physicians and researchers. Although a number of computer-aided detection (CAD) systems have been developed for PE diagnosis, their performance is often evaluated using private datasets. In this paper, we introduce a new public dataset called FUMPE (standing for Ferdowsi University of Mashhad's PE dataset) which consists of three-dimensional PE-CTA images of 35 different subjects with 8792 slices in total. For each benchmark image, two expert radiologists provided the ground-truth with the assistance of a semi-automated image processing software tool. FUMPE is a challenging benchmark for CAD methods because of the large number (i.e., 3438) of PE regions and, more especially, because of the location of most of them (i.e., 67%) in lung peripheral arteries. Moreover, due to the reporting of the Qanadli score for each PE-CTA image, FUMPE is the first public dataset which can be used for the analysis of mortality and morbidity risks associated with PE. We also report some complementary prognosis information for each subject.
| S-EPMC6122162 | biostudies-literature
Project description:ObjectivesDeep-learning image reconstruction (DLIR) offers unique opportunities for reducing image noise without degrading image quality or diagnostic accuracy in coronary CT angiography (CCTA). The present study aimed at exploiting the capabilities of DLIR to reduce radiation dose and assess its impact on stenosis severity, plaque composition analysis, and plaque volume quantification.MethodsThis prospective study includes 50 patients who underwent two sequential CCTA scans at normal-dose (ND) and lower-dose (LD). ND scans were reconstructed with Adaptive Statistical Iterative Reconstruction-Veo (ASiR-V) 100%, and LD scans with DLIR. Image noise (in Hounsfield units, HU) and quantitative plaque volumes (in mm3) were assessed quantitatively. Stenosis severity was visually categorized into no stenosis (0%), stenosis (< 20%, 20-50%, 51-70%, 71-90%, 91-99%), and occlusion (100%). Plaque composition was classified as calcified, non-calcified, or mixed.ResultsReduction of radiation dose from ND scans with ASiR-V 100% to LD scans with DLIR at the highest level (DLIR-H; 1.4 mSv vs. 0.8 mSv, p < 0.001) had no impact on image noise (28 vs. 27 HU, p = 0.598). Reliability of stenosis severity and plaque composition was excellent between ND scans with ASiR-V 100% and LD scans with DLIR-H (intraclass correlation coefficients of 0.995 and 0.974, respectively). Comparison of plaque volumes using Bland-Altman analysis revealed a mean difference of - 0.8 mm3 (± 2.5 mm3) and limits of agreement between - 5.8 and + 4.1 mm3.ConclusionDLIR enables a reduction in radiation dose from CCTA by 43% without significant impact on image noise, stenosis severity, plaque composition, and quantitative plaque volume.Key points•Deep-learning image reconstruction (DLIR) enables radiation dose reduction by over 40% for coronary computed tomography angiography (CCTA). •Image noise remains unchanged between a normal-dose CCTA reconstructed by ASiR-V and a lower-dose CCTA reconstructed by DLIR. •There is no impact on the assessment of stenosis severity, plaque composition, and quantitative plaque volume between the two scans.
| S-EPMC8921160 | biostudies-literature