Deep Learning-Based Approach for Identifying and Measuring Focal Liver Lesions on Contrast-Enhanced MRI.

BACKGROUND: The number of focal liver lesions (FLLs) detected by imaging has increased worldwide, highlighting the need to develop a robust, objective system for automatically detecting FLLs. PURPOSE: To assess the performance of the deep learning-based artificial intelligence (AI) software in identifying and measuring lesions on contrast-enhanced magnetic resonance imaging (MRI) images in patients with FLLs. STUDY TYPE: Retrospective. SUBJECTS: 395 patients with 1149 FLLs. FIELD STRENGTH/SEQUENCE: The 1.5 T and 3 T scanners, including T1-, T2-, diffusion-weighted imaging, in/out-phase imaging, and dynamic contrast-enhanced imaging. ASSESSMENT: The diagnostic performance of AI, radiologist, and their combination was compared. Using 20 mm as the cut-off value, the lesions were divided into two groups, and then divided into four subgroups: <10, 10-20, 20-40, and ≥40 mm, to evaluate the sensitivity of radiologists and AI in the detection of lesions of different sizes. We compared the pathologic sizes of 122 surgically resected lesions with measurements obtained using AI and those made by radiologists. STATISTICAL TESTS: McNemar test, Bland-Altman analyses, Friedman test, Pearson's chi-squared test, Fisher's exact test, Dice coefficient, and intraclass correlation coefficients. A P-value <0.05 was considered statistically significant. RESULTS: The average Dice coefficient of AI in segmentation of liver lesions was 0.62. The combination of AI and radiologist outperformed the radiologist alone, with a significantly higher detection rate (0.894 vs. 0.825) and sensitivity (0.883 vs. 0.806). The AI showed significantly sensitivity than radiologists in detecting all lesions <20 mm (0.848 vs. 0.788). Both AI and radiologists achieved excellent detection performance for lesions ≥20 mm (0.867 vs. 0.881, P = 0.671). A remarkable agreement existed in the average tumor sizes among the three measurements (P = 0.174). DATA CONCLUSION: AI software based on deep learning exhibited practical value in automatically identifying and measuring liver lesions. TECHNICAL EFFICACY: Stage 2.

Quantitative analysis from ultrafast dynamic contrast-enhanced breast MRI using population-based versus individual arterial input functions, and comparison with semi-quantitative analysis.

PURPOSE: To evaluate the value of inline quantitative analysis of ultrafast dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) using a population-based arterial input function (P-AIF) compared with offline quantitative analysis with an individual AIF (I-AIF) and semi-quantitative analysis for diagnosing breast cancer. METHODS: This prospective study included 99 consecutive patients with 109 lesions (85 malignant and 24 benign). Model-based parameters (Ktrans, kep, and ve) and model-free parameters (washin and washout) were derived from CAIPIRINHA-Dixon-TWIST-VIBE (CDTV) DCE-MRI. Univariate analysis and multivariate logistic regression analysis with forward stepwise covariate selection were performed to identify significant variables. The AUC and F1 score were assessed for semi-quantitative and two quantitative analyses. RESULTS: kep from inline quantitative analysis with P-AIF for diagnosing breast cancer provided an AUC similar to kep from offline quantitative analysis with I-AIF (0.782 vs 0.779, p = 0.954), higher compared to washin from semi-quantitative analysis (0.782 vs 0.630, p = 0.034). Furthermore, the inline quantitative analysis with P-AIF achieved the larger F1 score (0.920) compared with offline quantitative analysis with I-AIF (0.780) and semi-quantitative analysis (0.480). There were no statistically significant differences for kep values between the two quantitative analysis schemes (p = 0.944). CONCLUSION: The inline quantitative analysis with P-AIF from CDTV in characterizing breast lesions could offer similar diagnostic accuracy to offline quantitative analysis with I-AIF, and higher diagnostic accuracy to semi-quantitative analysis.

Comparison of Free-Breathing 3D Phase-Resolved Functional Lung (PREFUL) MRI With Dynamic 19 F Ventilation MRI in Patients With Obstructive Lung Disease and Healthy Volunteers.

BACKGROUND: Non-contrast-enhanced 1 H magnetic resonance imaging (MRI) with full lung coverage shows promise for assessment of regional lung ventilation but a comparison with direct ventilation measurement using 19 F MRI is lacking. PURPOSE: To compare ventilation parameters calculated using 3D phase-resolved functional lung (PREFUL) MRI with 19 F MRI. STUDY TYPE: Prospective. POPULATION: Fifteen patients with asthma, 14 patients with chronic obstructive lung disease, and 13 healthy volunteers. FIELD STRENGTH/SEQUENCE: A 3D gradient-echo pulse sequence with golden-angle increment and stack-of-stars encoding at 1.5 T. ASSESSMENT: All participants underwent 3D PREFUL MRI and 19 F MRI. For 3D PREFUL, static regional ventilation (RVent) and dynamic flow-volume cross-correlation metric (FVL-CM) were calculated. For both parameters, ventilation defect percentage (VDP) values and ventilation defect (VD) maps (including a combination of both parameters [VDPCombined ]) were determined. For 19 F MRI, images from eight consecutive breaths under volume-controlled inhalation of perfluoropropane were acquired. Time-to-fill (TTF) and wash-in (WI) parameters were extracted. For all 19 F parameters, a VD map was generated and the corresponding VDP values were calculated. STATISTICAL TESTS: For all parameters, the relationship between the two techniques was assessed using a Spearman correlation (r). Differences between VDP values were compared using Bland-Altman analysis. For regional comparison of VD maps, spatial overlap and Sørensen-Dice coefficients were computed. RESULTS: 3D PREFUL VDP values were significantly correlated to VDP measures by 19 F (r range: 0.59-0.70). For VDPRVent , no significant bias was observed with VDP of the third and fourth breath (bias range = -6.8:7.7%, P range = 0.25:0.30). For VDPFVL-CM , no significant bias was found with VDP values of fourth-eighth breaths (bias range = -2.0:12.5%, P range = 0.12:0.75). The overall spatial overlap of all VD maps increased with each breath, ranging from 61% to 81%, stabilizing at the fourth breath. DATA CONCLUSION: 3D PREFUL MRI parameters showed moderate to strong correlation with 19 F MRI. Depending on the 3D PREFUL VD map, the best regional agreement was found to 19 F VD maps of third-fifth breath. LEVEL OF EVIDENCE: 3 TECHNICAL EFFICACY: Stage 2.

T1 mapping as a quantitative imaging biomarker for diagnosing cervical cancer: a comparison with diffusion kurtosis imaging.

BACKGROUND: T1 mapping can potentially quantitatively assess the intrinsic properties of tumors. This study was conducted to explore the ability of T1 mapping in distinguishing cervical cancer type, grade, and stage and compare the diagnostic performance of T1 mapping with diffusion kurtosis imaging (DKI). METHODS: One hundred fifty-seven patients with pathologically confirmed cervical cancer were enrolled in this prospectively study. T1 mapping and DKI were performed. The native T1, difference between native and postcontrast T1 (T1diff), mean kurtosis (MK), mean diffusivity (MD), and apparent diffusion coefficient (ADC) were calculated. Cervical squamous cell carcinoma (CSCC) and adenocarcinoma (CAC), low- and high-grade carcinomas, and early- and advanced-stage groups were compared using area under the receiver operating characteristic (AUROC) curves. RESULTS: The native T1 and MK were higher, and the MD and ADC were lower for CSCC than for CAC (all p < 0.05). Compared with low-grade CSCC, high-grade CSCC had decreased T1diff, MD, ADC, and increased MK (p < 0.05). Compared with low-grade CAC, high-grade CAC had decreased T1diff and increased MK (p < 0.05). Native T1 was significantly higher in the advanced-stage group than in the early-stage group (p < 0.05). The AUROC curves of native T1, MK, ADC and MD were 0,772, 0.731, 0.715, and 0.627, respectively, for distinguishing CSCC from CAC. The AUROC values were 0.762 between high- and low-grade CSCC and 0.835 between high- and low-grade CAC, with T1diff and MK showing the best discriminative values, respectively. For distinguishing between advanced-stage and early-stage cervical cancer, only the AUROC of native T1 was statistically significant (AUROC = 0.651, p = 0.002). CONCLUSIONS: Compared with DKI-derived parameters, native T1 exhibits better efficacy for identifying cervical cancer subtype and stage, and T1diff exhibits comparable discriminative value for cervical cancer grade.

Diffusion-sensitized magnetic resonance imaging highlights placental microstructural damage in patients with previous SARS-CoV-2 pregnancy infection.

INTRODUCTION: Severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) has been a major global health problem since December 2019. This work aimed to investigate whether pregnant women's mild and moderate SARS-CoV-2 infection was associated with microstructural and vascular changes in the placenta observable in vivo by Intravoxel Incoherent Motion (IVIM) at different gestational ages (GA). METHODS: This was a retrospective, nested case-control of pregnant women during the SARS-CoV-2 pandemic (COVID-19 group, n = 14) compared to pre-pandemic healthy controls (n = 19). MRI IVIM protocol at 1.5T was constituted of diffusion-weighted (DW) images with TR/TE = 3100/76 ms and 10 b-values (0,10,30,50,75,100,200,400,700,1000s/mm2). Differences between IVIM parameters D (diffusion), and f (fractional perfusion) quantified in the two groups were evaluated using the ANOVA test with Bonferroni correction and linear correlation between IVIM metrics and GA, COVID-19 duration, the delay time between a positive SARS-CoV-2 test and MRI examination (delay-time exam+) was studied by Pearson-test. RESULTS: D was significantly higher in the COVID-19 placentas compared to that of the age-matched healthy group (p < 0.04 in fetal and p < 0.007 in maternal site). No significant difference between f values was found in the two groups suggesting no-specific microstructural damage with no perfusion alteration (potentially quantified by f) in mild/moderate SARS-Cov-2 placentas. A significant negative correlation was found between D and GA in the COVID-19 placentas whereas no significant correlation was found in the control placentas reflecting a possible accelerated senescence process due to COVID-19. DISCUSSION: We report impaired microstructural placental development during pregnancy and the absence of perfusion-IVIM parameter changes that may indicate no perfusion changing through microvessels and microvilli in the placentas of pregnancies with mild/moderate SARS-Cov-2 after reaching negativity.

Effect of CFTR modulator therapy with elexacaftor/tezacaftor/ivacaftor on pulmonary ventilation derived by 3D phase-resolved functional lung MRI in cystic fibrosis patients.

OBJECTIVES: To investigate whether 3D phase-resolved functional lung (PREFUL)-MRI parameters are suitable to measure response to elexacaftor/tezacaftor/ivacaftor (ETI) therapy and their association with clinical outcomes in cystic fibrosis (CF) patients. METHODS: Twenty-three patients with CF (mean age: 21; age range: 14-46) underwent MRI examination at baseline and 8-16 weeks after initiation of ETI. Morphological and 3D PREFUL scans assessed pulmonary ventilation. Morphological images were evaluated using a semi-quantitative scoring system, and 3D PREFUL scans were evaluated by ventilation defect percentage (VDP) values derived from regional ventilation (RVent) and cross-correlation maps. Improved ventilation volume (IVV) normalized to body surface area (BSA) between baseline and post-treatment visit was computed. Forced expiratory volume in 1 second (FEV1) and mid-expiratory flow at 25% of forced vital capacity (MEF25), as well as lung clearance index (LCI), were assessed. Treatment effects were analyzed using paired Wilcoxon signed-rank tests. Treatment changes and post-treatment agreement between 3D PREFUL and clinical parameters were evaluated by Spearman's correlation. RESULTS: After ETI therapy, all 3D PREFUL ventilation markers (all p < 0.0056) improved significantly, except for the mean RVent parameter. The BSA normalized IVVRVent was significantly correlated to relative treatment changes of MEF25 and mucus plugging score (all |r| > 0.48, all p < 0.0219). In post-treatment analyses, 3D PREFUL VDP values significantly correlated with spirometry, LCI, MRI global, morphology, and perfusion scores (all |r| > 0.44, all p < 0.0348). CONCLUSIONS: 3D PREFUL MRI is a very promising tool to monitor CFTR modulator-induced regional dynamic ventilation changes in CF patients. CLINICAL RELEVANCE STATEMENT: 3D PREFUL MRI is sensitive to monitor CFTR modulator-induced regional ventilation changes in CF patients. Improved ventilation volume correlates with the relative change of mucus plugging, suggesting that reduced endobronchial mucus is predominantly responsible for regional ventilation improvement. KEY POINTS: • 3D PREFUL MRI-derived ventilation maps show significantly reduced ventilation defects in CF patients after ETI therapy. • Significant post-treatment correlations of 3D PREFUL ventilation measures especially with LCI, FEV1 %pred, and global MRI score suggest that 3D PREFUL MRI is sensitive to measure improved regional ventilation of the lung parenchyma due to reduced inflammation induced by ETI therapy in CF patients. • 3D PREFUL MRI-derived improved ventilation volume (IVV) correlated with MRI mucus plugging score changes suggesting that reduced endobronchial mucus is predominantly responsible for regional ventilation improvement 8-16 weeks after ETI therapy.

Whole tumor based texture analysis of magnetic resonance diffusion imaging for colorectal liver metastases: A prospective study for diffusion model comparison and early response biomarker.

PURPOSE: To evaluate and compare the diagnostic value of diffusion-related texture analysis parameters obtained from various magnetic resonance diffusion models as early predictors of the clinical response to chemotherapy in patients with colorectal liver metastases (CRLM). METHODS: Patients (n = 145) with CRLM were prospectively and consecutively enrolled and scanned using diffusion-weighted imaging (DWI)-magnetic resonance imaging (MRI)/intravoxel incoherent motion (IVIM)/diffusion kurtosis imaging (DKI) before (baseline) and two-three weeks after (follow-up) commencing chemotherapy. Therapy response was evaluated based on the Response Evaluation Criteria in Solid Tumors (RECIST, version 1.1). The histogram and texture parameters of each diffusion-related parametric map were analysed between the responding and non-responding groups, screened using LASSO, and fitted with binary logistic regression models. The diagnostic efficacy of each model in the early prediction of CRLM was analysed, and the corresponding receiver operating characteristic (ROC) curve was drawn. The area under the curve (AUC) and 95% confidence intervals (CI) were calculated. RESULTS: Of the 145 analysed patients, 69 were in the responding group and 76 were in the non-responding group. Among all models, the difference value based on the histogram and texture features of the DKI-derived parameters performed best for the early prediction of CRLM treatment efficacy. The AUC of the DKI model in the validation set reached 0.795 (95% CI 0.652-0.938). Among the IVIM-derived parameters, the difference model based on D and D* performed best, and the AUC in the validation set reached 0.737 (95% CI 0.586-0.889). Finally, in the DWI sequence, the model comprising baseline features performed the best, with an AUC of 0.699 (95% CI 0.537-0.86) in the validation set. CONCLUSIONS: Baseline DWI parameters and follow-up changes in IVIM and DKI parameters predicted the chemotherapeutic response in patients with CRLM. In addition, as very early predictors, DKI-derived parameters were more effective than DWI- and IVIM-related parameters, in which changes in D-parameters performed best.

Intravoxel incoherent motion imaging to assess the acute effects of moderate-intensity continuous training and high-intensity interval training on thigh muscles.

This study investigated the use of intravoxel incoherent motion imaging (IVIM) to compare skeletal muscle perfusion during and after high-intensity interval training (HIIT) and moderate-intensity continuous training (MICT) to determine the impact on fat oxidation outcomes. Twenty overweight volunteers were recruited for the study. Each participant received one HIIT intervention and one MICT intervention using a cycling ergometer. Participants underwent a magnetic resonance imaging scan before, immediately after, and 1 and 2 h after each intervention. The IVIM parameters (D, fD*) of the rectus femoris, vastus lateralis, and biceps femoris long head were obtained. Changes in IVIM parameters of these muscles after both exercise interventions were compared using a two-factor repeated measures analysis of variance. In the rectus femoris, the fD* increased immediately after exercise intervention (d = 0.69 × 10-3  mm2 /s, p < 0.0083) and 2 h after exercise intervention (d = 0.64 × 10-3  mm2 /s, p < 0.0083) compared with before exercise. The increase in the fD* in the HIIT group was greater than that in the MICT group (d = 0.32, p = 0.023). In the vastus lateralis, the fD* increased immediately after the exercise intervention (d = 0.53 × 10-3  mm2 /s, p < 0.001) and returned to the pre-exercise level 1 h after exercising. The increase in the fD* in the HIIT group was lower than that in the MICT group (d = -0.21, p = 0.015). For the biceps femoris long head, the fD* was not significantly different between the two exercise interventions before and after exercise. Furthermore, the fD* 60 min after the HIIT intervention correlated with maximal oxygen consumption (VO2max), whereas fD* immediately after the MICT intervention correlated with VO2max. In summary, IVIM parameters can be used to evaluate differences in muscle perfusion between HIIT and MICT, and show a correlation with VO2max.

Diagnostic accuracy of perfusion-weighted phase-resolved functional lung magnetic resonance imaging in patients with chronic pulmonary embolism.

Purpose: This study aimed to evaluate the diagnostic performance of perfusion-weighted phase-resolved functional lung (PW-PREFUL) magnetic resonance imaging (MRI) in patients with chronic pulmonary embolism (CPE). Materials and methods: This study included 86 patients with suspected chronic thromboembolic pulmonary hypertension (CTEPH), who underwent PREFUL MRI and ventilation/perfusion (V/Q) single-photon emission computed tomography/computed tomography (SPECT/CT). PREFUL MRI was performed at 1.5 T using a balanced steady-state free precession sequence during free breathing. Color-coded PW images and quantitative parameters were obtained by postprocessing. Meanwhile, V/Q SPECT/CT imaging was performed as a reference standard. Hypoperfused areas in the lungs were scored for each lobe and segment using V/Q SPECT/CT images and PW-PREFUL MR images, respectively. Normalized perfusion (QN) and perfusion defect percentage (QDP) were calculated for all slices. For intra- and interobserver variability, the MRI images were analyzed 2 months after the first analysis by the same radiologist and another radiologist (11 years of lung MRI experience) blinded to the results of the first reader. Results: Of the 86 enrolled patients, 77 met the inclusion criteria (36 diagnosed with CPE using V/Q SPECT/CT and 41 diagnosed with non-CPE etiology). For the PW-PREFUL MRI, the sensitivity, specificity, accuracy, and positive and negative predictive values for the diagnosis of CPE were 97, 95, 96, 95, and 98% at the patient level; 91, 94, 93, 91, and 94% at the lobe level, and 85, 94, 92, 88, and 94% at the segment level, respectively. The detection of segmental and subsegmental hypoperfusion using PW-PREFUL MRI revealed a moderate agreement with V/Q SPECT/CT (κ = 0.65; 95% confidence interval: 0.61-0.68). The quantitative results indicated that the QN was lower in the CPE group than in the non-CPE group [median score (interquartile range, IQR) 6.3 (2.8-9.2) vs. 13.0 (8.8-16.7), p < 0.001], and the QDP was higher [median score (IQR) 33.8 (15.7-51.7) vs. 2.2 (1.4-2.9), p < 0.001]. Conclusion: PREFUL MRI could be an alternative test to detect CPE without requiring breath-hold, contrast agents, or ionizing radiation.

Influence of gadolinium, field-strength and sequence type on quantified perfusion values in phase-resolved functional lung MRI.

PURPOSE: The purpose of this study is to evaluate the influences of gadolinium-based contrast agents, field-strength and different sequences on perfusion quantification in Phase-Resolved Functional Lung (PREFUL) MRI. MATERIALS AND METHODS: Four cohorts of different subjects were imaged to analyze influences on the quantified perfusion maps: 1) at baseline and after 2 weeks to obtain the reproducibility (26 COPD patients), 2) before and after the administration of gadobutrol (11 COPD, 2 PAH and 1 asthma), 3) at 1.5T and 3T (12 healthy, 4 CF), and 4) with different acquisition sequences spoiled gradient echo (SPGR) and balanced steady-state free precession (bSSFP) (11 COPD, 7 healthy). Wilcoxon-signed rank test, Bland-Altman plots, voxelwise Pearson correlations, normalized histogram analyses with skewness and kurtosis and two-sample Kolmogorov-Smirnov tests were performed. P value ≤ 0.05 was considered statistically significant. RESULTS: In all cohorts, linear correlations of the perfusion values were significant with correlation coefficients of at least 0.7 considering the entire lung (P<0.01). The reproducibility cohort revealed stable results with a similar distribution. In the gadolinium cohort, the quantified perfusion increased significantly (P<0.01), and no significant change was detected in the histogram analysis. In the field-strength cohort, no significant change of the quantified perfusion was shown, but a significant increase of skewness and kurtosis at 3T (P = 0.01). In the sequence cohort, the quantified perfusion decreased significantly in the bSSFP sequence (P<0.01) together with a significant decrease of skewness and kurtosis (P = 0.02). The field-strength and sequence cohorts had differing probability distribution in the two-sample Kolmogorov-Smirnov tests. CONCLUSION: We observed a high susceptibility of perfusion quantification to gadolinium, field-strength or MRI sequence leading to distortion and deviation of the perfusion values. Future multicenter studies should strictly adhere to the identical study protocols to generate comparable results.

T2*-corrected Q-Dixon and reduced-FOV diffusion kurtosis imaging (DKI) parameters: correlation with QCT-derived bone mineral density (BMD) and ability to identify abnormal BMD and osteoporosis in postmenopausal women.

Background: Bone marrow fat increases when the bone volume decreases. The composition of the bone marrow microenvironment can also become altered. Assessments of bone marrow fat and bone marrow structural heterogeneity have the potential to predict abnormal bone mineral density (BMD) and osteoporosis. This study aimed to investigate the diagnostic performance of T2*-corrected Q-Dixon and reduced-field-of-view (FOV) diffusion kurtosis imaging (DKI) parameters in determining abnormal BMD and osteoporosis in postmenopausal women. Methods: In this prospective study, the individuals who were eligible for inclusion included postmenopausal women (over 50-year-old) with suspected osteoporosis based on experiencing low back pain. This mono-center study was conducted in tertiary care in China. All of the patients were recruited by using the consecutive sampling method. Subjects who underwent T2*-corrected Q-Dixon and reduced-FOV DKI sequences were enrolled. Fat fraction (FF), T2*, mean kurtosis (MK), and mean diffusivity (MD) values were measured on L1, L2, and L3 vertebral bodies. Quantitative computed tomography (QCT) examinations served as the reference standard. All of the subjects were divided into three groups: normal (BMD >120 mg/cm3), osteopenia (BMD 80-120 mg/cm3), and osteoporosis (BMD <80 mg/cm3). One-way analysis of variance, correlation coefficient analysis, and receiver operating characteristic curve analysis were performed. Results: Among all of the enrolled subjects, 52 were in the normal group, 51 were in the osteopenia group, and 52 were in the osteoporosis group. There were significant differences in FF, T2*, MK, and MD values between the three groups (P<0.001, P<0.001, P<0.001, and P=0.003, respectively). FF, T2*, and MK values exhibited significant negative correlations with BMD values (r=-0.739, P<0.001; r=-0,676, P<0.001; and r=-0.626, P<0.001, respectively). Excellent discriminatory capacity was observed in the Q-Dixon [area under the curve (AUC): 0.976, 95% confidence interval (CI): 0.955-0.997] differentiation between normal and abnormal BMD subjects. It was significantly better than the DKI (AUC: 0.812, 95% CI: 0.741-0.882) parameter combination (P<0.001), whereas the DKI model (AUC: 0.825, 95% CI: 0.739-0.910) performed comparably to the Q-Dixon model (AUC: 0.798, 95% CI: 0.710-0.886) for screening osteoporosis (P=0.57). Conclusions: FF and T2* values measured by using T2*-corrected Q-Dixon, as well as MK and MD values measured by using reduced-FOV DKI, may serve as potential imaging biomarkers in assessing abnormal BMD and osteoporosis in postmenopausal women.

Ultrafast DCE-MRI for discriminating pregnancy-associated breast cancer lesions from lactation related background parenchymal enhancement.

OBJECTIVE: To investigate the utility of ultrafast dynamic-contrast-enhanced (DCE) MRI in visualization and quantitative characterization of pregnancy-associated breast cancer (PABC) and its differentiation from background-parenchymal-enhancement (BPE) among lactating patients. MATERIALS AND METHODS: Twenty-nine lactating participants, including 10 PABC patients and 19 healthy controls, were scanned on 3-T MRI using a conventional DCE protocol interleaved with a golden-angle radial sparse parallel (GRASP) ultrafast sequence for the initial phase. The timing of the visualization of PABC lesions was compared to lactational BPE. Contrast-noise ratio (CNR) was compared between the ultrafast and conventional DCE sequences. The differences in each group's ultrafast-derived kinetic parameters including maximal slope (MS), time to enhancement (TTE), and area under the curve (AUC) were statistically examined using the Mann-Whitney test and receiver operator characteristic (ROC) curve analysis. RESULTS: On ultrafast MRI, breast cancer lesions enhanced earlier than BPE (p < 0.0001), enabling breast cancer visualization freed from lactation BPE. A higher CNR was found for ultrafast acquisitions vs. conventional DCE (p < 0.05). Significant differences in AUC, MS, and TTE values were found between the tumor and BPE (p < 0.05), with ROC-derived AUC of 0.86 ± 0.06, 0.82 ± 0.07, and 0.68 ± 0.08, respectively. The BPE grades of the lactating PABC patients were reduced as compared with the healthy lactating controls (p < 0.005). CONCLUSION: Ultrafast DCE MRI allows BPE-free visualization of lesions, improved tumor conspicuity, and kinetic quantification of breast cancer during lactation. Implementation of this method may assist in the utilization of breast MRI for lactating patients. CLINICAL RELEVANCE: The ultrafast sequence appears to be superior to conventional DCE MRI in the challenging evaluation of the lactating breast. Thus, supporting its possible utilization in the setting of high-risk screening during lactation and the diagnostic workup of PABC. KEY POINTS: • Differences in the enhancement slope of cancer relative to BPE allowed the optimal visualization of PABC lesions on mid-acquisitions of ultrafast DCE, in which the tumor enhanced prior to the background parenchyma. • The conspicuity of PABC lesions on top of the lactation-related BPE was increased using an ultrafast sequence as compared with conventional DCE MRI. • Ultrafast-derived maps provided further characterization and parametric contrast between PABC lesions and lactation-related BPE.

Prediction of upgrade to clinically significant prostate cancer in patients under active surveillance: Performance of a fully automated AI-algorithm for lesion detection and classification.

BACKGROUND: Multiparametric MRI (mpMRI) improves the detection of aggressive prostate cancer (PCa) subtypes. As cases of active surveillance (AS) increase and tumor progression triggers definitive treatment, we evaluated whether an AI-driven algorithm can detect clinically significant PCa (csPCa) in patients under AS. METHODS: Consecutive patients under AS who received mpMRI (PI-RADSv2.1 protocol) and subsequent MR-guided ultrasound fusion (targeted and extensive systematic) biopsy between 2017 and 2020 were retrospectively analyzed. Diagnostic performance of an automated clinically certified AI-driven algorithm was evaluated on both lesion and patient level regarding the detection of csPCa. RESULTS: Analysis of 56 patients resulted in 93 target lesions. Patient level sensitivity and specificity of the AI algorithm was 92.5%/31% for the detection of ISUP ≥ 1 and 96.4%/25% for the detection of ISUP ≥ 2, respectively. The only case of csPCa missed by the AI harbored only 1/47 Gleason 7a core (systematic biopsy; previous and subsequent biopsies rendered non-csPCa). CONCLUSIONS: AI-augmented lesion detection and PI-RADS scoring is a robust tool to detect progression to csPCa in patients under AS. Integration in the clinical workflow can serve as reassurance for the reader and streamline reporting, hence improve efficiency and diagnostic confidence.

Histogram analysis of breast diffusion kurtosis imaging: a comparison between readout-segmented and single-shot echo-planar imaging sequence.

Background: Histogram analysis of the diffusion-weighted imaging (DWI) parameters is widely used to differentiate the breast lesions. However, histogram analysis of the diffusion-kurtosis imaging (DKI) parameters for the single-shot echo-planar imaging (ss-EPI) and readout-segmented echo planar imaging (rs-EPI) sequences has not been compared in breast cancer. Thus, this study is to investigate the diagnostic accuracy and reliability of the histogram parameters derived from the rs-EPI and ss-EPI sequences of DKI parameters in distinguishing between the benign and malignant breast lesions. Methods: This single-center, retrospective cohort study enrolled 205 consecutive patients with breast lesions (65 benign and 140 malignant). The patients underwent breast magnetic resonance imaging (MRI) with a 3T scanner using the rs-EPI and ss-EPI sequences with 4 b values (0, 50, 1,000, and 2,000 s/mm2). The regions of interest (ROIs) were manually delineated for all the lesion images from both the sequences, and the histogram parameters were extracted from the apparent diffusion coefficient (ADC) and apparent diffusional kurtosis (Kapp) maps. Statistical analysis was performed using the Kolmogorov-Smirnov test, the student's t-test, and the receiver operating characteristic (ROC) curves. Results: The mean, 25th, 50th, 75th, and 100th percentiles, skewness, and kurtosis values derived from apparent diffusion for non-Gaussian distribution (Dapp) and Kapp maps showed good or excellent intra-observer agreement (ICC: 0.695 to 0.863).The mean and the 25th, 50th, 75th, and 100th percentile values for Dapp were significantly lower and the mean and the 25th, 50th, 75th, and 100th percentile values for Kapp were significantly higher in the malignant breast lesions compared with those in the benign breast lesions for both the rs-EPI and ss-EPI sequences (all P<0.05). The majority of the histogram Kapp and Dapp parameters (except skewness and kurtosis) for the benign and malignant lesions showed significant differences between the ss-EPI and the rs-EPI sequences (P<0.05). ROC curve analysis showed that the AUC values for the 75th percentile of Kapp (0.854 for rs-EPI, 0.844 for ss-EPI) and the 25th percentile of Dapp (0.866 for rs-EPI, 0.858 for ss-EPI) were highest for both DKI sequences. The diagnostic performance of the rs-EPI sequence was better than the ss-EPI sequence for all the histogram parameters except the skewness value of Dapp. Conclusions: Histogram parameters from the rs-EPI sequence were more reliable and accurate in differentiating malignant and benign breast lesions than those from the ss-EPI sequence.

Role of diffusion kurtosis imaging in evaluating microstructural changes in spinal cord of patients with cervical spondylosis.

STUDY DESIGN: Analytical cross-sectional study. PURPOSE: To study the role of diffusion kurtosis imaging (DKI) in evaluating microstructural changes in patients with cervical spondylosis. OVERVIEW OF LITERATURE: Cervical spondylosis is a common progressive degenerative disorder of the spine. Conventional magnetic resonance imaging (MRI) can only detect the changes in the spinal cord once there are visual signal changes; hence, it underestimates the extent of the injury. Newer imaging techniques like Diffusion Tensor and Kurtosis Imaging can evaluate the microstructural changes in cervical spinal cord before the obvious signal changes appear. METHODS: Conventional MRI, diffusion tensor imaging (DTI), and DKI scans were performed for 90 cervical spondylosis patients on 1.5-T MR Siemens Magnetom aera after obtaining informed consent. Eight patients were excluded due to poor image quality. Fractional anisotropy (FA) colour maps and diffusion kurtosis (DK) maps corresponding to spinal cord cross sections at C2-C3 intervertebral disc level (control) and at the most stenotic levels were obtained. Modified Japanese Orthopaedic Association (mJOA) scoring was used for clinical assessment of the spinal cord function. The changes in DTI and DKI parameters and their correlation with mJOA scores were analysed by SPSS 23 software. RESULTS: In our study, mean FA and mean kurtosis (MK) values at the stenotic level (0.54, 1.02) were significantly lower than values at the non-stenotic segment (0.70, 1.27). The mean diffusivity (MD) value at the stenotic segment (1.25) was significantly higher than in the non-stenotic segment (1.09). We also observed a strong positive correlation between mJOA score and FA and MK values and a negative correlation between mJOA score and MD values, suggesting a correlation of FA, MK, and MD with the clinical severity of the disease. CONCLUSION: Addition of DTI and DKI sequences helps in early identification of the disease without any additional cost incurred by the patient.

Automated deep-learning system in the assessment of MRI-visible prostate cancer: comparison of advanced zoomed diffusion-weighted imaging and conventional technique.

BACKGROUND: Deep-learning-based computer-aided diagnosis (DL-CAD) systems using MRI for prostate cancer (PCa) detection have demonstrated good performance. Nevertheless, DL-CAD systems are vulnerable to high heterogeneities in DWI, which can interfere with DL-CAD assessments and impair performance. This study aims to compare PCa detection of DL-CAD between zoomed-field-of-view echo-planar DWI (z-DWI) and full-field-of-view DWI (f-DWI) and find the risk factors affecting DL-CAD diagnostic efficiency. METHODS: This retrospective study enrolled 354 consecutive participants who underwent MRI including T2WI, f-DWI, and z-DWI because of clinically suspected PCa. A DL-CAD was used to compare the performance of f-DWI and z-DWI both on a patient level and lesion level. We used the area under the curve (AUC) of receiver operating characteristics analysis and alternative free-response receiver operating characteristics analysis to compare the performances of DL-CAD using f- DWI and z-DWI. The risk factors affecting the DL-CAD were analyzed using logistic regression analyses. P values less than 0.05 were considered statistically significant. RESULTS: DL-CAD with z-DWI had a significantly better overall accuracy than that with f-DWI both on patient level and lesion level (AUCpatient: 0.89 vs. 0.86; AUClesion: 0.86 vs. 0.76; P < .001). The contrast-to-noise ratio (CNR) of lesions in DWI was an independent risk factor of false positives (odds ratio [OR] = 1.12; P < .001). Rectal susceptibility artifacts, lesion diameter, and apparent diffusion coefficients (ADC) were independent risk factors of both false positives (ORrectal susceptibility artifact = 5.46; ORdiameter, = 1.12; ORADC = 0.998; all P < .001) and false negatives (ORrectal susceptibility artifact = 3.31; ORdiameter = 0.82; ORADC = 1.007; all P ≤ .03) of DL-CAD. CONCLUSIONS: Z-DWI has potential to improve the detection performance of a prostate MRI based DL-CAD. TRIAL REGISTRATION: ChiCTR, NO. ChiCTR2100041834 . Registered 7 January 2021.

Comparison of a Deep Learning-Accelerated vs. Conventional T2-Weighted Sequence in Biparametric MRI of the Prostate.

BACKGROUND: Demand for prostate MRI is increasing, but scan times remain long even in abbreviated biparametric MRIs (bpMRI). Deep learning can be leveraged to accelerate T2-weighted imaging (T2WI). PURPOSE: To compare conventional bpMRIs (CL-bpMRI) with bpMRIs including a deep learning-accelerated T2WI (DL-bpMRI) in diagnosing prostate cancer. STUDY TYPE: Retrospective. POPULATION: Eighty consecutive men, mean age 66 years (47-84) with suspected prostate cancer or prostate cancer on active surveillance who had a prostate MRI from December 28, 2020 to April 28, 2021 were included. Follow-up included prostate biopsy or stability of prostate-specific antigen (PSA) for 1 year. FIELD STRENGTH AND SEQUENCES: A 3 T MRI. Conventional axial and coronal T2 turbo spin echo (CL-T2), 3-fold deep learning-accelerated axial and coronal T2-weighted sequence (DL-T2), diffusion weighted imaging (DWI) with b = 50 sec/mm2 , 1000 sec/mm2 , calculated b = 1500 sec/mm2 . ASSESSMENT: CL-bpMRI and DL-bpMRI including the same conventional diffusion-weighted imaging (DWI) were presented to three radiologists (blinded to acquisition method) and to a deep learning computer-assisted detection algorithm (DL-CAD). The readers evaluated image quality using a 4-point Likert scale (1 = nondiagnostic, 4 = excellent) and graded lesions using Prostate Imaging Reporting and Data System (PI-RADS) v2.1. DL-CAD identified and assigned lesions of PI-RADS 3 or greater. STATISTICAL TESTS: Quality metrics were compared using Wilcoxon signed rank test, and area under the receiver operating characteristic curve (AUC) were compared using Delong's test. SIGNIFICANCE: P = 0.05. RESULTS: Eighty men were included (age: 66 ± 9 years; 17/80 clinically significant prostate cancer). Overall image quality results by the three readers (CL-T2, DL-T2) are reader 1: 3.72 ± 0.53, 3.89 ± 0.39 (P = 0.99); reader 2: 3.33 ± 0.82, 3.31 ± 0.74 (P = 0.49); reader 3: 3.67 ± 0.63, 3.51 ± 0.62. In the patient-based analysis, the reader results of AUC are (CL-bpMRI, DL-bpMRI): reader 1: 0.77, 0.78 (P = 0.98), reader 2: 0.65, 0.66 (P = 0.99), reader 3: 0.57, 0.60 (P = 0.52). Diagnostic statistics from DL-CAD (CL-bpMRI, DL-bpMRI) are sensitivity (0.71, 0.71, P = 1.00), specificity (0.59, 0.44, P = 0.05), positive predictive value (0.23, 0.24, P = 0.25), negative predictive value (0.88, 0.88, P = 0.48). CONCLUSION: Deep learning-accelerated T2-weighted imaging may potentially be used to decrease acquisition time for bpMRI. EVIDENCE LEVEL: 3. TECHNICAL EFFICACY: Stage 2.

Capability of arterial spin labeling and intravoxel incoherent motion diffusion-weighted imaging to detect early kidney injury in chronic kidney disease.

OBJECTIVES: To prospectively investigate the capability of arterial spin labeling (ASL) and intravoxel incoherent motion diffusion-weighted imaging (IVIM-DWI) for the identification of early kidney injury in chronic kidney disease (CKD) patients with normal estimated glomerular filtration rate (eGFR). METHODS: Fifty-four CKD patients confirmed by renal biopsy (normal eGFR group [eGFR ≥ 90 mL/min/1.73 m2]: n = 26; abnormal eGFR group [eGFR < 90 mL/min/1.73 m2]: n = 28) and 20 healthy volunteers (HV) were recruited. All subjects were examined by IVIM-DWI and ASL imaging. Renal blood flow (RBF) derived from ASL, true diffusion coefficient (D), pseudo-diffusion coefficient (D*), and perfusion fraction (f) derived from IVIM-DWI were measured from the renal cortex. One-way analysis of variance was used to compare MRI parameters among the three groups. The correlation between eGFR and MRI parameters was evaluated by Spearman correlation analysis. Diagnostic performances of MRI parameters for detecting kidney injury were assessed by receiver operating characteristic (ROC) curves. RESULTS: The renal cortical D, D*, f, and RBF values showed statistically significant differences among the three groups. eGFR was positively correlated with MRI parameters (D: r = 0.299, D*: r = 0.569, f: r = 0.733, RBF: r = 0.586). The areas under the curve (AUCs) for discriminating CKD patients from HV were 0.725, 0.752, 0.947, and 0.884 by D, D*, f, and RBF, respectively. D, D*, f, RBF, and eGFR identified CKD patients with normal eGFR with AUCs of 0.735, 0.612, 0.917, 0.827, and 0.733, respectively, and AUC of f value was significantly larger than that of eGFR. CONCLUSION: IVIM-DWI and ASL were useful for detecting underlying pathologic injury in early CKD patients with normal eGFR. KEY POINTS: • The renal cortical f and RBF values in the control group were significantly higher than those in the normal eGFR group. • A negative correlation was observed between the renal cortical D, D*, f, and RBF values and SCr and 24 h-UPRO, while eGFR was significantly positively correlated with renal cortical D, D*, f, and RBF values. • The AUC of renal cortical f values was statistically larger than that of eGFR for the discrimination between the CKD with normal eGFR group and the control group.

Evaluation of functional and metabolic tumor volume using voxel-wise analysis in childhood rhabdomyosarcoma.

BACKGROUND: Cross-sectional imaging-based morphological characteristics of pediatric rhabdomyosarcoma have failed to predict outcomes. OBJECTIVE: To evaluate the feasibility and possible value of generating tumor sub-volumes using voxel-wise analysis of metabolic and functional data from positron emission tomography/magnetic resonance imaging (PET/MR) or PET/computed tomography (CT) and MRI in rhabdomyosarcoma. MATERIALS AND METHODS: Thirty-four examinations in 17 patients who received PET/MRI or PET/CT plus MRI were analyzed. The volume of interest included total tumor volume before and after therapy. Apparent diffusion coefficients (ADC) and standard uptake values (SUV) were determined voxel-wise. Voxels were assigned to three different groups based on ADC and SUV: "viable tumor tissue," "intermediate tissue" or "possible necrosis." In a second approach, data were grouped into three clusters using the Gaussian mixture model. The ratio of these clusters to total tumor volume and changes due to chemotherapy were correlated with clinical and histopathological data. RESULTS: After chemotherapy, the proportion of voxels in the different groups changed significantly. A significant reduction of the proportion of voxels assigned to cluster 1 was found, from a mean of 36.4% to 2.5% (P < 0.001). There was a significant increase in the proportion of voxels in cluster 3 following chemotherapy from 24.8% to 81.6% (P = 0.02). The proportion of voxels in cluster 2 differed depending on the presence or absence of tumor recurrence, falling from 48% to 10% post-chemotherapy in the group with no tumor recurrence (P < 0.05) and from 29% to 23% (P > 0.05) in the group with tumor recurrence. CONCLUSION: Voxel-wise evaluation of multimodal data in rhabdomyosarcoma is feasible. Our initial results suggest that the different distribution of sub-volumes before and after therapy may have prognostic significance.

Inter- and intravisit repeatability of free-breathing MRI in pediatric cystic fibrosis lung disease.

PURPOSE: The purpose of this study is to assess the intra- and interscan repeatability of free-breathing phase-resolved functional lung (PREFUL) MRI in stable pediatric cystic fibrosis (CF) lung disease in comparison to static breath-hold hyperpolarized 129-xenon MRI (Xe-MRI) and pulmonary function tests. METHODS: Free-breathing 1-hydrogen MRI and Xe-MRI were acquired from 15 stable pediatric CF patients and seven healthy age-matched participants on two visits, 1 month apart. Same-visit MRI scans were also performed on a subgroup of the CF patients. Following the PREFUL algorithm, regional ventilation (RVent) and regional flow volume loop cross-correlation maps were determined from the free-breathing data. Ventilation defect percentage (VDP) was determined from RVent maps (VDPRVent ), regional flow volume loop cross-correlation maps (VDPCC ), VDPRVent ∪ VDPCC , and multi-slice Xe-MRI. Repeatability was evaluated using Bland-Altman analysis, coefficient of repeatability (CR), and intraclass correlation. RESULTS: Minimal bias and no significant differences were reported for all PREFUL MRI and Xe-MRI VDP parameters between intra- and intervisits (all P > 0.05). Repeatability of VDPRVent , VDPCC , VDPRVent ∪ VDPCC , and multi-slice Xe-MRI were lower between the two-visit scans (CR = 14.81%, 15.36%, 16.19%, and 9.32%, respectively) in comparison to the same-day scans (CR = 3.38%, 2.90%, 1.90%, and 3.92%, respectively). pulmonary function tests showed high interscan repeatability relative to PREFUL MRI and Xe-MRI. CONCLUSION: PREFUL MRI, similar to Xe-MRI, showed high intravisit repeatability but moderate intervisit repeatability in CF, which may be due to inherent disease instability, even in stable patients. Thus, PREFUL MRI may be considered a suitable outcome measure for future treatment response studies.