Automated LVO detection and collateral scoring on CTA using a 3D self-configuring object detection network: a multi-center study.

The use of deep learning (DL) techniques for automated diagnosis of large vessel occlusion (LVO) and collateral scoring on computed tomography angiography (CTA) is gaining attention. In this study, a state-of-the-art self-configuring object detection network called nnDetection was used to detect LVO and assess collateralization on CTA scans using a multi-task 3D object detection approach. The model was trained on single-phase CTA scans of 2425 patients at five centers, and its performance was evaluated on an external test set of 345 patients from another center. Ground-truth labels for the presence of LVO and collateral scores were provided by three radiologists. The nnDetection model achieved a diagnostic accuracy of 98.26% (95% CI 96.25-99.36%) in identifying LVO, correctly classifying 339 out of 345 CTA scans in the external test set. The DL-based collateral scores had a kappa of 0.80, indicating good agreement with the consensus of the radiologists. These results demonstrate that the self-configuring 3D nnDetection model can accurately detect LVO on single-phase CTA scans and provide semi-quantitative collateral scores, offering a comprehensive approach for automated stroke diagnostics in patients with LVO.

A Comparative Study of Multiparametric MRI Sequences in Measuring Prostate Cancer Index Lesion Volume.

Objectives: To compare the effectiveness of individual multiparametric prostate MRI (mpMRI) sequences-T2W, diffusion-weighted imaging (DWI) and apparent diffusion coefficient (ADC), and dynamic contrast-enhanced (DCE)-in assessing prostate cancer (PCa) index lesion volume using whole-mount pathology as the ground-truth; to assess the impact of an endorectal coil (ERC) on the measurements. Materials and Methods: We retrospectively enrolled 72 PCa patients who underwent 3T mpMRI with (n = 39) or without (n = 33) an ERC. A pathologist drew the index lesion borders on whole-mount pathology using planimetry (whole-mountvol). A radiologist drew the borders of the index lesion on each mpMRI sequence-T2Wvol, DWIvol, ADCvol, and DCEvol. Additionally, we calculated the maximum index lesion volume for each patient (maxMRIvol). The correlation and differences between mpMRI and whole-mount pathology in measuring the index lesion volume and the impact of an ERC were investigated. Results: The median T2Wvol, DWIvol, ADCvol, DCEvol, and maxMRIvol were 0.68 cm3, 0.97 cm3, 0.98 cm3, 0.82 cm3, and 1.13 cm3. There were good positive correlations between whole-mountvol and mpMRI sequences. However, all mpMRI-derived volumes underestimated the median whole-mountvol volume of 1.97 cm3 (P ≤ 0.001), with T2Wvol having the largest volumetric underestimation while DWIvol and ADCvol having the smallest. The mean relative index lesion volume underestimations of maxMRIvol were 39.16% ± 32.58% and 7.65% ± 51.91% with and without an ERC (P = 0.002). Conclusion: T2Wvol, DWIvol, ADCvol, DCEvol, and maxMRIvol substantially underestimate PCa index lesion volume compared with whole-mount pathology, with T2Wvol having the largest volume underestimation. Additionally, using an ERC exacerbates the volume underestimation.

A joint convolutional-recurrent neural network with an attention mechanism for detecting intracranial hemorrhage on noncontrast head CT.

To investigate the performance of a joint convolutional neural networks-recurrent neural networks (CNN-RNN) using an attention mechanism in identifying and classifying intracranial hemorrhage (ICH) on a large multi-center dataset; to test its performance in a prospective independent sample consisting of consecutive real-world patients. All consecutive patients who underwent emergency non-contrast-enhanced head CT in five different centers were retrospectively gathered. Five neuroradiologists created the ground-truth labels. The development dataset was divided into the training and validation set. After the development phase, we integrated the deep learning model into an independent center's PACS environment for over six months for assessing the performance in a real clinical setting. Three radiologists created the ground-truth labels of the testing set with a majority voting. A total of 55,179 head CT scans of 48,070 patients, 28,253 men (58.77%), with a mean age of 53.84 ± 17.64 years (range 18-89) were enrolled in the study. The validation sample comprised 5211 head CT scans, with 991 being annotated as ICH-positive. The model's binary accuracy, sensitivity, and specificity on the validation set were 99.41%, 99.70%, and 98.91, respectively. During the prospective implementation, the model yielded an accuracy of 96.02% on 452 head CT scans with an average prediction time of 45 ± 8 s. The joint CNN-RNN model with an attention mechanism yielded excellent diagnostic accuracy in assessing ICH and its subtypes on a large-scale sample. The model was seamlessly integrated into the radiology workflow. Though slightly decreased performance, it provided decisions on the sample of consecutive real-world patients within a minute.

Endovascular management of anterior falcotentorial dural arteriovenous fistulas: importance of functionality of deep venous system and existence of accompanying choroidal arteriovenous malformation.

BACKGROUND: Anterior falcotentorial junction dural arteriovenous fistulas (AFDAVFs) are the most deeply located and most complex type of tentorial fistula due to their location and vascular anatomy. We aimed to make new angiographic definitions of AFDAVF nidus and functionality of the deep venous system of the brain and thereby provide a safer approach for endovascular treatment. METHODS: We retrospectively examined 18 patients with AFDAVF who received endovascular treatment at our neuroradiology department between June 2002 and May 2019. Pre- and post-treatment clinical assessments were performed using the modified Rankin Scale. AFDAVF niduses were defined as mixed-type or pure-dural-type on the basis of whether choroidal arteriovenous malformation was coexisting or not, respectively. The deep venous system was denoted as functional or nonfunctional. RESULTS: We included 13 men and 5 women (mean (range) age, 47.2 (31-62) years). We evaluated 15 patients with pure-dural-type AFDAVFs and three with mixed-type AFDAVFs. Complete occlusion of the fistula was achieved in 15/18 patients. Three patients had transient neurologic symptoms. In two patients these were due to mild thalamic ischemia and in the third patient was due to tectal venous ischemia, all in mixed-type AFDAVF. One patient also developed Parinaud syndrome due to compression of the tectal plate by a thrombosed large vein of Galen. No patients died or developed permanent morbidity. CONCLUSION: Evaluating AFDAVFs as described here using our new subtyping model will help improve analysis of the malformation and development of a safer endovascular strategy, and hence may prevent periprocedural complications and improve treatment safety.

Inter-vendor performance of deep learning in segmenting acute ischemic lesions on diffusion-weighted imaging: a multicenter study.

There is little evidence on the applicability of deep learning (DL) in the segmentation of acute ischemic lesions on diffusion-weighted imaging (DWI) between magnetic resonance imaging (MRI) scanners of different manufacturers. We retrospectively included DWI data of patients with acute ischemic lesions from six centers. Dataset A (n = 2986) and B (n = 3951) included data from Siemens and GE MRI scanners, respectively. The datasets were split into the training (80%), validation (10%), and internal test (10%) sets, and six neuroradiologists created ground-truth masks. Models A and B were the proposed neural networks trained on datasets A and B. The models subsequently fine-tuned across the datasets using their validation data. Another radiologist performed the segmentation on the test sets for comparisons. The median Dice scores of models A and B were 0.858 and 0.857 for the internal tests, which were non-inferior to the radiologist's performance, but demonstrated lower performance than the radiologist on the external tests. Fine-tuned models A and B achieved median Dice scores of 0.832 and 0.846, which were non-inferior to the radiologist's performance on the external tests. The present work shows that the inter-vendor operability of deep learning for the segmentation of ischemic lesions on DWI might be enhanced via transfer learning; thereby, their clinical applicability and generalizability could be improved.

Does Employing a Flowchart Improve the Diagnostic Performance of Cardiac Magnetic Resonance Imaging in Left Ventricular Noncompaction?

BACKGROUND: To test the hypothesis that making a diagnosis of left ventricular noncompaction (LVNC) on cardiac magnetic resonance imaging (CMRI) using a noncompacted-to-compacted (NC/C) myocardium ratio > 2.3 would yield significant errors, and also to test a diagnostic flowchart in patients who undergo CMRI and have clinical and echocardiographic findings suggesting LVNC could improve the diagnosis of LVNC. METHODS: A total of 84 patients with LVNC and 162 controls consisting of patients with other diseases and healthy participants who had CMRI and echocardiograms were selected. The diagnostic flowchart of the study involved the use of CMRI with all available sequences for patients with a high pre-test probability of LVNC. Two blinded independent cardiologists evaluated echocardiograms, and patients with suggestive echocardiographic and clinical findings for LVNC were enrolled in the high pre-test probability of LVNC group. Two independent blinded radiologists established the diagnosis of LVNC based on NC/C ratio > 2.3 on CMRI, and they were allowed to re-assess the patients following the diagnostic flowchart. RESULTS: An NC/C ratio > 2.3 identified 83 of 84 LVNC patients, yet incorrectly classified 48 of the 162 controls as having LVNC. Radiologists changed their decision in 23 of 48 patients with incorrect diagnoses, resulted in improved specificity (70.4% to 84.6%). The use of the CMRI diagnostic flowchart in the high pre-test probability group yielded a high specificity (97.2%) and accuracy (95.9%). CONCLUSIONS: LVNC diagnosed by CMRI based on the NC/C criterion can lead to overdiagnosis, whereas only using CMRI in patients with a high pre-test probability of LVNC with all available sequences may improve the diagnostic performance.

Use of volumetric CT scanning to predict tumor staging and survival in pancreatic cancer patients that are to be administered curative resection.

BACKGROUND: This study evaluates the achievability of CT volumetry of pancreatic cancer and its correlation with pTNM stage and survival. METHODS: Tumor volume was measured from contrast enhanced CT images of 58 patients who undergo curative resection for pancreatic cancer using the Segment Editor module implemented in 3D-Slicer-a free open source software platform. Receiver operating characteristic (ROC) analysis was used to evaluate correlation between Tvol and pTNM staging. RESULTS: The preoperative images of 58 pancreatic adenocarcinoma patients were included. The mean Tvol of pancreatic cancer is an increasing trend with T stage (The mean T1vol = 1.75 cm3 , the mean T2vol = 11.43 cm3 , the mean T3vol = 14.98 cm3 , the mean T4vol = 19.6 cm3 ). There were statistical differences between volumes (p = .000). On ROC analysis, the area under the ROC curve (Az) of Tvol to differentiate T1 stage from ≥T2 stage was 0.966 (p = .000). At a cut-off value of 3.050 cm3 , sensitivity of 92.3%, and specificity of 83.3% were achieved. Az value of Tvol to differentiate ≤T2 from ≥T3 stage was 0.750 (p = .010). At a cut-off value of 10.250 cm3 , sensitivity of 72.7% and specificity of 66% were achieved. In addition Az value of Tvol to differentiate ≤T3 from ≥T4 stage was 0.652 and was not significant (p = .380). At a cut-off value of 11.2 cm3 , sensitivity of 66.7% and specificity of 63.6% were achieved. CONCLUSION: CT volumetry in pancreatic cancer is feasible with excellent reproducibility. It is one of the prognostic factors affecting survival in operated patients with pancreatic cancer.

Loss of skeletal muscle area and fat-free mass during dabrafenib/trametinib and vemurafenib/cobimetinib treatments in patients with BRAF-mutant metastatic malignant melanoma.

This study aimed to assess whether dabrafenib/trametinib and vemurafenib/cobimetinib treatments are associated with a change in skeletal muscle area (SMA) and total fat-free mass (FFM) assessed by computed tomography (CT), and to compare the efficacy and safety profile of these treatments in patients with metastatic melanoma. Thirty-one patients treated with B-Raf proto-oncogene, serine/threonine kinase/MAPK extracellular receptor kinase inhibitors were included between 2016 and 2019. Eighteen patients received dabrafenib/trametinib and remaining patients received vemurafenib/cobimetinib. CT scans were performed at baseline and at 4-6 months of follow-up to measure cross-sectional areas of SMA. FFM and skeletal muscle index (SMI) values were calculated. Of the patients, including 18 treated with dabrafenib/trametinib (58.1%) and 13 with vemurafenib/cobimetinib (41.9%); 58.1% were male, 41.9% were female and median age was 52 years. A significant decrease in SMA was observed after dabrafenib/trametinib and vemurafenib/cobimetinib treatments (P = 0.003 and P = 0.002, respectively). A significant decrease in FFM values was observed after dabrafenib/trametinib and vemurafenib/cobimetinib treatments (P = 0.003 and P = 0.002, respectively). Dose-limiting toxicity (DLT) was observed in 35.9% of the patients with sarcopenia. No significant difference was seen between the dabrafenib/trametinib and vemurafenib/cobimetinib groups in median progression-free survival (PFS) (11.9 vs. 7.3 months, respectively, P = 0.28) and in median overall survival (OS) (25.46 vs. 13.7 months, respectively, P = 0.41). Baseline sarcopenia was not significantly associated with PFS or OS (P = 0.172 and P = 0.326, respectively). We found a significant decrease in SMI values determined at 4-6 months compared to the values before treatment both in dabrafenib/trametinib and vemurafenib/cobimetinib groups. DLT was similar with both treatments. Baseline sarcopenia was not significantly associated with PFS or OS.

Machine learning-based quantitative texture analysis of conventional MRI combined with ADC maps for assessment of IDH1 mutation in high-grade gliomas.

PURPOSE: To assess the performance of texture analysis of conventional magnetic resonance imaging (MRI) and apparent diffusion coefficient (ADC) maps in predicting IDH1 status in high-grade gliomas (HGG). MATERIALS AND METHODS: A total of 142 patients with HGG were included in the study. IDH1 mutation was present in 48 of 142 HGG (33.8%). Patients were randomly divided into the training cohort (n = 96) and the validation cohort (n = 46). Texture features were extracted via regions of interest on axial T2WI FLAIR, post-contrast T1WI, and ADC maps covering the whole volume of the tumors. The training cohort was used to train the random forest classifier, and the diagnostic performance of the pre-trained model was tested on the validation cohort. RESULTS: The random forest model of conventional MRI sequences and ADC images achieved diagnostic accuracy of 82.2% and 80.4% in predicting IDH1 status in the validation cohorts, respectively. The combined model of T2WI FLAIR, post-contrast T1WI, and ADC images exhibited the highest diagnostic accuracy equating 86.94% in the validation cohort. CONCLUSION: Texture analysis of conventional MRI sequences enhanced by ML analysis can accurately predict the IDH1 status of HGG. Adding textural analysis of ADC maps to conventional MRI results in incremental diagnostic performance.

De Novo Development of a Cerebral Arteriovenous Malformation: Case Report and Review of the Literature.

BACKGROUND: Cerebral arteriovenous malformations (AVMs) have been commonly regarded as congenital. However, this suspected origin has been challenged by reports of de novo cerebral AVM. CASE DESCRIPTION: We have described a 25-year-old man without any known history of cerebrovascular disease, in whom cranial imaging demonstrated the de novo appearance of a pial AVM. Initial magnetic resonance imaging at 11 years of age had revealed the presence of a 2-cm parafalcine arachnoid cyst. Computed tomography and magnetic resonance imaging performed 14 years later showed a new cerebral AVM in the left frontal lobe, which was then confirmed angiographically. CONCLUSIONS: The findings from our case report and from 9 similar reports challenge the traditional theory that AVMs constitute congenital lesions.