Diffusion-Weighted MRI-Based Virtual Elastography and Shear-Wave Elastography for the Assessment of Breast Lesions.

BACKGROUND: Diffusion-weighted imaging (DWI)-based virtual MR elastography (DWI-vMRE) in the assessment of breast lesions is still in the research stage. PURPOSE: To investigate the usefulness of elasticity values on DWI-vMRE in the evaluation of breast lesions, and the correlation with the values calculated from shear-wave elastography (SWE). STUDY TYPE: Prospective. POPULATION/SUBJECTS: 153 patients (mean age ± standard deviation: 55 ± 12 years) with 153 pathological confirmed breast lesions (24 benign and 129 malignant lesions). FIELD STRENGTH/SEQUENCE: 1.5-T MRI, multi-b readout segmented echo planar imaging (b-values of 0, 200, 800, and 1000 sec/mm2 ). ASSESSMENT: For DWI-vMRE assessment, lesions were manually segmented using apparent diffusion coefficient (ADC0-1000 ) map, then the region of interests were copied to the map of shifted-ADC (sADC200-800 , sADC 200-1500 ). For SWE assessment, the shear modulus of the lesions was measured by US elastic modulus (µUSE ). Intraclass/interclass kappa coefficients were calculated to measure the consistency. STATISTICAL TESTS: Pearson's correlation was used to assess the relationship between sADC and µUSE . A receiver operating characteristic analysis with the area under the curve (AUC) was performed to compare the diagnostic accuracy between benign and malignant breast lesions of sADC and µUSE . A P value <0.05 was considered statistically significant. RESULTS: There were significant differences between benign and malignant breast lesions of µUSE (24.17 ± 10.64 vs. 37.20 ± 12.61), sADC200-800 (1.38 ± 0.31 vs. 0.97 ± 0.18 × 10-3 mm2 /sec), and sADC200-1500 (1.14 ± 0.30 vs. 0.78 ± 0.13 × 10-3 mm2 /sec). In all breast lesions, a moderate but significant correlation was observed between µUSE and sADC200-800 /sADC200-1500 (r = -0.49/-0.44). AUC values to differentiate benign from malignant lesions were as follows: µUSE , 0.78; sADC200-800 , 0.89; sADC200-1500 , 0.89. DATA CONCLUSIONS: Both SWE and DWI-vMRE could be used for the differentiation of benign versus malignant breast lesions. Furthermore, DWI-vMRE with the use of sADC show relatively higher AUC values than SWE. LEVEL OF EVIDENCE: 4 TECHNICAL EFFICACY: Stage 2.

Anti-HER2 therapy response assessment for guiding treatment (de-)escalation in early HER2-positive breast cancer using a novel deep learning radiomics model.

OBJECTIVES: Anti-HER2 targeted therapy significantly reduces risk of relapse in HER2 + breast cancer. New measures are needed for a precise risk stratification to guide (de-)escalation of anti-HER2 strategy. METHODS: A total of 726 HER2 + cases who received no/single/dual anti-HER2 targeted therapies were split into three respective cohorts. A deep learning model (DeepTEPP) based on preoperative breast magnetic resonance (MR) was developed. Patients were scored and categorized into low-, moderate-, and high-risk groups. Recurrence-free survival (RFS) was compared in patients with different risk groups according to the anti-HER2 treatment they received, to validate the value of DeepTEPP in predicting treatment efficacy and guiding anti-HER2 strategy. RESULTS: DeepTEPP was capable of risk stratification and guiding anti-HER2 treatment strategy: DeepTEPP-Low patients (60.5%) did not derive significant RFS benefit from trastuzumab (p = 0.144), proposing an anti-HER2 de-escalation. DeepTEPP-Moderate patients (19.8%) significantly benefited from trastuzumab (p = 0.048), but did not obtain additional improvements from pertuzumab (p = 0.125). DeepTEPP-High patients (19.7%) significantly benefited from dual HER2 blockade (p = 0.045), suggesting an anti-HER2 escalation. CONCLUSIONS: DeepTEPP represents a pioneering MR-based deep learning model that enables the non-invasive prediction of adjuvant anti-HER2 effectiveness, thereby providing valuable guidance for anti-HER2 (de-)escalation strategies. DeepTEPP provides an important reference for choosing the appropriate individualized treatment in HER2 + breast cancer patients, warranting prospective validation. CLINICAL RELEVANCE STATEMENT: We built an MR-based deep learning model DeepTEPP, which enables the non-invasive prediction of adjuvant anti-HER2 effectiveness, thus guiding anti-HER2 (de-)escalation strategies in early HER2-positive breast cancer patients. KEY POINTS: • DeepTEPP is able to predict anti-HER2 effectiveness and to guide treatment (de-)escalation. • DeepTEPP demonstrated an impressive prognostic efficacy for recurrence-free survival and overall survival. • To our knowledge, this is one of the very few, also the largest study to test the efficacy of a deep learning model extracted from breast MR images on HER2-positive breast cancer survival and anti-HER2 therapy effectiveness prediction.

TP53 Mutation Estimation Based on MRI Radiomics Analysis for Breast Cancer.

BACKGROUND: Noninvasive detection of TP53 mutations is useful for the molecular stratification of breast cancer. PURPOSE: To explore MRI radiomics features reflecting TP53 mutations in breast cancer and propose a classifier for detecting such mutations. STUDY TYPE: Retrospective. POPULATION/SUBJECTS: A total of 139 breast cancer patients with TP53 expression profiling (98 with TP53 mutations and 41 without TP53 mutations). FIELD STRENGTH/SEQUENCE: 1.5 T, T1-weighted (T1W) DCE-MRI. ASSESSMENT: Lesions were manually segmented using subtracted T1WI. A total of 944 radiomics features (including 744 wavelet-related features) and 7 clinicopathological features were extracted from each lesion. Principal component analysis and Pearson's correlation analysis were used to preprocess the features. Linear discriminant analysis, logistic regression (LR), support vector machine (SVM), and random forest (RF) were used as the classifiers. STATISTICAL TESTS: Analysis of variance, Kruskal-Wallis and recursive features elimination were used to select features. Receiver operating characteristic (ROC) analysis was performed to compare the diagnostic accuracy. RESULTS: For the radiomics model, the validation cohorts AUCs of the four classifiers ranged from 0.69 (RF) to 0.74 (LR), and LR (0.74) attained the highest AUCs. For the clinicopathological-radiomics combined model, the validation AUCs of the four classifiers ranged from 0.68 (RF) to 0.86 (SVM), and SVM (0.86) attained highest AUCs. In the subgroup analysis of triple-negative (TN) and luminal type breast cancer, RF achieved the highest AUCs (0.83 and 0.94). DATA CONCLUSION: Clinicopathological-radiomics combined model with SVM could be used as noninvasive biomarkers for predicting TP53 mutations. RF was recommended for the detection of TP53 mutations in TN and luminal type breast cancer. LEVEL OF EVIDENCE: 3 TECHNICAL EFFICACY: Stage 2.

Impact of the transection plan on postoperative pancreatic fistulas occurring after robot-assisted distal pancreatectomy for nonmalignant pancreatic neoplasms.

OBJECTIVES: Postoperative pancreatic fistula (POPF) is the main complication of distal pancreatectomy (DP) and affects the prognosis of patients. The impact of several clinical factors mentioned in recent studies on POPF remains controversial. This study aimed to investigate the impact of a remnant pancreas and other perioperative factors on POPFs occurring after robot-assisted distal pancreatectomy (RDP) for nonmalignant pancreatic neoplasms. METHODS: A total of 197 patients who received robot-assisted distal pancreatectomy (RDP) for nonmalignant pancreatic neoplasms at the Pancreatic Disease Center, Ruijin Hospital Shanghai Jiaotong University School of Medicine from January 2018 to December 2020 were included in this retrospective study. According to the intraoperative transection plan, patients were divided into an RDP body group and an RDP tail group. Clinical and pathological features and perioperative factors affecting POPF were analyzed and compared between the two groups. RESULTS: The results showed that a transection plan involving the tail of the pancreas (OR = 2.133, 95% CI 1.109-4.103, p = 0.023) and spleen preservation (OR = 2.588, 95% CI 1.435-4.665, p = 0.001) independently increased the incidence of POPF in patients with nonmalignant pancreatic neoplasms treated by RDP. A transection plan involving the tail of the pancreas was also an independent risk factor (OR = 3.464, 95% CI 1.270-9.450, p = 0.015) for grade B/C POPF. Length of remnant pancreas > 6.23 cm was an independent risk factor for POPF (OR = 3.116, 95% CI 1.364-7.121, p = 0.007). Length of remnant pancreas > 9.82 cm was an independent risk factor for grade B/C POPF (OR = 3.340, 95% CI 1.386-8.051, p = 0.007). CONCLUSION: This retrospective study suggests that a transection plan involving the tail of the pancreas is an independent risk factor for POPF in patients with nonmalignant neoplasms treated by RDP. We also propose that the postoperative length of the remnant pancreas evaluated by computed tomography scans can be used to identify patients with a high risk of POPF in order to optimize the individualized strategy.

Radiomics-based machine learning analysis and characterization of breast lesions with multiparametric diffusion-weighted MR.

BACKGROUND: This study aimed to evaluate the utility of radiomics-based machine learning analysis with multiparametric DWI and to compare the diagnostic performance of radiomics features and mean diffusion metrics in the characterization of breast lesions. METHODS: This retrospective study included 542 lesions from February 2018 to November 2018. One hundred radiomics features were computed from mono-exponential (ME), biexponential (BE), stretched exponential (SE), and diffusion-kurtosis imaging (DKI). Radiomics-based analysis was performed by comparing four classifiers, including random forest (RF), principal component analysis (PCA), L1 regularization (L1R), and support vector machine (SVM). These four classifiers were trained on a training set with 271 patients via ten-fold cross-validation and tested on an independent testing set with 271 patients. The diagnostic performance of the mean diffusion metrics of ME (mADCall b, mADC0-1000), BE (mD, mD*, mf), SE (mDDC, mα), and DKI (mK, mD) were also calculated for comparison. The area under the receiver operating characteristic curve (AUC) was used to compare the diagnostic performance. RESULTS: RF attained higher AUCs than L1R, PCA and SVM. The AUCs of radiomics features for the differential diagnosis of breast lesions ranged from 0.80 (BE_D*) to 0.85 (BE_D). The AUCs of the mean diffusion metrics ranged from 0.54 (BE_mf) to 0.79 (ME_mADC0-1000). There were significant differences in the AUCs between the mean values of all diffusion metrics and radiomics features of AUCs (all P < 0.001) for the differentiation of benign and malignant breast lesions. Of the radiomics features computed, the most important sequence was BE_D (AUC: 0.85), and the most important feature was FO-10 percentile (Feature Importance: 0.04). CONCLUSIONS: The radiomics-based analysis of multiparametric DWI by RF enables better differentiation of benign and malignant breast lesions than the mean diffusion metrics.

High-Resolution Breast MRI Reconstruction Using a Deep Convolutional Generative Adversarial Network.

BACKGROUND: A generative adversarial network could be used for high-resolution (HR) medical image synthesis with reduced scan time. PURPOSE: To evaluate the potential of using a deep convolutional generative adversarial network (DCGAN) for generating HRpre and HRpost images based on their corresponding low-resolution (LR) images (LRpre and LRpost ). STUDY TYPE: This was a retrospective analysis of a prospectively acquired cohort. POPULATION: In all, 224 subjects were randomly divided into 200 training subjects and an independent 24 subjects testing set. FIELD STRENGTH/SEQUENCE: Dynamic contrast-enhanced (DCE) MRI with a 1.5T scanner. ASSESSMENT: Three breast radiologists independently ranked the image datasets, using the DCE images as the ground truth, and reviewed the image quality of both the original LR images and the generated HR images. The BI-RADS category and conspicuity of lesions were also ranked. The inter/intracorrelation coefficients (ICCs) of mean image quality scores, lesion conspicuity scores, and Breast Imaging Reporting and Data System (BI-RADS) categories were calculated between the three readers. STATISTICAL TEST: Wilcoxon signed-rank tests evaluated differences among the multireader ranking scores. RESULTS: The mean overall image quality scores of the generated HRpre and HRpost were significantly higher than those of the original LRpre and LRpost (4.77 ± 0.41 vs. 3.27 ± 0.43 and 4.72 ± 0.44 vs. 3.23 ± 0.43, P < 0.0001, respectively, in the multireader study). The mean lesion conspicuity scores of the generated HRpre and HRpost were significantly higher than those of the original LRpre and LRpost (4.18 ± 0.70 vs. 3.49 ± 0.58 and 4.35 ± 0.59 vs. 3.48 ± 0.61, P < 0.001, respectively, in the multireader study). The ICCs of the image quality scores, lesion conspicuity scores, and BI-RADS categories had good agreements among the three readers (all ICCs >0.75). DATA CONCLUSION: DCGAN was capable of generating HR of the breast from fast pre- and postcontrast LR and achieved superior quantitative and qualitative performance in a multireader study. LEVEL OF EVIDENCE: 3 TECHNICAL EFFICACY STAGE: 2 J. MAGN. RESON. IMAGING 2020;52:1852-1858.

Whole-lesion histogram and texture analyses of breast lesions on inline quantitative DCE mapping with CAIPIRINHA-Dixon-TWIST-VIBE.

PURPOSE: To investigate the diagnostic capability of whole-lesion (WL) histogram and texture analysis of dynamic contrast-enhanced (DCE) MRI inline-generated quantitative parametric maps using CAIPIRINHA-Dixon-TWIST-VIBE (CDTV) to differentiate malignant from benign breast lesions and breast cancer subtypes. MATERIALS AND METHODS: From February 2018 to November 2018, DCE MRI using CDTV was performed on 211 patients. The inline-generated parametric maps included Ktrans, kep, Ve, and IAUGC60. Histogram and texture features were extracted from the above parametric maps respectively based on a WL analysis. Student's t tests, one-way ANOVAs, Mann-Whitney U tests, Jonckheere-Terpstra tests, and ROC curves were used for statistical analysis. RESULTS: Compared with benign breast lesions, malignant breast lesions showed significantly higher Ktrans_median, 5th percentile, entropy, and diff-entropy, IAUGC60_median, 5th percentile, entropy, and diff-entropy, kep_mean, median, 5th percentile, entropy, and diff-entropy, and Ve_95th percentile, diff-variance, and contrast, and significantly lower kep_skewness and Ve_SD, entropy, diff-entropy, and skewness (all p ≤ 0.011). The combination of all the extracted parameters yielded an AUC of 0.85 (sensitivity 76%, specificity 86%). kep_contrast showed a significant difference among different subtypes of breast cancer (p = 0.006). kep_skewness showed a significant difference between lymph node-positive and lymph node-negative breast cancer (p = 0.007). The IAGC60_5th percentile had an AUC of 0.71 (sensitivity 50%, specificity 91%) for differentiating between high- and low-proliferation groups of breast cancer. CONCLUSIONS: The WL histogram and texture analyses of CDTV-DCE-derived parameters may give additional information for further evaluation of breast cancer. KEY POINTS: • Inline DCE mapping with CDTV is effective and time-saving. • WL histogram and texture-extracted features could distinguish breast cancer from benign lesions accurately. • kep_contrast, kep_skewness, and IAUGC60_5th percentile could predict breast cancer subtypes, lymph node metastasis, and proliferation abilities, respectively.

Diffusion tensor imaging of microstructural alterations in the trigeminal nerve due to neurovascular contact/compression.

BACKGROUND: Several recent studies have focused on microstructural changes in the trigeminal nerve in trigeminal neuralgia using diffusion tensor imaging (DTI). However, alterations after microvascular decompression (MVD) have rarely been investigated. Furthermore, the trigeminal nerve of asymptomatic individuals also presenting with neurovascular contact/compression (NVC) has not yet been studied. METHODS: Thirty-four patients suffering from trigeminal neuralgia and 34 healthy age-matched controls, who were identified as having unilateral NVC signs, underwent both DTI and high-resolution magnetic resonance imaging (MRI) for comparison. All trigeminal neuralgia patients underwent a post-surgical MRI scan after 7 days and a follow-up MRI scan within 6-8 months after surgery. The apparent diffusion coefficients (ADCs) and fractional anisotropy (FA) values were measured from coronal images in which the nerves from the root exit point to the distal segment were clearly shown. RESULTS: In 34 trigeminal neuralgia patients, the absolute FA value was significantly lower on the affected side (mean FA, 0.34 ± 0.03) than on the unaffected side (mean FA, 0.37 ± 0.05, p < 0.001). The FA ratio was also significantly different between the trigeminal neuralgia group (RsFA, 0.92 ± 0.06) and the control group (RsFA, 0.99 ± 0.09) (p = 0.001). The absolute ADC value between the two sides in patients and the ratios of ADC between the trigeminal neuralgia and control groups did not show any significant differences (p = 0.21 and 0.29, respectively). However, in 34 healthy subjects presenting with signs of NVC, neither the FA value nor the ADC showed a difference between sides (p > 0.05). The FA ratio of patients showed a significant increase on two follow-up MRI scans compared to the preoperative FA (p = 0.02 and 0.002, respectively), while the ADC ratio showed a significant decrease at 6 months after MVD (p = 0.004). CONCLUSION: This study of trigeminal neuralgia due to NVC found that DTI indexes could reflect alterations in the affected trigeminal nerve. Furthermore, a reversible change after MVD surgery could be potentially valuable for monitoring the change in white matter of the trigeminal nerve.

Molecular evaluation of thrombosis using X-ray phase contrast imaging with microbubbles targeted to P-selectin in mice.

OBJECTIVES: X-ray phase contrast imaging (PCI) provides excellent image contrast by utilizing the phase shift. The introduction of microbubbles into tissues can cause a phase shift to make microbubbles visibly identified on PCI. In this study, we assessed the feasibility of targeted microbubble-based PCI for the detection of thrombosis. METHODS: The absorption and phase contrast images of P-selectin-targeted microbubbles (MBP) were obtained and compared in vitro. MBP, control IgG-targeted microbubbles (MBC), and unbound microbubbles (MBU) were tested for binding specificity on thrombi expressing P-selectin. MBP were used as molecular PCI probes to evaluate P-selectin expression in a mouse model of arteriovenous shunt thrombosis that was created using PE tubes in the bypass outside of the mouse body. RESULTS: PCI clearly showed the microbubbles not viewable via absorption contrast imaging (ACI). In vitro attachment of MBP (91.60 ± 11.63) to thrombi was significantly higher than attachment of MBC (17.80 ± 4.02, P < 0.001) or MBU (9.80 ± 2.59, P < 0.001). In the mouse model of arteriovenous shunt thrombosis, the binding affinity of MBP (15.50 ± 6.25) was significantly greater than that of MBC (0.50 ± 0.84, P < 0.001) or MBU (0.33 ± 0.52, P < 0.001). CONCLUSIONS: Our results indicate that molecular PCI may be considered as a novel and promising imaging modality for the investigation of thrombosis. KEY POINTS: • Small thrombi are rarely detected by conventional radiography. • Phase contrast imaging (PCI) provides higher contrast and spatial resolution than conventional radiography. • P-selectin targeted microbubbles detected by PCI may suggest early thrombosis.

Diffusion-Weighted Imaging-guided MR Spectroscopy in Breast Lesions using Readout-Segmented Echo-Planar Imaging.

PURPOSE: To investigate the feasibility and effectiveness of diffusion-weighted imaging (DWI)-guided magnetic resonance spectroscopy (MRS) using readout-segmented echo-planar imaging (RS-EPI) to characterise breast lesions. MATERIALS AND METHODS: A total of 258 patients with 258 suspicious breast lesions larger than 1 cm in diameter were examined using DWI-guided, single-voxel MRS with RS-EPI. The mean total choline-containing compound (tCho) signal-to-noise ratio (SNR) and concentration were used for the interpretation of MRS data. T-tests, χ(2)-tests, receiver operating characteristic (ROC) curve analyses and Pearson correlations were conducted for statistical analysis. RESULTS: Histologically, 183 lesions were malignant, and 75 lesions were benign. Both the mean tCho SNR and concentration of malignant lesions were higher than those of benign lesions (6.23 ± 3.30 AU/mL vs. 1.26 ± 1.75 AU/mL and 3.17 ± 2.03 mmol/kg vs. 0.86 ± 0.83 mmol/kg, respectively; P < 0.0001). For a tCho SNR of 2.0 AU/mL and a concentration of 1.76 mmol/kg, the corresponding areas under the ROC curves were 0.93 and 0.90, respectively. The mean tCho SNR and concentration negatively correlated with apparent diffusion coefficients calculated from RS-EPI, with correlation coefficients of -0.54 and -0.48, respectively. CONCLUSION: DWI-guided MRS using RS-EPI is feasible and accurate for characterising breast lesions. KEY POINTS: • The mean tCho SNR and concentration negatively correlated with ADCs. • DWI-guided MRS using RS-EPI is feasible. • DWI-guided MRS using RS-EPI accurately characterises breast lesions.

Breast Cancer: Diffusion Kurtosis MR Imaging-Diagnostic Accuracy and Correlation with Clinical-Pathologic Factors.

PURPOSE: To assess diagnostic accuracy with diffusion kurtosis imaging (DKI) in patients with breast lesions and to evaluate the potential association between DKI-derived parameters and breast cancer clinical-pathologic factors. MATERIALS AND METHODS: Institutional review board approval and written informed consent were obtained. Data from 97 patients (mean age ± standard deviation, 45.7 years ± 13.1; range, 19-70 years) with 98 lesions (57 malignant and 41 benign) who were treated between January 2014 and April 2014 were retrospectively analyzed. DKI (with b values of 0-2800 sec/mm(2)) and conventional diffusion-weighted imaging data were acquired. Kurtosis and diffusion coefficients from DKI and apparent diffusion coefficients from diffusion-weighted imaging were measured by two radiologists. Student t test, Wilcoxon signed-rank test, Jonckheere-Terpstra test, receiver operating characteristic curves, and Spearman correlation were used for statistical analysis. RESULTS: Kurtosis coefficients were significantly higher in the malignant lesions than in the benign lesions (1.05 ± 0.22 vs 0.65 ± 0.11, respectively; P < .0001). Diffusivity and apparent diffusion coefficients in the malignant lesions were significantly lower than those in the benign lesions (1.13 ± 0.27 vs 1.97 ± 0.33 and 1.02 ± 0.18 vs 1.48 ± 0.33, respectively; P < .0001). Significantly higher specificity for differentiation of malignant from benign lesions was shown with the use of kurtosis and diffusivity coefficients than with the use of apparent diffusion coefficients (83% [34 of 41] and 83% [34 of 41] vs 76% [31 of 41], respectively; P < .0001) with equal sensitivity (95% [54 of 57]). In patients with invasive breast cancer, kurtosis was positively correlated with tumor histologic grade (r = 0.75) and expression of the Ki-67 protein (r = 0.55). Diffusivity was negatively correlated with tumor histologic grades (r = -0.44) and Ki-67 expression (r = -0.46). CONCLUSION: DKI showed higher specificity than did conventional diffusion-weighted imaging for assessment of benign and malignant breast lesions. Patients with grade 3 breast cancer or tumors with high expression of Ki-67 were associated with higher kurtosis and lower diffusivity coefficients; however, this association must be confirmed in prospective studies.

Solid pseudopapillary tumour of the pancreas: distinct patterns of computed tomography manifestation for male versus female patients.

PURPOSE: The purpose of this study was to retrospectively assess the features of computed tomography (CT) images and clinical characteristics of male patients with solid pseudopapillary tumours (SPTs) and compare them with those of female patients. MATERIALS AND METHODS: Computed tomography images and clinical data of 102 patients with pathologically proven SPTs were reviewed. Details of the location, diameter, shape, encapsulation, calcification, internal composition, CT attenuation, and enhancement pattern of tumours were noted. Statistical analysis was performed using the χ (2) and t tests. RESULTS: Data from 16 males and 86 females were collected. Males were significantly older than females (38.5 years vs. 28.7 years; P = 0.004). Except for mean age, no significant statistical difference was observed between the clinical factors of SPTs in males and females. The mean tumour size in males was significantly smaller than that in females (5.3 vs. 7.6 cm; P = 0.037). Solid tumours were more common in males (8/16; 50 %) than in females (5/86; 5.8 %; P < 0.001). CONCLUSION: The imaging features of SPTs of males are different from those of females. In males, the finding of small, prominently solid tumours showing enhancement patterns typical of SPTs may suggest a diagnosis of SPT.

Whole-tumor histogram analysis of multiparametric breast magnetic resonance imaging to differentiate pure mucinous breast carcinomas from fibroadenomas with high-signal intensity on T2WI.

PURPOSE: To investigate the utility of whole-tumor histogram analysis based on multiparametric MRI in distinguishing pure mucinous breast carcinomas (PMBCs) from fibroadenomas (FAs) with strong high-signal intensity on T2-weighted imaging (T2-SHi). MATERIAL AND METHODS: The study included 20 patients (mean age, 55.80 ± 15.54 years) with single PBMCs and 29 patients (mean age, 42.31 ± 13.91 years) with single FAs exhibiting T2-SHi. A radiologist performed whole-tumor histogram analysis between PBMC and FA groups with T2-SHi using multiparametric MRI, including T2-weighted imaging (T2WI), diffusion weighted imaging (DWI) with apparent diffusion coefficient (ADC) maps, and the first (DCE_T1) and last (DCE_T4) phases of T1-weighted dynamic contrast-enhanced imaging (DCE) images, to extract 11 whole-tumor histogram parameters. Histogram parameters were compared between the two groups to identify significant variables using univariate analyses, and their diagnostic performance was assessed by receiver operating characteristic (ROC) curve analysis and logistic regression analyses. In addition, 15 breast lesions were randomly selected and histogram analysis was repeated by another radiologist to assess the intraclass correlation coefficient for each histogram feature. Pearson's correlation coefficients were used to analyze the correlations between histogram parameters and Ki-67 expression of PMBCs. RESULTS: For T2WI images, mean, median, maximum, 90th percentile, variance, uniformity, and entropy significantly differed in PBMCs and FAs with T2-SHi (all P < 0.05), yielding a combined area under the curve (AUC) of 0.927. For ADC maps, entropy was significantly lower in FAs with T2-SHi than in PMBCs (P = 0.03). In both DCE_T1 and DCE_T4 sequences, FAs with T2-SHi showed significantly higher minimum values than PBMCs (P = 0.007 and 0.02, respectively). The highest AUC value of 0.956 (sensitivity, 0.862; specificity, 0.944; positive predictive value, 0.962; negative predictive value, 0.810) was obtained when all significant histogram parameters were combined. CONCLUSIONS: Whole-tumor histogram analysis using multiparametric MRI is valuable for differentiating PBMCs from FAs with T2-SHi.

Exploration of anatomical distribution of brain metastasis from breast cancer at first diagnosis assisted by artificial intelligence.

Objectives: This study aimed to explore the spatial distribution of brain metastases (BMs) from breast cancer (BC) and to identify the high-risk sub-structures in BMs that are involved at first diagnosis. Methods: Magnetic resonance imaging (MRI) scans were retrospectively reviewed at our centre. The brain was divided into eight regions according to its anatomy and function, and the volume of each region was calculated. The identification and volume calculation of metastatic brain lesions were accomplished using an automatically segmented 3D BUC-Net model. The observed and expected rates of BMs were compared using 2-tailed proportional hypothesis testing. Results: A total of 250 patients with BC who presented with 1694 BMs were retrospectively identified. The overall observed incidences of the substructures were as follows: cerebellum, 42.1 %; frontal lobe, 20.1 %; occipital lobe, 9.7 %; temporal lobe, 8.0 %; parietal lobe, 13.1 %; thalamus, 4.7 %; brainstem, 0.9 %; and hippocampus, 1.3 %. Compared with the expected rate based on the volume of different brain regions, the cerebellum, occipital lobe, and thalamus were identified as higher risk regions for BMs (P value ≤ 5.6*10-3). Sub-group analysis according to the type of BC indicated that patients with triple-negative BC had a high risk of involvement of the hippocampus and brainstem. Conclusions: Among patients with BC, the cerebellum, occipital lobe and thalamus were identified as higher-risk regions than expected for BMs. The brainstem and hippocampus were high-risk areas of the BMs in triple negative breast cancer. However, further validation of this conclusion requires a larger sample size.

Hepatocellular carcinoma and focal nodular hyperplasia of the liver: differentiation with CT spectral imaging.

OBJECTIVES: To investigate the value of CT spectral imaging in differentiating hepatocellular carcinoma (HCC) from focal nodular hyperplasia (FNH) during the arterial phase (AP) and portal venous phase (PP). METHODS: Fifty-eight patients with 42 HCCs and 16 FNHs underwent spectral CT during AP and PP. The lesion-liver contrast-to-noise ratio (CNR) at different energy levels, normalised iodine concentrations (NIC) and the lesion-normal parenchyma iodine concentration ratio (LNR) were calculated. The two-sample t test compared quantitative parameters. Two readers qualitatively assessed lesion types according to imaging features. Sensitivity and specificity of the qualitative and quantitative studies were compared. RESULTS: In general, CNRs at low energy levels (40-70 keV) were higher than those at high energy levels (80-140 keV). NICs and LNRs for HCC differed significantly from those of FNH: mean NICs were 0.25 mg/mL ± 0.08 versus 0.42 mg/mL ± 0.12 in AP and 0.52 mg/mL ± 0.14 versus 0.86 mg/mL ± 0.18 in PP. Mean LNRs were 2.97 ± 0.50 versus 6.15 ± 0.62 in AP and 0.99 ± 0.12 versus 1.22 ± 0.26 in PP. NICs and LNRs for HCC were lower than those of FNH. LNR in AP had the highest sensitivity and specificity in differentiating HCC from FNH. CONCLUSIONS: CT spectral imaging may help to increase detectability of lesions and accuracy of differentiating HCC from FNH. KEY POINTS: • CT spectral imaging may help to detect hepatocellular carcinoma (HCC). • CT spectral imaging may help differentiate HCC from focal nodular hyperplasia. • Quantitative analysis of iodine concentration provides greater diagnostic confidence. • Treatment can be given with greater confidence.

[MRI outstands mammogram in sensitivity of breast ductal carcinoma in situ:an analysis of 122 cases].

OBJECTIVES: To compare the sensitivity of mammogram and breast dedicated MRI in detecting ductal carcinoma in situ with microinvaion (DCIS-MI) and ductal carcinoma in situ (DCIS) lesions, and to further investigate the independent predictive factors of mammogram and MRI sensitivity. METHODS: From August 2009 to November 2011, 122 consecutive confirmed breast cancer patients who had received operations were recruited for this clinical research. These patients were divided into two groups including DCIS (72 cases) and DCIS-MI (50 cases) based on pathologic reports. All the patients were female, with mean ages of 52.6 years and 54.4 years. Preoperative bilateral breast mammogram, breast dedicated MRI depictions and reports as well as histopathological reports were collected. RESULTS: Sensitivity of MRI outstood mammogram in each subgroups: 84.7% vs. 42.4% in DCIS (χ(2) = 27.028, P = 0.000), 94.0% vs. 80.0% in DCIS-MI group (χ(2) = 4.540, P = 0.040). And further analysis showed that MRI was more sensitive to high nuclear grade DCIS and DCIS-MI lesions than low nuclear grade ones (OR = 3.471, P = 0.031). RESULTS: of logistic regression analysis proved microcalcification was an independent predictive factor of mammogram sensitivity (OR = 11.287, P = 0.001). CONCLUSIONS: Sensitivity of breast dedicated MRI is superior to mammogram in detecting DCIS and DCIS-MI groups. Lesions with microcalcifiation is an independent predictive marker which meant that mammogram would achieve high detection rate in cancers presented calcification on mammogram image when compared with non-calcification. Diagnostic performance of breast MRI is less affected by clinical and pathological characteristics of the early stage breast cancer patients but further increased detection rate is observed in DCIS and DCIS-MI with high nuclear grade lesions which indicated that MRI could detect more early stage cancers with relative more aggression biological behaviour and provide these patients with early surgical interventions before possible progression to invasive breast cancers.

Multimodality MRI radiomics analysis of TP53 mutations in triple negative breast cancer.

Objectives: To explore the value of T1-weighted imaging (T1WI), T2-weighted imaging (T2WI) and diffusion-weighted imaging (DWI) radiomics features reflecting TP53 mutations in patients with triple negative breast cancer (TNBC). Study design: This retrospective study enrolled 91 patients with TNBC with TP53 testing (64 patients in the training cohort and 27 patients in the validation cohort). A total of 2832 radiomics features were extracted from the first phase of dynamic contrast-enhanced T1WI, T2WI and ADC maps. Analysis of variance (ANOVA) and the Kruskal-Wallis-test were used for feature selection. Then, linear discriminant analysis (LDA), multilayer perceptron (MLP), logistic regression (LR), LR with LASSO, decision tree (DT), naïve Bayes (NB), random forest (RF), and support vector machine (SVM) models were used for classification. Results: The validation AUCs of the eight classifiers ranged from 0.74 (NB) to 0.85 (SVM). SVM attained the highest AUC (0.85) and diagnostic accuracy (0.82) of all tested models. The top 3 ranking features in the SVM model were T1-square-first order-skewness (coefficient: 1.735), T2-wavelet-LHH-GLCM-joint energy, and T2-wavelet-LHH-GLCM-inverse difference moment (coefficient: -0.654, -0.634). Conclusions: Radiomics-based analysis with the SVM model is recommended for the detection of TP53 mutations in TNBC. Furthermore, T1WI- and T2WI-related features could be used as noninvasive biomarkers for predicting TP53 mutations.

Pancreatic index: A prognostic factor of upfront surgery for body or tail pancreatic ductal adenocarcinoma with vascular involvement-A retrospective study.

BACKGROUND: The pancreatic index (PI) is a useful preoperative imaging predictor for pancreatic ductal adenocarcinoma (PDAC). In this retrospective study, we determined the predictive effect of PI to distinguish patients of pancreatic body/tail cancer (PBTC) with vascular involvement who can benefit from upfront surgery. METHOD: All patients who received distal pancreatectomy for PDAC from 2016 to 2020 at the Pancreatic Disease Center, Ruijin Hospital, Shanghai Jiaotong University School of Medicine were considered for the study. A total of 429 patients with PBTC were assessed in relation to the value of PI. Fifty-five patients were eventually included and divided into low PI group and 29 patients in the normal PI group. RESULTS: The median overall survival (mOS) was significantly shorter in the low PI group (13.1 vs. 30.0 months, p = 0.002) in this study, and PI ≥ 0.78 (OR = 0.552, 95% CI: 0.301-0.904, p = 0.020) was an independent influencing factor confirmed by multivariate analysis. Subgroup analysis showed that PI was an independent prognostic factor for LA-PBTC (OR = 0.272, 95% CI: 0.077-0.969, p = 0.045). As for BR PBTC, PI (OR = 0.519, 95% CI: 0.285-0.947, p = 0.033) combined with carbohydrate antigen 125 (CA125) (OR = 2.806, 95% CI: 1.206-6.526, p = 0.017) and chemotherapy (OR = 0.327, 95% CI: 0.140-0.763, p = 0.010) were independent factors. CONCLUSION: This study suggests that the PI can be used as a predictive factor to optimize the surgical indication for PBTC with vascular involvement. Preoperative patients with normal PI and CA125 can achieve a long-term prognosis comparable to that of resectable PBTC patients.

A deep-learning radiomics-based lymph node metastasis predictive model for pancreatic cancer: a diagnostic study.

OBJECTIVES: Preoperative lymph node (LN) status is essential in formulating the treatment strategy among pancreatic cancer patients. However, it is still challenging to evaluate the preoperative LN status precisely now. METHODS: A multivariate model was established based on the multiview-guided two-stream convolution network (MTCN) radiomics algorithms, which focused on primary tumor and peri-tumor features. Regarding discriminative ability, survival fitting, and model accuracy, different models were compared. RESULTS: Three hundred and sixty-three pancreatic cancer patients were divided in to train and test cohorts by 7:3. The modified MTCN (MTCN+) model was established based on age, CA125, MTCN scores, and radiologist judgement. The MTCN+ model outperformed the MTCN model and the artificial model in discriminative ability and model accuracy. [Train cohort area under curve (AUC): 0.823 vs. 0.793 vs. 0.592; train cohort accuracy (ACC): 76.1 vs. 74.4 vs. 56.7%; test cohort AUC: 0.815 vs. 0.749 vs. 0.640; test cohort ACC: 76.1 vs. 70.6 vs. 63.3%; external validation AUC: 0.854 vs. 0.792 vs. 0.542; external validation ACC: 71.4 vs. 67.9 vs. 53.5%]. The survivorship curves fitted well between actual LN status and predicted LN status regarding disease free survival and overall survival. Nevertheless, the MTCN+ model performed poorly in assessing the LN metastatic burden among the LN positive population. Notably, among the patients with small primary tumors, the MTCN+ model performed steadily as well (AUC: 0.823, ACC: 79.5%). CONCLUSIONS: A novel MTCN+ preoperative LN status predictive model was established and outperformed the artificial judgement and deep-learning radiomics judgement. Around 40% misdiagnosed patients judged by radiologists could be corrected. And the model could help precisely predict the survival prognosis.

The value of whole-tumor histogram and texture analysis based on apparent diffusion coefficient (ADC) maps for the discrimination of breast fibroepithelial lesions: corresponds to clinical management decisions.

PURPOSE: This study aims to comprehensively evaluate the diagnostic value of quantitative parameters extracted from apparent diffusion coefficient (ADC) maps in distinguishing fibroepithelial tumors using whole-tumor histogram and texture analysis. MATERIALS AND METHODS: This retrospective study included 66 female patients with single phyllodes tumor (PT) and 29 female patients with single fibroadenoma (FA) who underwent preoperative magnetic resonance imaging. By independently performing whole-tumor histogram and texture analysis based on ADC maps, two radiologists extracted seven histogram parameters and four texture parameters. The extracted parameters were compared using univariate analysis to determine their ability to distinguish FAs from PTs, benign PTs from FAs, as well as benign PTs from borderline and malignant PTs. RESULTS: When FAs and PTs were compared, ADC_skewness values of PTs were significantly lower than those of FAs (p < 0.05), whereas other significant extracted parameter values of PTs were significantly higher than those of FAs (p ≤ 0.001); the area under the curve of significant parameters combined was 0.936. Regarding the differences between FAs and benign PTs, ADC_SD, ADC_95th percentile and ADC_kurtosis of FAs were significantly lower than those of benign PT group, and ADC_skewness was higher than that of benign PT group (all p < 0.05). Furthermore, ADC_SD, ADC_95th percentile and all texture parameters are significantly higher in the borderline and malignant PT group than in FA and benign PT group (p < 0.05). In addition, ADC_kurtosis of malignant PT group was significantly lower than that of borderline PT group (p = 0.045). CONCLUSION: The extracted whole-tumor histogram and texture features of ADC maps can improve differential diagnosis of breast fibroepithelial tumors and contribute to optimal selection for clinical management of patients with fibroepithelial tumors.