A Method for Efficient De-identification of DICOM Metadata and Burned-in Pixel Text.

De-identification of DICOM images is an essential component of medical image research. While many established methods exist for the safe removal of protected health information (PHI) in DICOM metadata, approaches for the removal of PHI "burned-in" to image pixel data are typically manual, and automated high-throughput approaches are not well validated. Emerging optical character recognition (OCR) models can potentially detect and remove PHI-bearing text from medical images but are very time-consuming to run on the high volume of images found in typical research studies. We present a data processing method that performs metadata de-identification for all images combined with a targeted approach to only apply OCR to images with a high likelihood of burned-in text. The method was validated on a dataset of 415,182 images across ten modalities representative of the de-identification requests submitted at our institution over a 20-year span. Of the 12,578 images in this dataset with burned-in text of any kind, only 10 passed undetected with the method. OCR was only required for 6050 images (1.5% of the dataset).

Duke Liver Dataset: A Publicly Available Liver MRI Dataset with Liver Segmentation Masks and Series Labels.

The Duke Liver Dataset contains 2146 abdominal MRI series from 105 patients, including a majority with cirrhotic features, and 310 image series with corresponding manually segmented liver masks.

Exercise-induced irregular right heart flow dynamics in adolescents and young adults born preterm.

BACKGROUND: Preterm birth has been linked to an elevated risk of heart failure and cardiopulmonary disease later in life. With improved neonatal care and survival, most infants born preterm are now reaching adulthood. In this study, we used 4D flow cardiovascular magnetic resonance (CMR) coupled with an exercise challenge to assess the impact of preterm birth on right heart flow dynamics in otherwise healthy adolescents and young adults who were born preterm. METHODS: Eleven young adults and 17 adolescents born preterm (< 32 weeks of gestation and < 1500 g birth weight) were compared to 11 young adult and 18 adolescent age-matched controls born at term. Stroke volume, cardiac output, and flow in the main pulmonary artery were quantified with 4D flow CMR. Kinetic energy and vorticity were measured in the right ventricle. All parameters were measured at rest and during exercise at a power corresponding to 70% VO2max for each subject. Multivariate linear regression was used to perform age-adjusted term-preterm comparisons. RESULTS: With exercise, stroke volume increased 10 ± 21% in term controls and decreased 4 ± 18% in preterm born subjects (p = 0.007). This resulted in significantly reduced capacity to increase cardiac output in response to exercise stress for the preterm group (58 ± 26% increase in controls, 36 ± 27% increase in preterm, p = 0.004). Elevated kinetic energy (KEterm = 71 ± 22 nJ, KEpreterm = 87 ± 38 nJ, p = 0.03) and vorticity (ωterm = 79 ± 16 s-1, ωpreterm = 94 ± 32 s-1, p = 0.01) during diastole in the right ventricle (RV) suggested altered RV flow dynamics in the preterm subjects. Streamline visualizations showed altered structure to the diastolic filling vortices in those born preterm. CONCLUSIONS: For the participants examined here, preterm birth appeared to result in altered right-heart flow dynamics as early as adolescence, especially during diastole. Future studies should evaluate whether the altered dynamics identified here evolves into cardiopulmonary disease later in life. Trial registration None.

Altered Right Ventricular Filling at Four-dimensional Flow MRI in Young Adults Born Prematurely.

PURPOSE: To use four-dimensional (4D) flow MRI to measure intraventricular flow in young adults who were born prematurely to investigate mechanisms that may account for increased heart failure risk in this population. MATERIALS AND METHODS: In this secondary analysis of a prospective study, a total of 56 young adults participated in an observational cardiac 4D flow MRI study from 2016 to 2020. There were 35 participants who had been born moderately to extremely prematurely (birth weight <1500 g or gestational age ≤32 weeks; 23 women; mean age, 26 years ± 4) and 21 term-born participants (11 women; mean age, 25 years ± 3). Participants underwent cardiac MRI, including cine cardiac structure and function assessment, as well as 4D flow MRI. In each ventricle, normalized kinetic energy (KE/end diastolic volume) and flow through the atrioventricular valve were computed and compared between term-born and preterm participants at systolic and diastolic (early diastolic filling rate [E wave] and late diastolic filling [atrial contraction] rate [A wave]) time points by using Wilcoxon rank-sum tests. RESULTS: Preterm-born participants had lower right ventricular (RV) E wave/A wave (E/A) KE ratios (2.4 ± 1.7 vs 3.5 ± 1.4; P <.01) and lower E/A peak filling rate ratios (computed from RV volume-time curves; 2.3 ± 1.3 vs 3.5 ± 2.5; P = .03). Additionally, viscous energy dissipation was increased during systole (5.7 µW/mL ± 3.0 vs 4.2 µW/mL ± 1.6; P = .03), increased during late diastole (3.9 µW/mL ± 4.0 vs 2.2 µW/mL ± 1.6; P = .03), and summed over the cardiac cycle (2.4 µJ/mL ± 1.0 vs 1.9 µJ/mL ± 0.6; P = .02) in preterm relative to term participants. CONCLUSION: These results suggest that RV diastolic filling is altered in young adults who were born moderately to severely prematurely.Supplemental material is available for this article. Keywords: Adults, Cardiac, Comparative Studies, MR-Imaging, Right Ventricle © RSNA, 2021.

Decreased ventricular size and mass mediate the reduced exercise capacity in adolescents and adults born premature.

BACKGROUND: Premature birth is associated with lower levels of cardiorespiratory fitness (CRF) but the underlying mechanisms responsible remain unclear. This study assessed whether differences in cardiac morphology or function mediate differences in CRF among adolescents and young adults born preterm. METHODS: Adolescents and young adults born moderately to extremely premature (gestational age ≤ 32 weeks or birth weight < 1500 g) and age-matched term born participants underwent resting cardiac MRI and maximal exercise testing. Mediation analysis assessed whether individual cardiovascular variables accounted for a significant proportion of the difference in maximal aerobic capacity between groups. RESULTS: Individuals born preterm had lower VO2max than those born term (41.7 ±â€¯8.6 v 47.5 ±â€¯8.7, p < 0.01). Several variables differed between term and preterm born subjects, including systolic and diastolic blood pressure, mean pulmonary artery pressure, indexed left ventricular end-diastolic volume (LVEDVi), right ventricular end-diastolic volume (RVEDVi), LV mass (LVMi), LV stroke volume index (LVSVi), and LV strain (p < 0.05 for all). Of these variables, LVEDVi, RVEDVi, LVSVi, LVMi, and LV longitudinal strain were significantly related to VO2max (p < 0.05 for all). Significant portions of the difference in VO2max between term and preterm born subjects were mediated by LVEDVi (74.3%, p = 0.010), RVEDVi (50.6%, p = 0.016), and LVMi (43.0%, p = 0.036). CONCLUSIONS: Lower levels of CRF in adolescents and young adults born preterm are mediated by differences in LVEDVi, RVEDVi, and LVMi. This may represent greater risk for long-term cardiac morbidity and mortality in preterm born individuals.

Feasibility of Cardiovascular Four-dimensional Flow MRI during Exercise in Healthy Participants.

PURPOSE: To explore the feasibility of using four-dimensional (4D) flow MRI to quantify blood flow and kinetic energy (KE) in the heart during strenuous exercise. MATERIALS AND METHODS: For this prospective study, cardiac 4D flow MRI was performed in 11 healthy young adult participants (eight men, three women; mean age, 26 years ± 1 [standard deviation]) at rest and during exercise with an MRI-compatible exercise stepper between March 2016 and July 2017. Flow was measured in the ascending aorta (AAo) and main pulmonary artery (MPA). KE was quantified in the left and right ventricle. Significant changes in flow and KE during exercise were identified by using t tests. Repeatability was assessed with inter- and intraobserver comparisons and an analysis of internal flow consistency. RESULTS: Nine participants successfully completed both rest and exercise imaging. Internal flow consistency analysis in systemic and pulmonary circulation showed average relative differences of 10% at rest and 16% during exercise. For flow measurements in the AAo and MPA, relative differences between observers never exceeded 6% in any vessel and showed excellent correlation, even during exercise. Relative differences were increased for KE, typically on the order of 30%, with poor interobserver correlation between measurements. CONCLUSION: Four-dimensional flow MRI can quantify increases in flow in the AAo and MPA during strenuous exercise and is highly repeatable. KE had reduced repeatability because of suboptimal segmentation methods and requires further development before clinical implementation. Supplemental material is available for this article. © RSNA, 2020See also the commentary by Markl and Lee in this issue.

Reduced regional flow in the left ventricle after anterior acute myocardial infarction: a case control study using 4D flow MRI.

BACKGROUND: Acute myocardial infarction (AMI) alters left ventricular (LV) hemodynamics, resulting in decreased global LV ejection fraction and global LV kinetic energy. We hypothesize that anterior AMI effects localized alterations in LV flow and developed a regional approach to analyze these local changes with 4D flow MRI. METHODS: 4D flow cardiac magnetic resonance (CMR) data was compared between 12 anterior AMI patients (11 males; 66 ± 12yo; prospectively acquired in 2016-2017) and 19 healthy volunteers (10 males; 40 ± 16yo; retrospective from 2010 to 2011 study). The LV cavity was contoured on short axis cine steady-state free procession CMR and partitioned into three regions: base, mid-ventricle, and apex. 4D flow data was registered to the short axis segmentation. Peak systolic and diastolic through-plane flows were compared region-by-region between groups using linear models of flow with age, sex, and heart rate as covariates. RESULTS: Peak systolic flow was reduced in anterior AMI subjects compared to controls in the LV mid-ventricle (fitted reduction = 3.9 L/min; P = 0.01) and apex (fitted reduction = 1.4 L/min; P = 0.02). Peak diastolic flow was also lower in anterior AMI subjects compared to controls in the apex (fitted reduction = 2.4 L/min; P = 0.01). CONCLUSIONS: A regional method to analyze 4D LV flow data was applied in anterior AMI patients and controls. Anterior AMI patients had reduced regional flow relative to controls.

Uteroplacental and Fetal 4D Flow MRI in the Pregnant Rhesus Macaque.

BACKGROUND: Pregnancy complications are often associated with poor uteroplacental vascular adaptation and standard diagnostics are unable to reliably quantify flow in all uteroplacental vessels and have poor sensitivity early in gestation. PURPOSE: To investigate the feasibility of using 4D flow MRI to assess total uteroplacental blood flow in pregnant rhesus macaques as a precursor to human studies. STUDY TYPE: Retrospective feasibility study. ANIMAL MODEL: Fifteen healthy, pregnant rhesus macaques ranging from the 1st trimester to 3rd trimester of gestation. FIELD STRENGTH/SEQUENCE: Abdominal 4D flow MRI was performed on a 3.0T scanner with a radially undersampled phase contrast (PC) sequence. Reference ferumoxytol-enhanced angiograms were acquired with a 3D ultrashort echo time sequence with a center-out radial trajectory. ASSESSMENT: Repeatability of flow measurements was assessed with scans performed same-day and on consecutive days in the uterine arteries and ovarian veins. In-flow was compared against out-flow in the uterus, umbilical cord, and fetal heart with a conservation of mass analysis. Conspicuity of uteroplacental vessels was qualitatively compared between PC angiograms derived from 4D flow data and ferumoxytol-enhanced angiograms. STATISTICAL TESTS: Bland-Altman analysis was used to quantify same-day and consecutive-day repeatability. RESULTS: Same-day flow measurements showed an average difference between scans of 13% in both the uterine arteries and ovarian veins, while consecutive-day measurements showed average differences of 22% and 24%, respectively. Comparisons of in-flow and out-flow showed average differences of 15% in the uterus, 8% in fetal heart, and 15% in the umbilical cord. PC angiograms showed similar depiction of main uteroplacental vessels as high-resolution, ferumoxytol-enhanced angiograms. DATA CONCLUSION: 4D flow MRI could be used in the rhesus macaque for repeatable flow measurements in the uteroplacental and fetal vasculature, setting the stage for future human studies. LEVEL OF EVIDENCE: 2 Technical Efficacy: Stage 1 J. Magn. Reson. Imaging 2019;49:534-545.

MRI assessment of aortic flow in patients with pulmonary arterial hypertension in response to exercise.

BACKGROUND: While primarily a right heart disease, pulmonary arterial hypertension (PAH) can impact left heart function and aortic flow through a shifted interventricular septum from right ventricular pressure overload and reduced left ventricular preload, among other mechanisms. In this study, we used phase contrast (PC) MRI and a modest exercise challenge to examine the effects of PAH on systemic circulation. While exercise challenges are typically performed with ultrasound in the clinic, MRI exercise studies allow for more reproducible image alignment, more accurate flow quantification, and improved tissue contrast. METHODS: Six PAH patients and fifteen healthy controls (8 older age-matched, 7 younger) exercised in the magnet bore with an MRI-compatible exercise device that allowed for scanning immediately following cessation of exercise. PC scans were performed in the ascending aorta during a breath hold immediately after modest exercise to non-invasively measure stroke volume (SV), cardiac output (CO), aortic peak systolic flow (PSF), and aortic wall stiffness via relative area change (RAC). RESULTS: Images following exercise showed mild blurring, but were high enough quality to allow for segmentation of the aorta. While SV was approximately 30% lower in PAH patients (SVPAH,rest = 67 ± 16 mL; SVPAH,stress = 90 ± 42 mL) than age-matched controls (SV,older,rest = 93 ± 16 mL; SVolder,stress = 133 ± 40 mL) at both rest and following exercise, CO was similar for both groups following exercise (COPAH,stress = 10.8 ± 5.7 L/min; COolder,stress = 11.8 ± 5.0 L/min). This was achieved through a compensatory increase in heart rate in the PAH subjects (74% increase as compared to 29% in age-matched controls). The PAH subjects also demonstrated reduced aortic peak systolic flow relative to the healthy controls (PSFPAH,rest = 309 ± 52 mL/s; PSFolder,rest = 416 ± 114 mL/s; PSFPAH,stress = 388 ± 113 mL/s; PSFolder,stress = 462 ± 176 mL/s). PAH patients and older controls demonstrated stiffer aortic walls when compared to younger controls (RACPAH,rest = 0.15 ± 0.05; RAColder,rest = 0.17 ± 0.05; RACyoung,rest = 0.28 ± 0.08). CONCLUSIONS: PC MRI following a modest exercise challenge was capable of detecting differences in left heart dynamics likely induced from PAH. These results demonstrated that PAH can have a significant influence on systemic flow, even when the patient has no prior left heart disease. Image quantification following exercise could likely be improved in future studies through the implementation of free-breathing or real-time MRI acquisitions. TRIAL REGISTRATION: Retrospectively registered on 02/26/2018 (TRN: NCT03523910 ).