Spatiotemporal Mammography-based Deep Learning Model for Improved Breast Cancer Risk Prediction.

Breast cancer is one of the most prevalent cancers among women. It is the second leading cause of death in cancer-related deaths. Early detection and personalized risk assessment can reduce the mortality rate and improve survival rates.

Bifurcation analysis of motoneuronal excitability mechanisms under normal and ALS conditions.

Introduction: Bifurcation analysis allows the examination of steady-state, non-linear dynamics of neurons and their effects on cell firing, yet its usage in neuroscience is limited to single-compartment models of highly reduced states. This is primarily due to the difficulty in developing high-fidelity neuronal models with 3D anatomy and multiple ion channels in XPPAUT, the primary bifurcation analysis software in neuroscience.

Methods: To facilitate bifurcation analysis of high-fidelity neuronal models under normal and disease conditions, we developed a multi-compartment model of a spinal motoneuron (MN) in XPPAUT and verified its firing accuracy against its original experimental data and against an anatomically detailed cell model that incorporates known MN non-linear firing mechanisms.

Bilateral Analysis Boosts the Performance of Mammography-based Deep Learning Models in Breast Cancer Risk Prediction.

Breast cancer is one of the leading causes of death among women. Early prediction of breast cancer can significantly improve the survival rates. Breast density was proven as a reliable risk factor.

Improved Automatic Grading of Diabetic Retinopathy Using Deep Learning and Principal Component Analysis.

Diabetic retinopathy (DR) is one of the most common chronic diseases around the world. Early screening and diagnosis of DR patients through retinal fundus is always preferred. However, image screening and diagnosis is a highly time-consuming task for clinicians.

Improved Centerline Extraction in Fully Automated Coronary Ostium Localization and Centerline Extraction Framework using Deep Learning.

Coronary artery extraction in cardiac CT angiography (CCTA) image volume is a necessary step for any quantitative assessment of stenoses and atherosclerotic plaque. In this work, we propose a fully automated workflow that depends on convolutional networks to extract the centerlines of the coronary arteries from CCTA image volumes, starting from identifying the ostium points and then tracking the vessel till its end based on its radius and direction. First, a regression U-Net is employed to identify the ostium points in the image volume, then these points are fed to an orientation and radius predictor CNN model to track and extract each artery till its end point.

Detecting liver fibrosis using a machine learning-based approach to the quantification of the heart-induced deformation in tagged MR images.

Liver disease causes millions of deaths per year worldwide, and approximately half of these cases are due to cirrhosis, which is an advanced stage of liver fibrosis that can be accompanied by liver failure and portal hypertension. Early detection of liver fibrosis helps in improving its treatment and prevents its progression to cirrhosis. In this work, we present a novel noninvasive method to detect liver fibrosis from tagged MRI images using a machine learning-based approach.

Imaging sequence for joint myocardial T mapping and fat/water separation.

Purpose: To develop and evaluate an imaging sequence to simultaneously quantify the epicardial fat volume and myocardial T relaxation time.

Methods: We introduced a novel simultaneous myocardial T mapping and fat/water separation sequence (joint T -fat/water separation). Dixon reconstruction is performed on a dual-echo data set to generate water/fat images.

Gray blood late gadolinium enhancement cardiovascular magnetic resonance for improved detection of myocardial scar.

Background: Low scar-to-blood contrast in late gadolinium enhanced (LGE) MRI limits the visualization of scars adjacent to the blood pool. Nulling the blood signal improves scar detection but results in lack of contrast between myocardium and blood, which makes clinical evaluation of LGE images more difficult.

Methods: GB-LGE contrast is achieved through partial suppression of the blood signal using T magnetization preparation between the inversion pulse and acquisition.

Improved segmented modified Look-Locker inversion recovery T1 mapping sequence in mice.

Object: To develop and evaluate a 2D modified Look-Locker (MOLLI) for high-resolution T1 mapping in mice using a 3T MRI scanner.

Materials And Methods: To allow high-resolution T1 mapping in mice at high heart rates a multi-shot ECG-triggered 2D MOLLI sequence was developed. In the proposed T1 mapping sequence the optimal number of sampling points and pause cardiac cycles following an initial adiabatic inversion pulse was investigated in a phantom.

Improved dark blood late gadolinium enhancement (DB-LGE) imaging using an optimized joint inversion preparation and T magnetization preparation.

Purpose: To develop a dark blood-late gadolinium enhancement (DB-LGE) sequence that improves scar-blood contrast and delineation of scar region.

Methods: The DB-LGE sequence uses an inversion pulse followed by T magnetization preparation to suppress blood and normal myocardium. Time delays inserted after preparation pulses and T -magnetization-prep duration are used to adjust tissue contrast.

Clinical performance of high-resolution late gadolinium enhancement imaging with compressed sensing.

Purpose: To evaluate diagnostic image quality of 3D late gadolinium enhancement (LGE) with high isotropic spatial resolution (∼1.4 mm ) images reconstructed from randomly undersampled k-space using LOw-dimensional-structure Self-learning and Thresholding (LOST).

Materials And Methods: We prospectively enrolled 270 patients (181 men; 55 ± 14 years) referred for myocardial viability assessment.

Relationship between native papillary muscle T time and severity of functional mitral regurgitation in patients with non-ischemic dilated cardiomyopathy.

Background: Functional mitral regurgitation is one of the severe complications of non-ischemic dilated cardiomyopathy (DCM). Non-contrast native T mapping has emerged as a non-invasive method to evaluate myocardial fibrosis. We sought to evaluate the potential relationship between papillary muscle T time and mitral regurgitation in DCM patients.

Comparison of spoiled gradient echo and steady-state free-precession imaging for native myocardial T1 mapping using the slice-interleaved T1 mapping (STONE) sequence.

Cardiac T1 mapping allows non-invasive imaging of interstitial diffuse fibrosis. Myocardial T1 is commonly calculated by voxel-wise fitting of the images acquired using balanced steady-state free precession (SSFP) after an inversion pulse. However, SSFP imaging is sensitive to B1 and B0 imperfection, which may result in additional artifacts.

Myocardial Native T1 Time in Patients With Hypertrophic Cardiomyopathy.

In hypertrophic cardiomyopathy (HC), there are significant variations in left ventricular (LV) wall thickness and fibrosis, which necessitates a volumetric coverage. Slice-interleaved T1 (STONE) mapping sequence allows for the assessment of native T1 time with complete coverage of LV myocardium. The aims of this study were to evaluate spatial heterogeneity of native T1 time in patients with HC.

Reproducibility of myocardial T and T relaxation time measurement using slice-interleaved T and T mapping sequences.

Purpose: To assess measurement reproducibility and image quality of myocardial T and T maps using free-breathing slice-interleaved T and T mapping sequences at 1.5 Tesla (T).

Materials And Methods: Eleven healthy subjects (33 ± 16 years; 6 males) underwent a slice-interleaved T and T mapping test/retest cardiac MR study at 1.

Native Myocardial T1 as a Biomarker of Cardiac Structure in Non-Ischemic Cardiomyopathy.

Diffuse myocardial fibrosis is involved in the pathology of nonischemic cardiomyopathy (NIC). Recently, the application of native (noncontrast) myocardial T1 measurement has been proposed as a method for characterizing diffuse interstitial fibrosis. To determine the association of native T1 with myocardial structure and function, we prospectively studied 39 patients with NIC (defined as left ventricular ejection fraction (LVEF) ≤ 50% without cardiac magnetic resonance (CMR) evidence of previous infarction) and 27 subjects with normal LVEF without known overt cardiovascular disease.

Left ventricular native T1 time and the risk of atrial fibrillation recurrence after pulmonary vein isolation in patients with paroxysmal atrial fibrillation.

Background: Native T1 mapping has emerged as a noninvasive non-contrast magnetic resonance imaging (MRI) method to assess for diffuse myocardial fibrosis. However, LV native T1 time in AF patients and its clinical relevance are unclear.

Methods: Fifty paroxysmal AF patients referred for PVI (60 ± 8 years, 37 male) and 11 healthy control subjects (57 ± 8 years, 10 male) were studied.

Free-breathing slice-interleaved myocardial T2 mapping with slice-selective T2 magnetization preparation.

Purpose: To develop and evaluate a free-breathing slice-interleaved T2 mapping sequence by proposing a new slice-selective T2 magnetization preparation (T2 prep) sequence that allows interleaved data acquisition for different slices in subsequent heartbeats.

Methods: We developed a slice-selective T2 prep for myocardial T2 mapping by adding slice-selective gradients to a conventional single-slice T2 prep sequence. In this sequence, five slices are acquired during five consecutive heartbeats, each using a slice-selective T2 prep.

Joint myocardial T1 and T2 mapping using a combination of saturation recovery and T2 -preparation.

Purpose: To develop a heart-rate independent breath-held joint T1 -T2 mapping sequence for accurate simultaneous estimation of coregistered myocardial T1 and T2 maps.

Methods: A novel preparation scheme combining both a saturation pulse and T2 -preparation in a single R-R interval is introduced. The time between these two pulses, as well as the duration of the T2 -preparation is varied in each heartbeat, acquiring images with different T1 and T2 weightings, and no magnetization dependence on previous images.

Impact of motion correction on reproducibility and spatial variability of quantitative myocardial T2 mapping.

Background: To evaluate and quantify the impact of a novel image-based motion correction technique in myocardial T2 mapping in terms of measurement reproducibility and spatial variability.

Methods: Twelve healthy adult subjects were imaged using breath-hold (BH), free breathing (FB), and free breathing with respiratory navigator gating (FB + NAV) myocardial T2 mapping sequences. Fifty patients referred for clinical CMR were imaged using the FB + NAV sequence.

Accelerated free breathing ECG triggered contrast enhanced pulmonary vein magnetic resonance angiography using compressed sensing.

Background: To investigate the feasibility of accelerated electrocardiogram (ECG)-triggered contrast enhanced pulmonary vein magnetic resonance angiography (CE-PV MRA) with isotropic spatial resolution using compressed sensing (CS).

Methods: Nineteen patients (59±13 y, 11 M) referred for MR were scanned using the proposed accelerated free breathing ECG-triggered 3D CE-PV MRA sequence (FOV=340×340×110 mm3, spatial resolution=1.5×1.

Free-breathing combined three-dimensional phase sensitive late gadolinium enhancement and T1 mapping for myocardial tissue characterization.

Purpose: To develop a novel MR sequence for combined three-dimensional (3D) phase-sensitive (PS) late gadolinium enhancement (LGE) and T1 mapping to allow for simultaneous assessment of focal and diffuse myocardial fibrosis.

Methods: In the proposed sequence, four 3D imaging volumes are acquired with different T1 weightings using a combined saturation and inversion preparation, after administration of a gadolinium contrast agent. One image is acquired fully sampled with the inversion time selected to null the healthy myocardial signal (the LGE image).

Accelerated three-dimensional cine phase contrast imaging using randomly undersampled echo planar imaging with compressed sensing reconstruction.

The aim of this study was to implement and evaluate an accelerated three-dimensional (3D) cine phase contrast MRI sequence by combining a randomly sampled 3D k-space acquisition sequence with an echo planar imaging (EPI) readout. An accelerated 3D cine phase contrast MRI sequence was implemented by combining EPI readout with randomly undersampled 3D k-space data suitable for compressed sensing (CS) reconstruction. The undersampled data were then reconstructed using low-dimensional structural self-learning and thresholding (LOST).

MR myocardial perfusion imaging: insights on techniques, analysis, interpretation, and findings.

Coronary microcirculatory dysfunction has a fundamental role in the pathophysiology of ischemic coronary artery disease (CAD) as well as various other cardiovascular disorders. Invasive coronary angiography remains the standard of reference for diagnosis of CAD. Nevertheless, it has been well acknowledged that the degree of luminal narrowing of epicardial coronary lesions detected at angiography is a poor predictor of the functional severity of the lesion.

An augmented Lagrangian based compressed sensing reconstruction for non-Cartesian magnetic resonance imaging without gridding and regridding at every iteration.

Background: Non-Cartesian trajectories are used in a variety of fast imaging applications, due to the incoherent image domain artifacts they create when undersampled. While the gridding technique is commonly utilized for reconstruction, the incoherent artifacts may be further removed using compressed sensing (CS). CS reconstruction is typically done using conjugate-gradient (CG) type algorithms, which require gridding and regridding to be performed at every iteration.