Dynamic Contrast-enhanced MRI Sequences Research Articles

Breast cancer classification through multivariate radiomic time series analysis in DCE-MRI sequences

Breast cancer is the most prevalent disease that poses a significant threat to women’s health. Despite the Dynamic Contrast-Enhanced MRI (DCE-MRI) has been widely used for breast cancer classification, its diagnostic performance is still suboptimal. In this work, the Radiomic workflow was implemented to classify the whole DCE-MRI sequence based on the distinction in contrast agent uptake between benign and malignant lesions. The radiomic features extracted from each of the seven time instants within the DCE-MRI sequence were fed into a multi-instant features selection strategy to select the discriminative features for time series classification. Several time series classification algorithms including Rocket, MultiRocket, K-Nearest Neighbor, Time Series Forest, and Supervised Time Series Forest were compared. Firstly, a univariate classification was performed to find the five most informative radiomic series, and then, a multivariate time series classification was implemented via a voting mechanism. The Multivariate Rocket model was the most accurate (Accuracy = 0.852, AUC-ROC = 0.852, Specificity = 0.823, Sensitivity = 0.882). The intelligible radiomic features enabled model findings explanations and clinical validation. In particular, the Energy and TotalEnergy were among the most important features, and the most descriptive for the change in signal intensity, which is the main effect of the contrast agent.

Understanding Spatial Correlation Between Multiparametric MRI Performance and Prostate Cancer.

Multiparametric MRI (mpMRI) has shown a substantial impact on prostate cancer (PCa) diagnosis. However, the understanding of the spatial correlation between mpMRI performance and PCa location is still limited. To investigate the association between mpMRI performance and tumor spatial location within the prostate using a prostate sector map, described by Prostate Imaging Reporting and Data System (PI-RADS) v2.1. Retrospective. One thousand one hundred forty-three men who underwent mpMRI before radical prostatectomy between 2010 and 2022. 3.0 T. T2-weighted turbo spin-echo, a single-shot spin-echo EPI sequence for diffusion-weighted imaging, and a gradient echo sequence for dynamic contrast-enhanced MRI sequences. Integrated relative cancer prevalence (rCP), detection rate (DR), and positive predictive value (PPV) maps corresponding to the prostate sector map for PCa lesions were created. The relationship between tumor location and its detection/missing by radiologists on mpMRI compared to WMHP as a reference standard was investigated. A weighted chi-square test was performed to examine the statistical differences for rCP, DR, and PPV of the aggregated sectors within the zone, anterior/posterior, left/right prostate, and different levels of the prostate with a statistically significant level of 0.05. A total of 1665 PCa lesions were identified in 1143 patients, and from those 1060 lesions were clinically significant (cs)PCa tumors (any Gleason score [GS] ≥7). Our sector-based analysis utilizing weighted chi-square tests suggested that the left posterior part of PZ had a high likelihood of missing csPCa lesions at a DR of 67.0%. Aggregated sector analysis indicated that the anterior or apex locations in PZ had the significantly lowest csPCa detection at 67.3% and 71.5%, respectively. Spatial characteristics of the per-lesion-based mpMRI performance for diagnosis of PCa were studied. Our results demonstrated that there is a spatial correlation between mpMRI performance and locations of PCa on the prostate. 4 TECHNICAL EFFICACY: Stage 2.

Radiomics Nomogram Based on Dual-Sequence MRI for Assessing Ki-67 Expression in Breast Cancer.

Radiomics has been extensively applied in predicting Ki-67 in breast cancer (BC). However, this is often confined to the exploration of a single sequence, without considering the varying sensitivity and specificity among different sequences. To develop a nomogram based on dual-sequence MRI derived radiomic features combined with clinical characteristics for assessing Ki-67 expression in BC. Retrospective. 227 females (average age, 51 years) with 233 lesions and pathologically confirmed BC, which were divided into the training set (n = 163) and test set (n = 70). 3.0-T, T1-weighted dynamic contrast-enhanced MRI (DCE-MRI) and apparent diffusion coefficient (ADC) maps from diffusion-weighted MRI (EPI sequence). The regions of interest were manually delineated on ADC and DCE-MRI sequences. Three radiomics models of ADC, DCE-MRI, and dsMRI (combined ADC and DCE-MRI sequences) were constructed by logistic regression and the radiomics score (Radscore) of the best model was calculated. The correlation between Ki-67 expression and clinical characteristics such as receptor status, axillary lymph node (ALN) metastasis status, ADC value, and time signal intensity curve was analyzed, and the clinical model was established. The Radscore was combined with clinical predictors to construct a nomogram. The independent sample t-test, Mann-Whitney U test, Chi-squared test, Interclass correlation coefficients (ICCs), single factor analysis, least absolute shrinkage and selection operator (LASSO), logistic regression, receiver operating characteristics, Delong test, Hosmer_Lemeshow test, calibration curve, decision curve. A P-value <0.05 was considered statistically significant. In the test set, the prediction efficiency of the dsMRI model (AUC = 0.862) was higher than ADC model (AUC = 0.797) and DCE-MRI model (AUC = 0.755). With the inclusion of estrogen receptor (ER) and ALN metastasis, the nomogram displayed quality improvement (AUC = 0.876), which was superior to the clinical model (AUC = 0.787) and radiomics model. The nomogram based on dsMRI radiomic features and clinical characteristics may be able to assess Ki-67 expression in BC. 3 TECHNICAL EFFICACY: Stage 3.

Role of dynamic contrast-enhanced MRI in predicting severe acute radiation-induced rectal injury in patients with rectal cancer.

To explore the potential of dynamic contrast-enhanced MRI (DCE-MRI) quantitative parameters in predicting severe acute radiation-induced rectal injury (RRI) in rectal cancer. This retrospective study enrolled 49 patients with rectal cancer who underwent neoadjuvant chemoradiotherapy and rectal MRI including a DCE-MRI sequence from November 2014 to March 2021. Two radiologists independently measured DCE-MRI quantitative parameters, including the forward volume transfer constant (Ktrans), rate constant (kep), fractional extravascular extracellular space volume (ve), and the thickness of the rectal wall farthest away from the tumor. These parameters were compared between mild and severe acute RRI groups based on histopathological assessment. Receiver operating characteristic curve analysis was performed to analyze statistically significant parameters. Forty-nine patients (mean age, 54years ± 12 [standard deviation]; 37 men) were enrolled, including 25 patients with severe acute RRI. Ktrans was lower in severe acute RRI group than mild acute RRI group (0.032min-1 vs 0.054min-1; p = 0.008), but difference of other parameters (kep, ve and rectal wall thickness) was not significant between these two groups (all p > 0.05). The area under the receiver operating characteristic curve of Ktrans was 0.72 (95% confidence interval: 0.57, 0.84). With a Ktrans cutoff value of 0.047min-1, the sensitivity and specificity for severe acute RRI prediction were 80% and 54%, respectively. Ktrans demonstrated moderate diagnostic performance in predicting severe acute RRI. Dynamic contrast-enhanced MRI can provide non-invasive and objective evidence for perioperative management and treatment strategies in rectal cancer patients with acute radiation-induced rectal injury. • To our knowledge, this study is the first to evaluate the predictive value of contrast-enhanced MRI (DCE-MRI) quantitative parameters for severe acute radiation-induced rectal injury (RRI) in patients with rectal cancer. • Forward volume transfer constant (Ktrans), derived from DCE-MRI, exhibited moderate diagnostic performance (AUC = 0.72) in predicting severe acute RRI of rectal cancer, with a sensitivity of 80% and specificity of 54%. • DCE-MRI is a promising imaging marker for distinguishing the severity of acute RRI in patients with rectal cancer.

Breast Cancer Surrogate Subtype Classification Using Pretreatment Multi-Phase Dynamic Contrast-Enhanced Magnetic Resonance Imaging Radiomics: A Retrospective Single-Center Study.

This study aimed to explore the potential of multi-phase dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) radiomics for classifying breast cancer surrogate subtypes. This retrospective study analyzed 360 breast cancers from 319 patients who underwent pretreatment DCE-MRI between January 2015 and January 2019. The cohort consisted of 33 triple-negative, 26 human epidermal growth factor receptor 2 (HER2)-positive, 109 luminal A-like, 144 luminal B-like HER2-negative, and 48 luminal B-like HER2-positive lesions. A total of 1781 radiomic features were extracted from manually segmented breast cancers in each DCE-MRI sequence. The model was internally validated and selected using ten times repeated five-fold cross-validation on the primary cohort, with further evaluation using a validation cohort. The most successful models were logistic regression models applied to the third post-contrast subtraction images. These models exhibited the highest area under the curve (AUC) for discriminating between luminal A like vs. others (AUC: 0.78), luminal B-like HER2 negative vs. others (AUC: 0.57), luminal B-like HER2 positive vs. others (AUC: 0.60), HER2 positive vs. others (AUC: 0.81), and triple negative vs. others (AUC: 0.83). In conclusion, the radiomic features extracted from multi-phase DCE-MRI are promising for discriminating between breast cancer subtypes. The best-performing models relied on tissue changes observed during the mid-stage of the imaging process.

A Soft-Attention Guidance Stacked neural Network for neoadjuvant chemotherapy’s pathological response diagnosis using breast dynamic contrast-enhanced MRI

Pathological complete response (pCR) to Neoadjuvant chemotherapy (NACT)is a significant clinical indicator for diagnosing patient outcomes and overall survival. However, the current diagnosis of pCR suffers from two limitations: firstly, it requires an invasive biopsy, intensifying the patient’s pain and risk of infection. Secondly, current radiomics-based computer methods inefficiently utilize intrinsic characteristics of medical images. In this paper, we present a Soft-Attention Guidance Stacked neural Network (SAGS-Net) to address these limitations and predict the NACT responses from breast dynamic contrast-enhanced MRI (DCE-MRI). Particularly, We first design a self-adapting feature selection strategy to extract peritumoral features while excluding outside noises accurately. Then, SAGS-Net is built by stacking position-based spatial models that generate discriminative feature representation. Each stacked model inside has its semantic feature branch as a control gate for pCR-related feature selection. The semantic feature branch combined with the residual learning mechanism as a feature selector enhances valuable information and suppresses redundant information, simulating the process of radiologists making a clinical diagnosis based on domain knowledge. Finally, the incremental stacked network architecture assisted with the soft-attention strategy can gradually refine attention-aware features in complex DCE-MRI sequences. Experimental results based on real-world clinical datasets confirmed that the proposed SAGS-Net obtain superior performance with 93% AUC, and it provides a new way that leverage DCE-MRI sequences to predict NACT responses.

Enhanced liver X receptor signalling reduces brain injury and promotes tissue regeneration following experimental intracerebral haemorrhage: roles of microglia/macrophages

BackgroundInflammation-exacerbated secondary brain injury and limited tissue regeneration are barriers to favourable prognosis after intracerebral haemorrhage (ICH). As a regulator of inflammation and lipid metabolism, Liver X receptor (LXR) has...

MRI-Based Breast Cancer Classification and Localization by Multiparametric Feature Extraction and Combination Using Deep Learning.

Deep learning (DL) have been reported feasible in breast MRI. However, the effectiveness of DL method in mpMRI combinations for breast cancer detection has not been well investigated. To implement a DL method for breast cancer classification and detection using feature extraction and combination from multiple sequences. Retrospective. A total of 569 local cases as internal cohort (50.2 ± 11.2 years; 100% female), divided among training (218), validation (73) and testing (278); 125 cases from a public dataset as the external cohort (53.6 ± 11.5 years; 100% female). T1-weighted imaging and dynamic contrast-enhanced MRI (DCE-MRI) with gradient echo sequences, T2-weighted imaging (T2WI) with spin-echo sequences, diffusion-weighted imaging with single-shot echo-planar sequence and at 1.5-T. A convolutional neural network and long short-term memory cascaded network was implemented for lesion classification with histopathology as the ground truth for malignant and benign categories and contralateral breasts as healthy category in internal/external cohorts. BI-RADS categories were assessed by three independent radiologists as comparison, and class activation map was employed for lesion localization in internal cohort. The classification and localization performances were assessed with DCE-MRI and non-DCE sequences, respectively. Sensitivity, specificity, area under the curve (AUC), DeLong test, and Cohen's kappa for lesion classification. Sensitivity and mean squared error for localization. A P-value <0.05 was considered statistically significant. With the optimized mpMRI combinations, the lesion classification achieved an AUC = 0.98/0.91, sensitivity = 0.96/0.83 in the internal/external cohorts, respectively. Without DCE-MRI, the DL-based method was superior to radiologists' readings (AUC 0.96 vs. 0.90). The lesion localization achieved sensitivities of 0.97/0.93 with DCE-MRI/T2WI alone, respectively. The DL method achieved high accuracy for lesion detection in the internal/external cohorts. The classification performance with a contrast agent-free combination is comparable to DCE-MRI alone and the radiologists' reading in AUC and sensitivity. 3. Stage 2.

Intravoxel incoherent motion and dynamic contrast-enhanced magnetic resonance imaging for neoadjuvant chemotherapy response evaluation in patients with osteosarcoma

ObjectivesThis study aims to explore the role of quantitative intravoxel incoherent motion (IVIM) and dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) parameters in characterizing changes in osteosarcoma (OS) patients receiving neoadjuvant chemotherapy (NACT). Material and methodsTwenty-seven patients with histologically proven OS were examined prospectively and divided into good-response group (n = 14) and poor-response group (n = 13). IVIM and DCE-MRI sequences were performed at baseline (pre-NACT) and after three cycles of NACT (post-NACT). Apparent diffusion coefficient (ADC) and IVIM bi-exponential model parameters, including diffusion coefficient (D-Bi), perfusion coefficient (D*-Bi), and perfusion fraction (f-Bi), were evaluated. DCE-MRI parameters, including quantitative parameters (volume transfer constant [Ktrans], elimination rate constant [Kep], and extravascular extracellular space volume ratio [Ve]) and semi-quantitative parameters (initial area under the gadolinium curve [IAUGC] and contrast enhancement rate [CER]), were also measured. ResultsD-Bi, D*-Bi, and f-Bi post-NACT and ΔD-Bi were statistically different between the good- and poor-response groups (Z1 = − 3.348, Z2 = − 2.572, Z3 = − 2.378, t = 2.235, P < 0.05). ADC, f-Bi, Ktrans, IAUGC, Kep, and CER post-NACT were statistically different from those at pre-NACT (P < 0.05). The receiver operating characteristic curve showed that f-Bi post-NACT had the best performance among all parameters, with area under the curve of 0.769, sensitivity of 1, and specificity of 0.538. The correlation analysis showed that the efficacy of NACT was negatively correlated with D-Bi, D*-Bi post-NACT, and ΔD-Bi (r1 = − 0.530, r2 = − 0.411, r3 = − 0.434, P1 = 0.008, P2 = 0.046, P3 = 0.034) and significantly positively correlated with f-Bi post-NACT (r = 0.482, P = 0.017). ConclusionsThe IVIM quantitative parameters D-Bi, D*-Bi, and f-Bi post-NACT and ΔD-Bi could be used as noninvasive imaging biomarkers for early response assessment of NACT in OS.

Whole-tumor histogram analysis of diffusion-weighted imaging and dynamic contrast-enhanced MRI for soft tissue sarcoma: correlation with HIF-1alpha expression.

To investigate the correlation of histogram metrics from diffusion-weighted imaging (DWI) and dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) parameters with HIF-1alpha expression in soft tissue sarcoma (STS). We enrolled 71 patients with STS who underwent 3.0-T MRI, including conventional MRI, DWI, and DCE-MRI sequences. Location, maximum tumor diameter, envelope, T2-weighted tumor heterogeneity, peritumoral edema, peritumoral enhancement, necrosis, tail-like pattern, bone invasion, and vessel/nerve invasion and/or encasement were determined using conventional MRI images. The whole-tumor histogram metrics were calculated on the apparent diffusion coefficient (ADC), Ktrans, Kep, and Ve maps. Independent-samples t test and one-way ANOVA were used for testing the differences between normally distributed categorical data with HIF-1alpha expression. Pearson and Spearman correlations and multiple linear regression analyses were performed to determine the correlations between histogram metrics and HIF-1alpha expression. Survival curves were plotted using the Kaplan-Meier method. Regarding conventional MRI features, only highly heterogeneous on T2-weighted images (55.6 ± 19.9% vs. 45.4 ± 20.5%, p = 0.041) and more than 50% necrotic area (57.3 ± 20.4% vs. 43.9 ± 19.7%, p = 0.002) were prone to indicate STS with higher HIF-1alpha expression. Histogram metrics obtained from ADC (mean, median, 10th, and 25th percentile values), Ktrans (mean, median, 75th, and 90th percentile values), and Kep (90th percentile values) were significantly correlated with HIF-1alpha expression. Multiple linear regression analysis demonstrated that more than 50% necrosis, ADCskewness, Ktrans90th, and grade III were independently associated with HIF-1alpha expression. DWI and DCE-MRI histogram parameters were significantly correlated with HIF-1alpha expression in STS. • DWI and DCE-MRI histogram parameters are correlated with HIF-1alpha expression in STS. • More than 50% necrosis, ADCskewness, Ktrans90th, and grade III were independently associated with HIF-1alpha expression in STS.

Differentiation between hyperintense hepatocellular carcinoma and focal nodular hyperplasia in hepatobiliary phase with multimodal parameters of magnetic resonance imaging

Objective: To differentiate hyperintense hepatocellular carcinoma (HCC) with focal nodular hyperplasia (FNH) in the hepatobiliary phase by MRI multimodal parameters. Methods: A retrospective cross-sectional study method was adopted. Clinical data on 15 cases with hyperintense HCC and 15 cases with FNH in the hepatobiliary phase admitted to the First Affiliated Hospital of the Army Medical University from January 2012 to December 2019 were collected. All patients with solitary lesions who underwent Gd-EOB-DTPA-enhanced MRI examinations were included. Surgically resected specimens were verified by pathological and immunohistochemical examination. HCC and FNH imaging features were analyzed by two radiologists. Results: (1) HCC and FNH apparent diffusion coefficient (ADC) values were 1 205.07±239.65×10-3 mm2/s and 1 434.73±217.6×10-3 mm2/s, respectively, and the SIADC difference was statistically significant (P<0.05) between the two groups. (2) In the dynamic contrast-enhanced MRI sequence, 15 cases of HCC were significantly enhanced in the arterial phase, of which 13 cases were characterized by continuous enhancement, and 2 cases were characterized by wash-in and wash-out enhancement. There was no statistically significant difference (P>0.05) between the two groups. SIenhancement rate between HCC and FNH (1.39±0.60 vs. 1.33±0.50, P>0.05) had no significant difference. (3) HCC and FNH morphological features in the hepatobiliary phase included: annular hypointensity: HCC (8 cases) vs. FNH (0 cases); contrast filling defects: HCC (8 cases) vs. FNH (0 cases); linear hyposignal separation: HCC (10 cases) vs. FNH (0 cases); and stellate scars: HCC (0) vs. FNH (5 cases), and there were statistically significant differences (P<0.05) between the two groups . Conclusion: Multimodal MRI have significant value for differentiating hyperintense HCC and FNH in the hepatobiliary phase.

Attention-based Deep Learning for the Preoperative Differentiation of Axillary Lymph Node Metastasis in Breast Cancer on DCE-MRI.

Previous studies have explored the potential on radiomics features of primary breast cancer tumor to identify axillary lymph node (ALN) metastasis. However, the value of deep learning (DL) to identify ALN metastasis remains unclear. To investigate the potential of the proposed attention-based DL model for the preoperative differentiation of ALN metastasis in breast cancer on dynamic contrast-enhanced MRI (DCE-MRI). Retrospective. A total of 941 breast cancer patients who underwent DCE-MRI before surgery were included in the training (742 patients), internal test (83 patients), and external test (116 patients) cohorts. A 3.0 T MR scanner, DCE-MRI sequence. A DL model containing a 3D deep residual network (ResNet) architecture and a convolutional block attention module, named RCNet, was proposed for ALN metastasis identification. Three RCNet models were established based on the tumor, ALN, and combined tumor-ALN regions on the images. The performance of these models was compared with ResNet models, radiomics models, the Memorial Sloan-Kettering Cancer Center (MSKCC) model, and three radiologists (W.L., H.S., and F. L.). Dice similarity coefficient for breast tumor and ALN segmentation. Accuracy, sensitivity, specificity, intercorrelation and intracorrelation coefficients, area under the curve (AUC), and Delong test for ALN classification. The optimal RCNet model, that is, RCNet-tumor+ALN , achieved an AUC of 0.907, an accuracy of 0.831, a sensitivity of 0.824, and a specificity of 0.837 in the internal test cohort, as well as an AUC of 0.852, an accuracy of 0.828, a sensitivity of 0.792, and a specificity of 0.853 in the external test cohort. Additionally, with the assistance of RCNet-tumor+ALN , the radiologists' performance was improved (external test cohort, P< 0.05). DCE-MRI-based RCNet model could provide a noninvasive auxiliary tool to identify ALN metastasis preoperatively in breast cancer, which may assist radiologists in conducting more accurate evaluation of ALN status. 3 TECHNICAL EFFICACY: Stage 2.

Dynamic contrast-enhanced magnetic resonance imaging for evaluating early response to radiosurgery in patients with vestibular schwannoma.

This study aimed to use dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) to evaluate early treatment response in vestibular schwannoma (VS) patients after radiosurgery. Twenty-four VS patients who underwent gamma knife radiosurgery were prospectively followed up for at least four years. DCE-MRI sequences, in addition to standard MRI protocol, were obtained prior to radiosurgery, at 3 and 6months. Conventionally, treatment responses based on tumor volume changes were classified as regression or stable (RS), transient tumor enlargement (TTE), and continuous tumor enlargement (CTE). DCE-MRI parameters, such as Ktrans, Kep and Ve, were compared according to follow-up periods and between groups. The diagnostic performance was tested using receiver operating characteristic (ROC) curves. Changes in tumor volume were as follows at the last 48 months of follow-up: RS in 11 patients (45.8%), TTE in 10 patients (41.7%), and CTE in three patients (12.5%). The median time required to distinguish TTE from CTE using conventional MRI was 12months (range 9-18). The Ktrans and Ve were significantly decreased in patients with RS and TTE at 3 and 6months, but did not differ significantly in patients with CTE. There were no significant differences in Ktrans and Ve between patients with RS and TTE at 3 and 6months. Both Ktrans and Ve demonstrated high diagnostic performance in evaluating early treatment response to radiosurgery in patients with VS. DCE-MRI may aid in the monitoring and early prediction of treatment response in patients with VS following radiosurgery.

The role of diffusion-weighted imaging and dynamic contrast-enhanced magnetic resonance imaging for the diagnosis of diabetic foot osteomyelitis: a preliminary report

PurposeTo assess the role of dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) and diffusionweighted imaging (DWI) in diagnosing diabetic foot osteomyelitis (DFO).Material and methodsTwenty-five participants with suspected osteomyelitis were included, who underwent MRI including DCE-MRI and DWI sequences. It was subsequently followed by bone biopsy and microbiological analysis (gold standard). The participants were divided into 2 groups based on biopsy results: DFO-positive or DFO-negative. The semi-quantitative DCE-MRI parameters (SI0, SImax, SIrel, wash-in rate [WIR], and type of curve) and apparent diffusion coefficient (ADC) values were subsequently compared between the 2 groups.ResultsOut of the 25 cases, 19 were DFO-positive and 6 were DFO-negative on bone biopsy. The SI0, SImax, and WIR were significantly higher in DFO-positive cases (p-value 0.050, 0.023, and 0.004, respectively). No difference was seen in SIrel. 100% negative cases revealed type-I curve, and 94% of positive cases showed type-II curve. SI0 > 143.4 revealed a sensitivity of 94.7% and specificity of 83.3%. SImax had a sensitivity of 89.5% but lower specificity of 67.7% at a cut-off value of 408.35. The most significant difference was seen with WIR; p-value ~0.004. At the cut-off value of > 1.280, it had a specificity and sensitivity of 100% and 76%, respectively. Also, ADC values below 1.57 × 10-3 had a sensitivity of 88.2% and specificity of 80% for diagnosing DFO.ConclusionsDWI and DCE-MRI provide non-invasive sequences, which can help to increase the overall specificity and sensitivity of conventional MRI for the diagnosis of osteomyelitis, differentiating it from acute Charcot’s arthropathy.

Exploration of chest wall perforator vascular anatomy on standard breast MRI: A potential aid to chest wall perforator flap planning

Local chest wall perforator flaps (CWPFs) are a volume replacement technique permitting breast-conserving surgery in patients who otherwise may require a mastectomy. These flaps are based on one or more perforating arteries arising from the lateral chest wall that travel through the soft tissue and into the sub-dermal plexus to perfuse the flap. Examples include the lateral intercostal and lateral thoracic artery perforators (LICAP and LTAP, respectively). Cross-sectional imaging of perforating vessels is not routinely performed, and vessels are mapped pre- and peri-operatively using a hand-held acoustic doppler device. As many breast cancer patients undergo pre-operative MRI scanning for oncological purposes, we investigated the role of MRI in mapping the vascular anatomy to aid with the surgical planning of CWPFs. We collated data retrospectively on a cohort of breast cancer patients who underwent breast MRI as part of routine pre-operative imaging. Axial 3D high-resolution dynamic contrast-enhanced MRI sequences with multiplanar reconstructions were analysed by a consultant radiologist. The presence and calibre of lateral chest wall perforator vessels were assessed. Fifty patients were suitable for inclusion. A consistent pattern of lateral chest wall vasculature was observed. Forty-eight patients (96%) demonstrated a bilateral lateral thoracic artery (LTA) descending inferiorly along the chest wall with two-thirds of these communicating with perforating intercostal vessels. True independent LICAP vessels were identified in six patients. From our observations, lateral CWPFs are dependent on an intricate intercommunication between intercostal vessels and the LTA which in turn supply perforators to the lateral chest wall donor site.

A deep learning fusion model with evidence-based confidence level analysis for differentiation of malignant and benign breast tumors using dynamic contrast enhanced MRI

ObjectiveAccurate interpretation of breast magnetic resonance imaging (MRI) remains challenging due to the lack of highly-specialized experiences of radiologists. The purpose of this study is to develop a Convolutional Neural Network (CNN) model based on dynamic contrast enhanced (DCE) MRI images as a reliable evidence-based computer aided diagnostic tool for breast cancer diagnosis to augment diagnostic performance. MethodsThis retrospective study included a total of 130 patients (71 malignant and 59 benign tumors). Surgical removal or biopsy procedures were performed for pathological analysis as the ground truth of tumor malignancy. The CNN model fused analytical features of tumor geometric information and pharmacokinetic properties of tumor tissues for differentiation of malignant and benign breast tumors. The confidence level of the CNN model classification outcome was interpreted based on three types of evidence criteria: prediction probability, feature hotmap visualization, and contributive dynamic scan time points. ResultsThe CNN model achieved an overall diagnostic accuracy of 87.7%, precision of 91.2%, sensitivity of 86.1%, and AUC of 91.2% ± 4.0% across five-fold testing process. The evidence-based model confidence analysis precluded misclassification given that all three types of evidence were met. ConclusionThis CNN-empowered computer aided diagnostic tool with evidence-based confidence level analysis achieved high diagnostic performance in breast tumor classification based on a single DCE MRI sequence. SignificanceThis work indicated that it is feasible to shorten breast cancer MRI screening protocol by including one DCE sequence, as well as to augment radiologist’s efficiency by focusing on cases with lower model confidence.

Subsampled brain MRI reconstruction by generative adversarial neural networks.

A main challenge in magnetic resonance imaging (MRI) is speeding up scan time. Beyond improving patient experience and reducing operational costs, faster scans are essential for time-sensitive imaging, such as fetal, cardiac, or functional MRI, where temporal resolution is important and target movement is unavoidable, yet must be reduced. Current MRI acquisition methods speed up scan time at the expense of lower spatial resolution and costlier hardware. We introduce a practical, software-only framework, based on deep learning, for accelerating MRI acquisition, while maintaining anatomically meaningful imaging. This is accomplished by MRI subsampling followed by estimating the missing k-space samples via generative adversarial neural networks. A generator-discriminator interplay enables the introduction of an adversarial cost in addition to fidelity and image-quality losses used for optimizing the reconstruction. Promising reconstruction results are obtained from feasible sampling patterns of up to a fivefold acceleration of diverse brain MRIs, from a large publicly available dataset of healthy adult scans as well as multimodal acquisitions of multiple sclerosis patients and dynamic contrast-enhanced MRI (DCE-MRI) sequences of stroke and tumor patients. Clinical usability of the reconstructed MRI scans is assessed by performing either lesion or healthy tissue segmentation and comparing the results to those obtained by using the original, fully sampled images. Reconstruction quality and usability of the DCE-MRI sequences is demonstrated by calculating the pharmacokinetic (PK) parameters. The proposed MRI reconstruction approach is shown to outperform state-of-the-art methods for all datasets tested in terms of the peak signal-to-noise ratio (PSNR), the structural similarity index (SSIM), as well as either the mean squared error (MSE) with respect to the PK parameters, calculated for the fully sampled DCE-MRI sequences, or the segmentation compatibility, measured in terms of Dice scores and Hausdorff distance. The code is available on GitHub.

Contrast Medium or No Contrast Medium for Prostate Cancer Diagnosis. That Is the Question.

The benefits and drawbacks of the dynamic contrast-enhanced MRI sequence for prostate cancer diagnosis are increasingly being recognized, with many centers adopting the biparametric (bp) MRI approach as the default initial approach. The abandonment of the routine use of contrast medium requires an assessment of the loss of diagnostic power against the gains in operational logistics, costs, time, capacity, and side effects. It is the balance of these factors weighted against the clinical priorities of patients that determines which patient groups can safely avoid dynamic contrast enhancement. Although systematic reviews and individual studies are broadly supportive of the bpMRI approach, the pathway impacts for men with suspected cancer using the bpMRI approach are still not well documented for clinical practice. Robust prospectively acquired data for bpMRI regarding biopsy avoidance, detection of clinically significant and insignificant cancers, and for increasing the precision of tumor grade and volume are needed. There is a requirement for prospective, randomized, or blinded head-to-head, multicenter studies, addressing the noninferiority of biopsy yields prompted by bpMRI and multiparametric MRI approaches. These studies should more precisely define patient groups where the benefits and harms of contrast enhancement are aligned to their clinical priorities. Only then can we be confident in recommending bpMRI as an initial diagnostic approach for prostate cancer diagnosis. Level of Evidence 1 Technical Efficacy Stage 5.

Clinical Implementation of a Free-Breathing, Motion-Robust Dynamic Contrast-Enhanced MRI Protocol to Evaluate Pleural Tumors.

OBJECTIVE. The purpose of this study was to develop a motion insensitive clinical dynamic contrast-enhanced MRI (DCE-MRI) protocol to assess the response of pleural tumors in clinical trials. MATERIALS AND METHODS. Thirty-two patients with pleura-based lesions were administered contrast material and imaged with gradient-recalled echo DCE-MRI sequence variants: either a traditional cartesian k-space acquisition (FLASH), a time-resolved imaging with stochastic trajectories acquisition (TWIST), or a radial stack-of-stars acquisition (radial) sequence in addition to other standard-of-care imaging sequences. Each image acquisition's sensitivity to motion was evaluated by comparing the motion of the thoracic border in 3D throughout the acquisition. One-way ANOVA was used to compare the image quality between different acquisitions. The 95% CIs were calculated for mean thoracic border displacement. The effects of motion on kinetic parameter estimation were explored with simulations according to clinically acquired data. RESULTS. Radial was the most motion-robust sequence with subvoxel mean displacement in the superior-inferior direction (0.4 ± 1.2 [SD] mm). FLASH showed intermediate displacement (4.6 ± 2.0 mm), whereas TWIST was most sensitive to motion (6.4 ± 3.4 mm). Signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) of the images acquired with the radial sequence were on par or better than the FLASH and TWIST sequences when reconstructed with an improved density compensation algorithm. Simulations showed that motion on scans showing pleural-based lesions can lead to markedly inaccurate kinetic parameter estimation and inappropriate kinetic model convergence within a nested model analysis. CONCLUSION. A practical radial k-space trajectory sequence that provides motion-insensitive pharmacokinetic parameters was incorporated as part of the DCE-MRI protocol of pleural tumors. Validation and usefulness in clinical trials assessing response to therapy is needed.

Post-treatment prostate MRI.

Accurate early detection of recurrent prostate cancer after surgical or nonsurgical treatment is increasingly relevant in the era of evolving options for salvage therapy. The importance of differentiating between local tumor recurrence, distant metastatic disease, and a combination of both in a patient with biochemical recurrence of prostate cancer is essential for appropriate treatment selection. Magnetic resonance imaging (MRI) is the best test for localization and characterization of locally residual or recurrent prostate cancer. It is essential for the radiologist involved in prostate MRI interpretation to be familiar with key imaging findings and advantages of different sequences to reach a confident diagnosis in the post-treatment setting. In this pictorial review, we present imaging findings of post-treatment prostate MRI including expected post-treatment anatomy and imaging characteristics, and the typical appearances of local tumor recurrence after radical prostatectomy, radiation therapy, and focal therapy for prostate cancer. While a multi-parametric MRI approach remains key just as in the treatment-naïve gland, this review emphasizes the much greater importance of the dynamic contrast-enhanced MRI sequence for evaluation in the post-treatment setting.