A5074300820- Radiomics and Machine Learning in Medical Imaging
- Lung Cancer Diagnosis and Treatment
- Medical Imaging and Analysis
- Medical Image Segmentation Techniques
- Chronic Obstructive Pulmonary Disease (COPD) Research
- Advanced Neuroimaging Techniques and Applications
- Medical Imaging Techniques and Applications
- Machine Learning in Healthcare
- COVID-19 diagnosis using AI
- Advanced MRI Techniques and Applications
- Topic Modeling
- Functional Brain Connectivity Studies
- Total Knee Arthroplasty Outcomes
- Human Pose and Action Recognition
- Brain Tumor Detection and Classification
- Medical Imaging and Pathology Studies
- Spine and Intervertebral Disc Pathology
- Tracheal and airway disorders
- Domain Adaptation and Few-Shot Learning
- Spinal Fractures and Fixation Techniques
- Otolaryngology and Infectious Diseases
- Image and Object Detection Techniques
- Advanced Image and Video Retrieval Techniques
- Gastrointestinal Bleeding Diagnosis and Treatment
- Cell Image Analysis Techniques
Siemens Healthcare (United States)
2016-2019
Siemens (United States)
2004-2018
Siemens (Germany)
2006-2012
Princeton University
2003-2010
Université de Technologie de Compiègne
2006-2007
Centre National de la Recherche Scientifique
2006
Abstract Tractography based on non-invasive diffusion imaging is central to the study of human brain connectivity. To date, approach has not been systematically validated in ground truth studies. Based a simulated data set with tracts, we organized an open international tractography challenge, which resulted 96 distinct submissions from 20 research groups. Here, report encouraging finding that most state-of-the-art algorithms produce tractograms containing 90% bundles (to at least some...
This paper describes a framework for establishing reference airway tree segmentation, which was used to quantitatively evaluate fifteen different extraction algorithms in standardized manner. Because of the sheer difficulty involved manually constructing complete standard from scratch, we propose construct using results all that are be evaluated. We start by subdividing each segmented into its individual branch segments. Each segment is then visually scored trained observers determine...
Abstract Fiber tractography based on non-invasive diffusion imaging is at the heart of connectivity studies human brain. To date, approach has not been systematically validated in ground truth studies. Based a simulated brain dataset with white matter tracts, we organized an open international challenge, which resulted 96 distinct submissions from 20 research groups. While most state-of-the-art algorithms reconstructed 90% bundles to least some extent, average they produced four times more...
Abstract With the rapid growth and increasing use of brain MRI, there is an interest in automated image classification to aid human interpretation improve workflow. We aimed train a deep convolutional neural network assess its performance identifying abnormal MRIs critical intracranial findings including acute infarction, hemorrhage mass effect. A total 13,215 clinical MRI studies were categorized training (74%), validation (9%), internal testing (8%) external datasets. Up eight contrasts...
We propose in this paper a novel approach to the automatic segmentation of lung nodules given volume interest (VOI) from high resolution multi-slice CT images by dynamically initializing and adjusting 3D template analyzing its cross correlation with structure interest. First, thresholding techniques are used separate background voxels. The interest, comprising nodule candidate possible attached vessels, is then extracted excluding any part chest wall inside VOI. Afterwards, proposed method...
Pulmonary diseases such as bronchiectasis, asthma, and emphysema are characterized by abnormalities in airway dimensions. Multi-slice computed tomography (MSCT) has become one of the primary means to depict these abnormalities, availability high-resolution near-isotropic data makes it possible evaluate airways at oblique angles scanner plane. However, currently, clinical evaluation is typically limited subjective visual inspection only: systematic take advantage not proved practical without...
Multi-slice computed tomography (CT) provides a promising technology for lung cancer detection and treatment. To optimize automatic detections of more complete spectrum nodules on CT requires multiple specialized algorithms in coherently integrated system. We have developed knowledge-based system nodule analysis, which integrates several robust novel to detect different types nodules, including those attached the chest wall, adjacent or fed by vessels, solitary simultaneously. The...
We present an algorithm for local surface smoothing in a defined Volume of Interest (VOI) cropped from 3D volume data, such as lung CT data. There is generally smooth and piecewise linear the VOI, with one or more bumps on surface. In can be nodules that are grown chest wall, which represent possibility cancer. Through smoothing, segmented wall its size measured diagnostic evidence. The has advantage high consistency robustness, useful segmentation module Compute Aided Diagnosis (CAD) system.
Bronchiectasis, the permanent dilatation of airways, is frequently evaluated by computed tomography (CT) in order to determine disease progression and response treatment. Normal airways have diameters approximately same size as their accompanying artery, most scoring systems for quantifying bronchiectasis severity ask physicians estimate broncho-arterial ratio. However, lack standardization coupled with inter-observer variability limits diagnostic sensitivity ability make reliable...
This paper describes an automatic algorithm to extract the knee frame of reference from 3D MR isotropic scans. The method ultimately seeks determine two lines that are tangent bottom condyles in axial and a coronal plane. It consists three major parts, initial detection joint using Hidden Markov Models, femur segmentation Random Walker segmentation, finally condyle detection. We demonstrate on 30 datasets our is very robust performs at same level as human reader.
Ziel: Klinische Prüfung eines Softwarealgorithmus, der bei Suche korrespondierender Lungenrundherde in CT-Verlaufskontrollen Unterstützung bieten soll, und Identifizierung Faktoren, die Rate korrekt lokalisierter Herde beeinflussen. Methode: 11 Patienten mit 22 Mehrdetektor-Spiral-CT-Untersuchungen des Thorax (Siemens Somatom VZ; Röhrenspannung 120 kVp; effektiver Röhrenstrom 20 oder 100 mAs; Kollimation 4 × 1 mm; Rekonstruktionsinkrement 0,8 mm) insgesamt 190 Rundherden wurden dem sog....
Chronic airway disease causes structural changes in the lungs including peribronchial thickening and dilatation. Multi-detector computed tomography (CT) yields detailed near-isotropic images of lungs, thus potential to obtain quantitative measurements lumen diameter wall thickness. Such would allow standardized assessment, physicians diagnose locate abnormalities, adapt treatment, monitor progress over time. However, due sheer number airways per patient, systematic analysis is infeasible...
In this work, we have developed a novel knowledge-driven quasi-global method for fast and robust registration of thoracic-abdominal CT cone beam (CBCT) scans. While the use CBCT in operating rooms has become common practice, there is an increasing demand on with pre-operative scans, many cases, One major challenges CT/CBCT from various fields view (FOVs) two imaging modalities. The proposed approach utilizes <i>a priori</i> knowledge anatomy to generate 2D targeted projection (ATP) images...
During the last decade, a multitude of novel quantitative and semiquantitative MRI techniques have provided new information about pathophysiology neurological diseases. Yet, selection most relevant contrasts for given pathology remains challenging. In this work, we developed validated method, Gated-Attention MEchanism Ranking multi-contrast in brain (GAMER MRI), to rank relative importance MR measures classification well understood ischemic stroke lesions. Subsequently, applied method...
Arijit Sehanobish, McCullen Sandora, Nabila Abraham, Jayashri Pawar, Danielle Torres, Anasuya Das, Murray Becker, Richard Herzog, Benjamin Odry, Ron Vianu. Proceedings of the 2022 Conference North American Chapter Association for Computational Linguistics: Human Language Technologies: Industry Track. 2022.
Lung nodules that exhibit growth over time are considered highly suspicious for malignancy. We present a completely automated system detection of growing lung nodules, using initial and follow-up multi-slice CT studies. The begins with automatic in the later study, generating preliminary list candidate nodules. Next an registering locations two studies matches each study to its corresponding position earlier study. Then method segmentation is applied matching location, computed volumes...
This paper presents a fast and efficient method to determine intervertebral disk orientation in magnetic resonance (MR) image of the spine. The algorithm originates from active contour theory enforces shape constraint avoid leaks through weak or non-existent boundaries. represents vertebra as rectangle, modeled semi-affine transformation applied unit square. A regional flow integrated along rectangle's perimeter updates achieve segmentation. Further constraints are added so that adjacent...