scholarly journals Contrast-Enhanced Black Blood MRI Sequence Is Superior to Conventional T1 Sequence in Automated Detection of Brain Metastases by Convolutional Neural Networks

Diagnostics ◽  
2021 ◽  
Vol 11 (6) ◽  
pp. 1016
Author(s):  
Jonathan Kottlors ◽  
Simon Geissen ◽  
Hannah Jendreizik ◽  
Nils Große Hokamp ◽  
Philipp Fervers ◽  
...  
Keyword(s):  
Neural Networks ◽  
Brain Metastases ◽  
Convolutional Neural Networks ◽  
Brain Lesion ◽  
Brain Lesions ◽  
Automated Detection ◽  
Lesion Detection ◽  
Black Blood ◽  
Mri Imaging ◽  
Contrast Enhanced

Background: in magnetic resonance imaging (MRI), automated detection of brain metastases with convolutional neural networks (CNN) represents an extraordinary challenge due to small lesions sometimes posing as brain vessels as well as other confounders. Literature reporting high false positive rates when using conventional contrast enhanced (CE) T1 sequences questions their usefulness in clinical routine. CE black blood (BB) sequences may overcome these limitations by suppressing contrast-enhanced structures, thus facilitating lesion detection. This study compared CNN performance in conventional CE T1 and BB sequences and tested for objective improvement of brain lesion detection. Methods: we included a subgroup of 127 consecutive patients, receiving both CE T1 and BB sequences, referred for MRI concerning metastatic spread to the brain. A pretrained CNN was retrained with a customized monolayer classifier using either T1 or BB scans of brain lesions. Results: CE T1 imaging-based training resulted in an internal validation accuracy of 85.5% vs. 92.3% in BB imaging (p < 0.01). In holdout validation analysis, T1 image-based prediction presented poor specificity and sensitivity with an AUC of 0.53 compared to 0.87 in BB-imaging-based prediction. Conclusions: detection of brain lesions with CNN, BB-MRI imaging represents a highly effective input type when compared to conventional CE T1-MRI imaging. Use of BB-MRI can overcome the current limitations for automated brain lesion detection and the objectively excellent performance of our CNN suggests routine usage of BB sequences for radiological analysis.

Fr613 AUTOMATED DETECTION OF BOWEL PREPARATION ADEQUACY WITH DEEP CONVOLUTIONAL NEURAL NETWORKS

2021 ◽  
Vol 160 (6) ◽  
pp. S-376
Author(s):  
Daniel J. Low ◽  
Zhuoqiao Hong ◽  
Anjishnu Mukherjee ◽  
Sechiv Jugnundan ◽  
Samir C. Grover
Keyword(s):  
Neural Networks ◽  
Convolutional Neural Networks ◽  
Bowel Preparation ◽  
Automated Detection ◽  
Deep Convolutional Neural Networks

scholarly journals Deep Learning for Detecting Building Defects Using Convolutional Neural Networks

Sensors ◽  
2019 ◽  
Vol 19 (16) ◽  
pp. 3556 ◽  
Author(s):  
Husein Perez ◽  
Joseph H. M. Tah ◽  
Amir Mosavi
Keyword(s):  
Neural Networks ◽  
Convolutional Neural Networks ◽  
Effective Means ◽  
Scale Up ◽  
Real Life ◽  
Health And Safety ◽  
Condition Assessment ◽  
Automated Detection ◽  
Proposed Model ◽  
Current Asset

Clients are increasingly looking for fast and effective means to quickly and frequently survey and communicate the condition of their buildings so that essential repairs and maintenance work can be done in a proactive and timely manner before it becomes too dangerous and expensive. Traditional methods for this type of work commonly comprise of engaging building surveyors to undertake a condition assessment which involves a lengthy site inspection to produce a systematic recording of the physical condition of the building elements, including cost estimates of immediate and projected long-term costs of renewal, repair and maintenance of the building. Current asset condition assessment procedures are extensively time consuming, laborious, and expensive and pose health and safety threats to surveyors, particularly at height and roof levels which are difficult to access. This paper aims at evaluating the application of convolutional neural networks (CNN) towards an automated detection and localisation of key building defects, e.g., mould, deterioration, and stain, from images. The proposed model is based on pre-trained CNN classifier of VGG-16 (later compaired with ResNet-50, and Inception models), with class activation mapping (CAM) for object localisation. The challenges and limitations of the model in real-life applications have been identified. The proposed model has proven to be robust and able to accurately detect and localise building defects. The approach is being developed with the potential to scale-up and further advance to support automated detection of defects and deterioration of buildings in real-time using mobile devices and drones.

IMpower150: Exploratory analysis of brain metastases development.

2020 ◽  
Vol 38 (15_suppl) ◽  
pp. 9587-9587 ◽  
Author(s):  
Federico Cappuzzo ◽  
Martin Reck ◽  
Mark A. Socinski ◽  
Tony S. K. Mok ◽  
Robert M. Jotte ◽  
...  
Keyword(s):  
Brain Metastases ◽  
Brain Lesion ◽  
Population Data ◽  
Brain Lesions ◽  
Exploratory Analysis ◽  
Phase Iii ◽  
Lesion Development ◽  
Brain Scans ◽  
Intention To Treat ◽  
Grade 3

9587 Background: In the global phase III IMpower150 study (NCT02366143), atezolizumab (atezo) + bevacizumab (bev) + chemo (carboplatin + paclitaxel [CP] (ABCP) showed significant improvements in PFS and OS vs BCP in patients with chemotherapy-naive metastatic NSCLC (Socinski et al. N Engl J Med 2018). Because bev has been shown to delay or prevent brain metastases progression in NSCLC (Fu et al. J Chemother 2016; Ilhan-Mutlu et al. Mol Can Ther 2016), exploratory analyses were conducted to assess the development of brain metastases in patients treated with ABCP, BCP and atezo + CP (ACP) in IMpower150. Methods: A total of 1202 patients (intention-to-treat [ITT] population) were randomized 1:1:1 to receive ABCP, ACP or BCP. Doses were given every 3 weeks: atezo 1200 mg, bev 15 mg/kg, carboplatin AUC 6 mg/mL/min and paclitaxel 200 mg/m2. Co-primary endpoints were investigator-assessed PFS and OS in ITT–wild-type (no EGFR or ALK alterations) patients. Exploratory analyses included the rate and time to development (TTD) of new brain metastases in the ITT population, regardless of the presence of baseline brain metastases, as well as safety. Brain scans were performed as clinically indicated, and analyses were based on investigator assessments. Results: With a minimum follow-up of 32.4 months in the ITT population (data cutoff: September 13, 2019), 100 patients had developed brain metastases, with the highest rate of new brain lesions seen in the ACP (11.9%) vs the ABCP (7.0%) and BCP (6.0%) arms (table). Median TTD was not reached in any arm; a trend toward delayed TTD was seen in the ABCP vs BCP arm (HR, 0.68 [95% CI: 0.39, 1.19]). Among patients with and without brain metastases, 17 (35.4%) and 155 (44.0%) in the ACP arm, 18 (64.3%) and 207 (56.7%) in the ABCP arm and 10 (41.7%) and 183 (49.5%) in the BCP arm had Grade 3-4 treatment-related adverse events, respectively. Conclusions: The ACP arm had the highest rate of new brain lesions, whereas the ABCP and BCP arms had similar, lower rates. Taken together with the trend toward delayed development of new brain lesions with ABCP, the data suggest that adding atezo to BCP may not reduce the rate of new brain lesion development but may delay the time to new lesion development. No new safety signals were observed in this exploratory analysis. Clinical trial information: NCT02366143 . [Table: see text]

scholarly journals Intraoperative Contrast-Enhanced Ultrasound for Brain Tumor Surgery

Neurosurgery ◽  
2014 ◽  
Vol 74 (5) ◽  
pp. 542-552 ◽  
Author(s):  
Francesco Prada ◽  
Alessandro Perin ◽  
Alberto Martegani ◽  
Luca Aiani ◽  
Luigi Solbiati ◽  
...  
Keyword(s):  
Brain Tumor ◽  
Real Time ◽  
Large Scale ◽  
Brain Lesion ◽  
Ultrasound Contrast Agent ◽  
Ultrasound Contrast Agents ◽  
Brain Lesions ◽  
Contrast Enhanced Ultrasound ◽  
Ultrasound Contrast ◽  
Contrast Enhanced

Abstract BACKGROUND: Contrast-enhanced ultrasound (CEUS) is a dynamic and continuous modality that offers a real-time, direct view of vascularization patterns and tissue resistance for many organs. Thanks to newer ultrasound contrast agents, CEUS has become a well-established, live-imaging technique in many contexts, but it has never been used extensively for brain imaging. The use of intraoperative CEUS (iCEUS) imaging in neurosurgery is limited. OBJECTIVE: To provide the first dynamic and continuous iCEUS evaluation of a variety of brain lesions. METHODS: We evaluated 71 patients undergoing iCEUS imaging in an off-label setting while being operated on for different brain lesions; iCEUS imaging was obtained before resecting each lesion, after intravenous injection of ultrasound contrast agent. A semiquantitative, offline interobserver analysis was performed to visualize each brain lesion and to characterize its perfusion features, correlated with histopathology. RESULTS: In all cases, the brain lesion was visualized intraoperatively with iCEUS. The afferent and efferent blood vessels were identified, allowing evaluation of the time and features of the arterial and venous phases and facilitating the surgical strategy. iCEUS also proved to be useful in highlighting the lesion compared with standard B-mode imaging and showing its perfusion patterns. No adverse effects were observed. CONCLUSION: Our study is the first large-scale implementation of iCEUS in neurosurgery as a dynamic and continuous real-time imaging tool for brain surgery and provides the first iCEUS characterization of different brain neoplasms. The ability of CEUS to highlight and characterize brain tumor will possibly provide the neurosurgeon with important information anytime during a surgical procedure.

scholarly journals Komparacija detektabilnosti mozdanih lezija u multipleks sklerozi pri pregledu magnetnom rezonancijom jacine magnetnog polja 1.0 i 3.0 Tesla

2007 ◽  
Vol 54 (3) ◽  
pp. 115-117 ◽  
Author(s):  
T.L. Stosic-Opincal ◽  
M. Gavrilov ◽  
S. Lavrnic ◽  
R. Milenkovic ◽  
V. Peric ◽  
...  
Keyword(s):  
Multiple Sclerosis ◽  
Spin Echo ◽  
Relative Sensitivity ◽  
Brain Lesions ◽  
Lesion Detection ◽  
Focal Lesion ◽  
Gadolinium Contrast ◽  
Contrast Enhanced ◽  
Fluid Attenuated Inversion Recovery ◽  
3.0 T

To estimate the relative sensitivity of MR examination for brain lesions in multiple sclerosis at 1.0 Tesla (T) and 3.0 T using identical acquisition conditions. 54 patients with multiple sclerosis were examined both at 1.0T (Siemens Impact Expert) and 3.0T (Philips Intera) using T1-weighted spin echo (T1W-SE) with and without gadolinium contrast injections, T2W SE and fluid attenuated inversion recovery (FLAIR) imaging. Images were examined independently by three experienced neuroradiologists using focal lesion counting. 3.0T scans compared with 1.0T scans demonstrate a 27.3%, increase in the number of detected contrast enhanced lesions and an 22.7% increase in the number of detected lesions on FLAIR MR tomograms. High field 3.0T MR imaging demonstrates better sensitivity in the detection of focal brain lesions in multiple sclerosis. This improvement is more apparent in contrast enhanced lesion detection and less noticeable in FLAIR detected lesions.

Sign in / Sign up

Export Citation Format

Share Document