Predicting furan content in a fried dough system using image analysis.

The aim of this paper is to test different models for predicting furan content in a dough system, based on partial least squares regression using colour images. Starch dough systems were fried at five temperatures between 150 and 190 °C and for 5, 7, 9, 11 and 13 min. The furan content was quantified using gas chromatography/mass spectrometry, while the corresponding images were simultaneously obtained and processed in order to extract 2914 features. Good furan content predictions were obtained using computer vision image chromatic features using correlation coefficient of prediction (Rp = 0.86). However, the best prediction correlation was obtained using the image textural features (Rp = 0.93), when the number of features was reduced to 10 by algorithms applications. These results suggest that furan content in fried dough systems can be predicted using features of computer vision images.

Accelerating dual cardiac phase images using undersampled radial phase encoding trajectories.

A three-dimensional dual-cardiac-phase (3D-DCP) scan has been proposed to acquire two data sets of the whole heart and great vessels during the end-diastolic and end-systolic cardiac phases in a single free-breathing scan. This method has shown accurate assessment of cardiac anatomy and function but is limited by long acquisition times. This work proposes to accelerate the acquisition and reconstruction of 3D-DCP scans by exploiting redundant information of the outer k-space regions of both cardiac phases. This is achieved using a modified radial-phase-encoding trajectory and gridding reconstruction with uniform coil combination. The end-diastolic acquisition trajectory was angularly shifted with respect to the end-systolic phase. Initially, a fully-sampled 3D-DCP scan was acquired to determine the optimal percentage of the outer k-space data that can be combined between cardiac phases. Thereafter, prospectively undersampled data were reconstructed based on this percentage. As gold standard images, the undersampled data were also reconstructed using iterative SENSE. To validate the method, image quality assessments and a cardiac volume analysis were performed. The proposed method was tested in thirteen healthy volunteers (mean age, 30years). Prospectively undersampled data (R=4) reconstructed with 50% combination led high quality images. There were no significant differences in the image quality and in the cardiac volume analysis between our method and iterative SENSE. In addition, the proposed approach reduced the reconstruction time from 40min to 1min. In conclusion, the proposed method obtains 3D-DCP scans with an image quality comparable to those reconstructed with iterative SENSE, and within a clinically acceptable reconstruction time.

Realistic aortic phantom to study hemodynamics using MRI and cardiac catheterization in normal and aortic coarctation conditions.

PURPOSE: To design and characterize a magnetic resonance imaging (MRI)-compatible aortic phantom simulating normal and aortic coarctation (AoCo) conditions and to compare its hemodynamics with healthy volunteers and AoCo patients. MATERIALS AND METHODS: The phantom is composed of an MRI-compatible pump, control unit, aortic model, compliance chamber, nonreturn, and shutoff valves. The phantom without and with AoCo (13, 11, and 9 mm) was studied using 2D and 3D phase-contrast data and with a catheterization unit to measure pressures. The phantom data were compared with the mean values of 10 healthy volunteers and two AoCo patients. RESULTS: Hemodynamic parameters in the normal phantom and healthy volunteers were: heart rate: 68/61 bpm, cardiac output: 3.5/4.5 L/min, peak flow and peak velocity (Vpeak) in the ascending aorta (AAo): 270/357 mL/s (significantly, P < 0.05) and 97/107 cm/s (not significantly, P = 0.16), and pressure in the AAo of the normal phantom of 131/58 mmHg. Hemodynamic parameters in the 13, 11, and 9 mm coarctation phantoms and Patients 1 and 2 were: heart rate: 75/75/75/97/78 bpm, cardiac output: 3.3/3.0/2.9/4.0/5.8 L/min, peak flow in the AAo: 245/265/215/244/376 mL/s, Vpeak in the AAo: 96/95/81/196/187 cm/s, Vpeak after the AoCo: 123/187/282/247/165 cm/s, pressure in the AAo: 124/56, 127/51, 133/50, 120/51 and 87/39 mmHg, and a trans-coarctation systolic pressure gradient: 7, 10, 30, 20, and 11 mmHg. CONCLUSION: We propose and characterize a normal and an AoCo phantom, whose hemodynamics, including velocity, flow, and pressure data are within the range of healthy volunteers and patients with AoCo. J. Magn. Reson. Imaging 2016;44:683-697.