• Investigative Magnetic Resonance Imaging
• /
• v.23 no.2
• /
• pp.114-124
• /
• 2019

Purpose: We investigate biases in the assessments of left ventricular function (LVF), by compressed sensing (CS)-cine magnetic resonance imaging (MRI). Materials and Methods: Cardiovascular cine images with short axis view, were obtained for 8 volunteers without CS. LVFs were assessed with subsampled data, with compression factors (CF) of 2, 3, 4, and 8. A semi-automatic segmentation program was used, for the assessment. The assessments by 3 CS methods (ITSC, FOCUSS, and view sharing (VS)), were compared to those without CS. Bland-Altman analysis and paired t-test were used, for comparison. In addition, real-time CS-cine imaging was also performed, with CF of 2, 3, 4, and 8 for the same volunteers. Assessments of LVF were similarly made, for CS data. A fixed compensation technique is suggested, to reduce the bias. Results: The assessment of LVF by CS-cine, includes bias and random noise. Bias appeared much larger than random noise. Median of end-diastolic volume (EDV) with CS-cine (ITSC or FOCUSS) appeared -1.4% to -7.1% smaller, compared to that of standard cine, depending on CF from (2 to 8). End-systolic volume (ESV) appeared +1.6% to +14.3% larger, stroke volume (SV), -2.4% to -16.4% smaller, and ejection fraction (EF), -1.1% to -9.2% smaller, with P < 0.05. Bias was reduced from -5.6% to -1.8% for EF, by compensation applied to real-time CS-cine (CF = 8). Conclusion: Loss of temporal resolution by adopting missing data from nearby cardiac frames, causes an underestimation for EDV, and an overestimation for ESV, resulting in underestimations for SV and EF. The bias is not random. Thus it should be removed or reduced for better diagnosis. A fixed compensation is suggested, to reduce bias in the assessment of LVF.

• Journal of IKEEE
• /
• v.23 no.4
• /
• pp.1408-1414
• /
• 2019

Deep artificial neural network with transfer learning is applied to compressed sensing cardiovascular MRI. Transfer learning is a method that utilizes structure, filter kernels, and weights of the network used in prior learning for current learning or application. The transfer learning is useful in accelerating learning speed, and in generalization of the neural network when learning data is limited. From a cardiac MRI experiment, with 8 healthy volunteers, the neural network with transfer learning was able to reduce learning time by a factor of more than five compared to that with standalone learning. Using test data set, reconstructed images with transfer learning showed lower normalized mean square error and better image quality compared to those without transfer learning.

• Investigative Magnetic Resonance Imaging
• /
• v.18 no.1
• /
• pp.7-16
• /
• 2014

Purpose : To evaluate the prevalence and pattern of perfusion defect (PD) on first-pass stress perfusion MR imaging in relation with the degree of left ventricular hypertrophy (LVH) and late gadolinium-enhancement (LGE) in patients with apical hypertrophic cardiomyopathy (APH). Materials and Methods: Cardiac MR imaging with first-pass stress perfusion, cine, and LGE sequence was performed in 26 patients with APH from January 2008 to December 2012. We analyzed a total of 416 segments for LV wall thickness on end-diastolic phase of cine images, and evaluated the number of hypertrophied segment and number of consecutive hypertrophied segment (NCH). We assessed the presence or absence of PD and LGE from all patients. If there was PD, we subdivided the pattern into sporadic (sporadic-PD) or ring (ring-PD). Using univariate logistic method, we obtained the independent predictor for presence of overall PD and ring-PD. Results: PD on stress perfusion MRI was observed in 20 patients (76.9%), 12 of them (60%) showed ring-PD. Maximal LV wall thickness and number of hypertrophied segment were independent predictors for overall PD (all, p < 0.05). NCH with more than 3 segments was an additional independent factor for ring-PD. However, LGE was not statistically related with PD in patients with APH. Conclusion: About three quarters of the patients with APH showed PD, most of them represented as ring-PD. LVH degree or distribution was related with pattern of PD, however, LGE was not related with PD. Therefore, the clinical significance of PD in the patients with APH seems to be different from those with non-APH, and further comparison study between the two groups should be carried out.

• Investigative Magnetic Resonance Imaging
• /
• v.17 no.3
• /
• pp.200-206
• /
• 2013

Purpose : To assess the effect of applying an automated heart model based measurements of left ventricle (LV) and compare with manual and semi-automated measurements at Cardiovascular MR Imaging. Materials and Methods: Sixty-two patients who underwent cardiac 1.5T MR imaging were included. Steady state free precession cine images of 20 phases per cardiac cycle were obtained in short axis views and both 2-chamber and 4-chamber views. Epicardial and endocardial contours were drawn in manual, automated, and semi-automated ways. Based on these acquired contour sets, the end-diastolic (ED) and end-systolic (ES) volumes, ejection fraction (EF), systolic volume (SV) and LV mass were calculated and compared. Results: In EDV and ESV, the differences among three measurement methods were not statistically significant (P = .399 and .145, respectively). However, in EF, SV, and LV mass, the differences were statistically significant (P=.001, <001, <001, respectively) and the measured value from automated method tend to be consistently higher than the values from other two methods. Conclusion: An automatic heart model-based method grossly overestimate EF, SV and LV mass compared with manual or semi-automated methods. Even though the method saves a considerable amount of efforts, further manual adjustment should be considered in critical clinical cases.

• Investigative Magnetic Resonance Imaging
• /
• v.12 no.1
• /
• pp.40-48
• /
• 2008

Purpose : We sought to determine the early change of ventricular geometry and function after concomitant surgeries of modified Dor procedure and mitral valve annuloplasty by using magnetic resonance imaging. Materials and Methods : We enrolled 21 patients with dilated heart failure who underwent modified Dor procedure (n=8), mitral valve annuloplasty (n=6), or both surgeries (n=7). Cine MRI was used to assess left ventricular dimensions and function before and after surgery. We measured the left ventricular end-diastolic and end-systolic volumes and the dimensions of the left ventricular long-axis and short-axis. Left ventricular stroke volume, ejection fraction, and sphericity index were calculated from these measurements. These parameters were analyzed and compared between three different surgery groups to explain the combined effect of the concomitant surgeries. Results : MRI was performed within average $12\;{\pm}\;15$ days (range 1-58 days) before and $38\;{\pm}\;50$ days (range 7- 231 days) after the surgery. The patients who underwent concomitant surgeries had more profound enlargement of left ventricle and decreased contractility prior to surgery than those in the patients who underwent single surgical procedure. Left ventricular end-diastolic volume and endsystolic volume significantly decreased in all patients regardless of surgery type after surgery. Ejection fraction significantly increased only in the patients who got modified Dor procedure without mitral valve annuloplasty (25.4% to 40.7%). Sphericity index increased in patients with modified Dor procedure but decreased in patients with mitral valve annuloplasty (0.65 to 0.78 vs. 0.75 to 0.65). In the patients who underwent concomitant surgeries showed no significant change in sphericity index after surgery. Conclusion : The early change of the left ventricular geometry and function after the concomitant surgeries with modified Dor procedure and mitral valve annuloplasty in patients with dilated heart failure includes a marked reduction in left ventricular volume and in stroke volume. The shape of the left ventricle does not change because the effect of sphericity index decrease from mitral valve annuloplasty is counteracted by the effect of sphericity index increase from modified Dor procedure. Improvement of left ventricular ejection fraction is not the early change after the concomitant surgeries.