The aim of this study was to investigate the feasibility ischemic stroke detection from computed tomography angiography source images (CTA-SI) using three-dimensional convolutional neural networks.CTA-SI 60 patients with a suspected acute middle cerebral artery were randomly selected for study; 30 used in network training, and subsequent testing performed remaining patients. training based on manually segmented lesions. Cerebral hemispheric comparison CTA non-contrast (NCCT) studied as...

Abstract Deep learning algorithms can be used to classify medical images. In distal radius fracture treatment, detection and radiographic assessment of displacement are critical steps. The aim this study was use pixel-level annotations fractures develop a deep model for precise detection. We randomly divided 3785 consecutive emergency wrist radiograph examinations from six hospitals training set (3399 examinations) test (386 examinations). the assess its validity. consensus three hand...

Summary Objective Navigated transcranial magnetic stimulation ( nTMS ) is becoming increasingly popular in noninvasive preoperative language mapping, as its results correlate well enough with those obtained by direct cortical DCS during awake surgery adult patients tumor. Reports the context of epilepsy or extraoperative adults are, however, sparse, and validation children lacking. Furthermore, little known about risk inducing epileptic seizures pediatric patients. We provide largest study...

Abstract Background Early diagnosis of the potentially fatal but curable chronic pulmonary embolism (CPE) is challenging. We have developed and investigated a novel convolutional neural network (CNN) model to recognise CPE from CT angiograms (CTPA) based on general vascular morphology in two-dimensional (2D) maximum intensity projection images. Methods A CNN was trained curated subset public dataset (RSPECT) with 755 CTPA studies, including patient-level labels CPE, acute (APE), or no...

Abstract Background Computed tomography angiography (CTA) imaging is needed in current guideline-based stroke diagnosis, and infarct core size one factor guiding treatment decisions. We studied the efficacy of a convolutional neural network (CNN) final volume prediction from CTA compared results to CT perfusion (CTP)-based commercially available software (RAPID, iSchemaView). Methods retrospectively selected 83 consecutive cases treated with thrombolytic therapy or receiving supportive care...

PurposeTo determine the effect of patient's vertical off-centering and scout direction on function automatic tube voltage selection (ATVS) current modulation (TCM) in chest computed tomography (CT).MethodsChest phantom was scanned with Siemens GE CT systems using three clinical protocols exploiting ATVS a fixed 120 kVp protocol. The scans were performed at five positions (−6 to +6 cm from scanner isocenter). effects (posterior-to-anterior, anterior-to-posterior, lateral) TCM studied by...

Abstract Objective We aimed to develop a vendor-neutral and interaction-free quality assurance protocol for measuring geometric accuracy of head brain magnetic resonance (MR) images. investigated the usability nonrigid image registration in analysis looked optimal parameters. Materials methods constructed 3D-printed phantom imaged it with 12 MR scanners using clinical sequences. registered geometric-ground-truth computed tomography (CT) acquisition images an open-source...

Computed tomography (CT) image noise is usually determined by standard deviation (SD) of pixel values from uniform regions. This study investigates how deep learning (DL) could be applied in head CT estimation.Two approaches were investigated for estimation a single acquisition image: direct using supervised DnCNN convolutional neural network (CNN) architecture, and subtraction denoised estimated with denoising UNet-CNN experimented unsupervised noise2noise training approaches. Noise was...

Artificial intelligence (AI) applications are becoming increasingly common in radiology. However, ensuring reliable operation and expected clinical benefits remains a challenge. A systematic testing process aims to facilitate deployment by confirming software applicability local patient populations, practises, adherence regulatory safety requirements, compatibility with existing systems. In this work, we present our developed based on practical experience. First, survey pre-evaluation is...

Developments in single photon emission tomography instrumentation and reconstruction methods present a potential for decreasing acquisition times. One of such recent options myocardial perfusion imaging (MPI) is IQ-SPECT. This study was motivated by the inconsistency reported ejection fraction (EF) left ventricular (LV) volume results between IQ-SPECT more conventional low-energy high-resolution (LEHR) collimation protocols. LEHR quantitative were compared while equivalent number iterations...

Abstract In stroke imaging, CT angiography (CTA) is used for detecting arterial occlusions. These images could also provide information on the extent of ischemia. The study aim was to develop and evaluate a convolutional neural network (CNN)–based algorithm segmenting acute ischemic lesions from CTA patients with suspected middle cerebral artery stroke. results were compared volumes reported by widely perfusion–based RAPID software (IschemaView). A 42-layer-deep CNN trained 50 manually...

To report initial experiences of automatic detection Crohn's disease (CD) using quantified motility in magnetic resonance enterography (MRE).From 302 patients, three datasets with roughly equal proportions CD and non-CD cases various illnesses were drawn for testing neural network training validation. All had unique MRE parameter configurations performed free breathing. Nine networks devised generation different regions interests (ROI): small bowel, all non-bowel. Additionally, a full-image...

The segmentation of 3D cell nuclei is essential in many tasks, such as targeted molecular radiotherapies (MRT) for metastatic tumours, toxicity screening, and the observation proliferating cells. In recent years, one popular method automatic has been deep learning enhanced marker-controlled watershed transform. this method, convolutional neural networks (CNNs) have used to create masks markers, algorithm instance segmentation. We studied whether could be improved densely cultivated via...

Chronic pulmonary embolism (CPE) is a life-threatening disease easily misdiagnosed on computed tomography. We investigated three-dimensional convolutional neural network (CNN) algorithm for detecting hypoperfusion in CPE from tomography angiography (CTPA).

Background Computed tomography perfusion (CTP) is the mainstay to determine possible eligibility for endovascular thrombectomy (EVT), but there still a need alternative methods in patient triage. Purpose To study ability of computed angiography (CTA)-based convolutional neural network (CNN) method predicting final infarct volume patients with large vessel occlusion successfully treated therapy. Materials and Methods The accuracy CTA source image-based CNN prediction was evaluated against...

PurposeTo determine the effects of patient vertical off-centering when using organ-based tube current modulation (OBTCM) in chest computed tomography (CT) with focus on breast dose.Materials and methodsAn anthropomorphic adult female phantom two different attachment sizes was scanned GE Revolution EVO Siemens Definition Edge CT systems clinical protocols anterior-to-posterior scouts. Scans without OBTCM were performed at table heights (GE: centered, ±6 cm, ± 3 cm; Siemens: −6 cm). The dose...