Feasibility of deep learning k-space-to-image reconstruction for diffusion weighted imaging in patients with breast cancers: Focus on image quality and reduced scan time.

PURPOSE: This study aimed to evaluate the feasibility of accelerated DLR (deep learning reconstruction) single-shot echo planar imaging (ss-EPI) for diffusion-weighted image (DWI) in patients with breast cancers in comparison to conventional ss-EPI. METHODS: Between August 2021 and February 2022, eighty-seven patients with pathologically proven breast cancer underwent DCE breast MRI including ss-EPI and DLR ss-EPI DWI sequences (TA, 3:36 min and 1:54 min, respectively) at 3 Tesla. In a randomized and blinded manner, two radiologists independently performed qualitative analyses for overall image quality using a 5-point scale of the following components: homogeneous fat suppression, image blurring, artifact, and lesion conspicuity. Quantitative analyses were performed by measurement of ADC values, SNR, CNR, and lesion contrast. RESULTS: DLR ss-EPI showed better image quality scores, CNR, and lesion contrast than ss-EPI (all P < 0.05) while reducing scan time by 47.2 %. DLR ss-EPI showed no significant difference in SNR and tumor ADC values compared to -ss-EPI (P = 0.307 and P = 0.123, respectively). CONCLUSIONS: DLR ss-EPI showed better results in the qualitative and quantitative analysis than conventional ss-EPI despite reducing scan time by 47.2%.

Image quality and whole-lesion histogram and texture analysis of diffusion-weighted imaging of breast MRI based on advanced ZOOMit and simultaneous multislice readout-segmented echo-planar imaging.

Objectives: To investigate the image quality and diagnostic capability a of whole-lesion histogram and texture analysis of advanced ZOOMit (A-ZOOMit) and simultaneous multislice readout-segmented echo-planar imaging (SMS-RS-EPI) to differentiate benign from malignant breast lesions. Study design: From February 2020 to October 2020, diffusion-weighted imaging (DWI) using SMS-RS-EPI and A-ZOOMit were performed on 167 patients. Three breast radiologists independently ranked the image datasets. The inter-/intracorrelation coefficients (ICCs) of mean image quality scores and lesion conspicuity scores were calculated between these three readers. Histogram and texture features were extracted from the apparent diffusion coefficient (ADC) maps, respectively, based on a WL analysis. Student's t-tests, one-way ANOVAs, Mann-Whitney U tests, and receiver operating characteristic curves were used for statistical analysis. Results: The overall image quality scores and lesion conspicuity scores for A-ZOOMit and SMS-RS-EPI showed statistically significant differences (4.92 ± 0.27 vs. 3.92 ± 0.42 and 4.93 ± 0.29 vs. 3.87 ± 0.47, p < 0.0001). The ICCs for the image quality and lesion conspicuity scores had good agreements among the three readers (all ICCs >0.75). To differentiate benign and malignant breast lesions, the entropy of ADCA-Zoomit had the highest area (0.78) under the ROC curve. Conclusions: A-ZOOMit achieved higher image quality and lesion conspicuity than SMS-RS-EPI. Entropy based on A-ZOOMit is recommended for differentiating benign from malignant breast lesions.

Modified Reduced Field-of-View Diffusion-Weighted Magnetic Resonance Imaging of the Prostate: Comparison With Reduced Field-of-View Imaging and Single Shot Echo-Planar Imaging.

OBJECTIVE: The objective of this study was to compare the image quality and apparent diffusion coefficient (ADC) of diffusion-weighted imaging (DWI) with modified reduced field of view (FOV) based on 2-dimensional (2D)-selective radiofrequency excitations by tilting the excitation plane in prostate with reduced FOV using parallel-transmit-accelerated 2D-selective radiofrequency excitation and single-shot echo planar imaging (ssEPI). METHODS: Fifty patients who underwent multiparametric magnetic resonance imaging including 3 DWIs were included. Two observers independently performed qualitative image analyses using 5-point scale. Apparent diffusion coefficient measurements were performed for quantitative analysis. RESULTS: Modified reduced FOV provided the highest qualitative scores for all categories compared with reduced FOV and ssEPI (P < 0.000). Both reduced FOV DWIs showed higher ADC values compared with ssEPI (P < 0.001); however, the ADC ratios between the lesion and peripheral zone were not significantly different (all P > 0.05). CONCLUSIONS: The modified reduced FOV DWI showed better overall image quality, differentiability of anatomic regions, and lesion conspicuity with fewer artifacts compared with DWI with reduced FOV and ssEPI.