Dynamic graph transformer network via dual-view connectivity for autism spectrum disorder identification.

Autism Spectrum Disorder (ASD) is a neurodevelopmental disorder that requires objective and accurate identification methods for effective early intervention. Previous population-based methods via functional connectivity (FC) analysis ignore the differences between positive and negative FCs, which provide the potential information complementarity. And they also require additional information to construct a pre-defined graph. Meanwhile, two challenging demand attentions are the imbalance of performance caused by the class distribution and the inherent heterogeneity of multi-site data. In this paper, we propose a novel dynamic graph Transformer network based on dual-view connectivity for ASD Identification. It is based on the Autoencoders, which regard the input feature as individual feature and without any inductive bias. First, a dual-view feature extractor is designed to extract individual and complementary information from positive and negative connectivity. Then Graph Transformer network is innovated with a hot plugging K-Nearest Neighbor (KNN) algorithm module which constructs a dynamic population graph without any additional information. Additionally, we introduce the PolyLoss function and the Vrex method to address the class imbalance and improve the model's generalizability. The evaluation experiment on 1102 subjects from the ABIDE I dataset demonstrates our method can achieve superior performance over several state-of-the-art methods and satisfying generalizability for ASD identification.

Evaluating the expression of tumorigenic long noncoding RNAs in circulating exosomes isolated from non-small-cell lung cancer patients.

Aim: To evaluate the correlation of long noncoding RNAs (lncRNAs) expression in circulating exosomes and the cancerous and noncancerous tissues in patients with non-small-cell lung carcinoma. Methods: The relative expression of the four lncRNAs including LUADT1, MALAT1, NEAT1 and MIAT between tumor tissue, adjacent noncancerous tissues and circulating exosomes were evaluated by quantitative reverse transcription PCR. Results & conclusion: The relative expression of the lncRNAs, including LUADT1, MALAT1 and NEAT1, was upregulated and MIAT was downregulated in tumor tissue compared with noncancerous tissue samples. The expression of lncRNAs in circulating exosomes was not significantly different from cancerous tissue. Our results indicate that the studied exosomal lncRNAs have a good potential to be further evaluated as prognostic/diagnostic biomarkers in patients with non-small-cell lung cancer.

Magnetic Resonance Imaging Radiomics-Based Nomogram From Primary Tumor for Pretreatment Prediction of Peripancreatic Lymph Node Metastasis in Pancreatic Ductal Adenocarcinoma: A Multicenter Study.

BACKGROUND: Determining the absence or presence of peripancreatic lymph nodal metastasis (PLNM) is important to the pathologic staging, prognostication, and guidance of treatment in pancreatic ductal adenocarcinoma (PDAC) patients. Computed tomography and MRI had a poor sensitivity and diagnostic accuracy in the assessment of PLNM. PURPOSES: To develop and validate a 3 T MRI primary tumor radiomics-based nomogram from multicenter datasets for pretreatment prediction of the PLNM in PDAC patients. STUDY TYPE: Retrospective. SUBJECTS: A total of 251 patients (156 men and 95 women; mean age, 60.85 ± 8.23 years) with histologically confirmed pancreatic ductal adenocarcinoma from three hospitals. FIELD STRENGTH AND SEQUENCES: A 3.0 T and fat-suppressed T1-weighted imaging. ASSESSMENT: Quantitative imaging features were extracted from fat-suppressed T1-weighted (FS T1WI) images at the arterial phase. STATISTICAL TESTS: Normally distributed data were compared by using t-tests, while the Mann-Whitney U test was used to evaluate non-normally distributed data. The diagnostic performances of the preoperative and postoperative nomograms were assessed in the external validation cohort with the area under receiver operating characteristics curve (AUC), calibration curve, and decision curve analysis (DCA). AUCs were compared with the De Long test. A p value below 0.05 was considered to be statistically significant. RESULTS: The AUCs of magnetic resonance imaging (MRI) Rad-score were 0.868 (95% confidence level [CI]: 0.613-0.852) and 0.772 (95% CI: 0.659-0.879) in the training and internal validation cohort, respectively. The preoperative and postoperative nomograms could accurately predict PLNM in the training cohort (AUC = 0.909 and 0.851) and were validated in both the internal and external cohorts (AUC = 0.835 and 0.805, 0.808 and 0.733, respectively). DCA indicated that the two novel nomograms are of similar clinical usefulness. DATA CONCLUSION: Pre-/postoperative nomograms and the constructed radiomics signature from primary tumor based on FS T1WI of arterial phase could serve as a potential tool to predict PLNM in patients with PDAC. EVIDENCE LEVEL: 3 TECHNICAL EFFICACY: Stage 2.

Multifocal autoimmune pancreatitis: A retrospective study in a single tertiary center of 26 patients with a 20-year literature review.

BACKGROUND: Multifocal-type autoimmune pancreatitis (AIP), sometimes forming multiple pancreatic masses, is frequently misdiagnosed as pancreatic malignancy in routine clinical practice. It is critical to know the imaging features of multifocal-type AIP to prevent misdiagnosis and unnecessary surgery. To the best of our knowledge, there have been no studies evaluating the value of diffusionweighted imaging (DWI), axial fat-suppressed T1 weighted image (T1WI), and dynamic contrast enhanced-computed tomography (DCE-CT) in detecting the lesions of multifocal-type AIP. AIM: To clarify the exact prevalence and radiological findings of multifocal AIP in our cohorts and compare the sensitivity of DWI, axial fat-suppressed T1WI, and DCE-CT for detecting AIP lesions. We also compared radiological features between multifocal AIP and pancreatic ductal adenocarcinoma with several key imaging landmarks. METHODS: Twenty-six patients with proven multifocal AIP were retrospectively included. Two blinded independent radiologists rated their confidence level in detecting the lesions on a 5-point scale and assessed the diagnostic performance of DWI, axial fat-suppressed T1WI, and DCE-CT. CT and magnetic resonance imaging of multifocal AIP were systematically reviewed for typical imaging findings and compared with the key imaging features of pancreatic ductal adenocarcinoma. RESULTS: Among 118 patients with AIP, 26 (22.0%) had multiple lesions (56 lesions). Ulcerative colitis was associated with multifocal AIP in 7.7% (2/26) of patients, and Crohn's disease was present in 15.3% (4/26) of patients. In multifocal AIP, multiple lesions, delayed homogeneous enhancement, multifocal strictures of the main pancreatic duct, capsule-like rim, lower apparent diffusion coefficient values, and elevated serum Ig4 level were observed significantly more frequently than pancreatic ductal adenocarcinoma, whereas the presence of capsule-like rim in multifocal-type AIP was lower in frequency than total AIP. Of these lesions of multifocal AIP, DWI detected 89.3% (50/56) and 82.1% (46/56) by the senior and junior radiologist, respectively. CONCLUSION: Multifocal AIP is not as rare as previously thought and was seen in 22.0% of our patients. The diagnostic performance of DWI for detecting multifocal AIP was best followed by axial fat-suppressed T1WI and DCE-CT.

MR imaging features of pancreatic schwannoma: a Chinese case series and a systematic review of 25 cases.

BACKGROUND: There is a paucity of existing literature centering on the magnetic resonance (MR) imaging features of pancreatic schwannomas, due to the neoplasm's nonspecific presentation and its rarity. We aimed to identify the characteristic imaging features of pancreatic schwannoma. METHODS: This retrospective search was conducted for histologically confirmed pancreatic schwannoma in multi-institutional database of pathology. Abdominal magnetic resonance imaging (MRI) was performed before histologic examination and their MR imaging studies were independently reviewed. The search yielded six adults (mean age, 46 years) with a definitive histologic postoperative diagnosis of single pancreatic schwannoma each. Additionally, a comprehensive English and Chinese literature review for pancreatic schwannoma and reported MR-imaging findings since 1961 was also conducted. MR imaging features of those cases in the literature were analyzed, summarized and compared with our case series. RESULTS: This rare entity appeared to be a well-circumscribed, exophytic, oval or round pancreatic mass with a mean greatest diameter of 3.7 cm. Five schwannomas were located in the pancreatic head-neck and one in the pancreatic tail. On MRI, all cases appeared hypointense on T1-weighted images, inhomogeneous hyperintense on T2-weighted images, and hyperintense on diffusion-weighted images. The mean apparent diffusion coefficient (ADC) values of pancreatic schwannoma were 1.11 ± 0.29 × 10- 3 mm2/s and significantly lower than the surrounding pancreas. The lesion-to-pancreas signal intensity ratio (SIR) at unenhanced T1-weighted images was 0.53 ± 0.07. On dynamic contrast-enhanced MRI, most of the lesions (67%, 4/6) showed homogeneously iso- or hypointense on arterial and portal venous phases, and hyperenhancement on delayed phase compared with the surrounding pancreas. In our analysis of the time intensity curves, all cases exhibited a gradual enhancement pattern. CONCLUSIONS: A well-circumscribed mass displaying inhomogeneous hyperintensity on T2, marked hypointensity on T1, hyperintensity on DWI, and with early slight enhancement at arterial phase and progressive enhancement at portal venous and delayed phase, may suggest the diagnosis of pancreatic schwannoma.

Cascaded MultiTask 3-D Fully Convolutional Networks for Pancreas Segmentation.

Automatic pancreas segmentation is crucial to the diagnostic assessment of diabetes or pancreatic cancer. However, the relatively small size of the pancreas in the upper body, as well as large variations of its location and shape in retroperitoneum, make the segmentation task challenging. To alleviate these challenges, in this article, we propose a cascaded multitask 3-D fully convolution network (FCN) to automatically segment the pancreas. Our cascaded network is composed of two parts. The first part focuses on fast locating the region of the pancreas, and the second part uses a multitask FCN with dense connections to refine the segmentation map for fine voxel-wise segmentation. In particular, our multitask FCN with dense connections is implemented to simultaneously complete tasks of the voxel-wise segmentation and skeleton extraction from the pancreas. These two tasks are complementary, that is, the extracted skeleton provides rich information about the shape and size of the pancreas in retroperitoneum, which can boost the segmentation of pancreas. The multitask FCN is also designed to share the low- and mid-level features across the tasks. A feature consistency module is further introduced to enhance the connection and fusion of different levels of feature maps. Evaluations on two pancreas datasets demonstrate the robustness of our proposed method in correctly segmenting the pancreas in various settings. Our experimental results outperform both baseline and state-of-the-art methods. Moreover, the ablation study shows that our proposed parts/modules are critical for effective multitask learning.

Assessment of acute obstructive pancreatitis by magnetic resonance imaging: Predicting the occurrence of pancreatic fistula following pancreatoduodenectomy.

The aim of the present study was to assess the effect of preoperative acute obstructive pancreatitis on the risk of the occurrence of pancreatic fistula (PF) following pancreaticoduodenectomy. A total of 22 patients who developed postoperative PF were carefully matched with 22 control patients without PF according to demographic data, pancreatic pathology, presenting symptoms and other surgery-associated parameters. These parameters were compared between these two groups. The mean pancreatic apparent diffusion coefficient (ADC) values in the fistula group were 1.14±0.31×10-3 mm2/s, which was significantly decreased compared with the non-fistula group (1.48±0.44×10-3 mm2/s) (P=0.005). The pancreas-muscle signal intensity (SI) ratio on fat-suppressed T1 weighted image (rT1) in the fistula group was 1.71±0.25, which was significantly increased compared with the non-fistula group (1.25±0.29) (P<0.001). The pancreas-muscle SI ratios on fat-suppressed T2 weighted image (T2WI) in the fistula group and the non-fistula group were 0.72±0.08 and 0.62±0.07, respectively (P=0.79). There was no significant difference in pancreas-muscle SI ratio on fat-suppressed T2-weighted image (rT2) value between these two groups. Based on the receiver operating characteristic curve, the optimal cut-off value of ADC as a criterion for prediction of pancreatic fistula was 1.29×10-3 mm2/s, which yielded a sensitivity of 77.3% and a specificity of 63.6%. In conclusion, the severity of acute obstructive pancreatitis was negatively associated with ADC values and pancreas-muscle SI ratio on rT1 images, which may be useful for predicting the occurrence of PF preoperatively.

Homogenously isoattenuating insulinoma on biphasic contrast-enhanced computed tomography: Little benefits of diffusion-weighted imaging for lesion detection.

The aim of the present study was to evaluate the diagnostic benefit of diffusion-weighted imaging (DWI) in the detection of homogenous isoattenuating insulinoma on biphasic contrast-enhanced computed tomography (CT) preoperatively and to determine which magnetic resonance (MR) sequences exhibited the best diagnostic performance. A total of 44 consecutive patients who underwent biphasic contrast-enhanced CT and conventional MR imaging (MRI), including DWI on a 3T scanner, were identified retrospectively. Apparent diffusion coefficient (ADC) values of insulinomas and the surrounding pancreatic parenchyma were compared using a Wilcoxon signed-rank test. Receiver operating characteristic analysis was used to compare the diagnostic accuracy of four randomized image sets [T2-weighted image (WI), axial T1WI, DWI and T2WI + DWI] for each reader. Axial T1-weighted MRI exhibited the highest relative sensitivity for each reader; DWI alone exhibited the lowest relative sensitivity and the lower inter-reader agreement. There was no significant difference in lesion detection between T2WI and T2WI + DWI image sets for each reader. The ADC values of the insulinoma were significantly lower compared with those of the surrounding parenchyma. In conclusion, DWI does not benefit the detection of homogenous isoattenuating insulinoma. Axial T1WI is the optimal pulse sequence. Quantitative assessment of the tumor ADC values may be a useful tool to characterize identified pancreatic neoplasms.

CT and MR imaging features in phosphaturic mesenchymal tumor-mixed connective tissue: A case report.

Phosphaturic mesenchymal tumor-mixed connective tissue (PMT-MCT) is rare and usually benign and slow-growing. The majority of these tumors is associated with sporadic tumor-induced osteomalacia (TIO) or rickets, affect middle-aged individuals and are located in the extremities. Previous imaging studies often focused on seeking the causative tumors of TIO, not on the radiological features of these tumors, especially magnetic resonance imaging (MRI) features. PMT-MCT remains a largely misdiagnosed, ignored or unknown entity by most radiologists and clinicians. In the present case report, a review of the known literature of PMT-MCT was conducted and the CT and MRI findings from three patient cases were described for diagnosing the small subcutaneous tumor. Typical MRI appearances of PMT-MCT were isointense relative to the muscles on T1-weighted imaging, and markedly hyperintense on T2-weighted imaging containing variably flow voids, with markedly heterogeneous/homogenous enhancement on post contrast T1-weighted fat-suppression imaging. Short time inversion recovery was demonstrated to be the optimal sequence in localizing the tumor.

Inflammatory Pseudotumor-Like Follicular Dendritic Cell Sarcoma of the Spleen: Computed Tomography Imaging Characteristics in 5 Patients.

OBJECTIVES: The purpose of this study was to retrospectively review the computed tomography (CT) and clinicopathological characteristics of inflammatory pseudotumor (IPT)-like follicular dendritic cell sarcoma (FDCS) of the spleen in 5 patients. METHODS: Clinical, pathologic, and CT imaging findings of 5 patients with IPT-like FDCS of the spleen were reviewed and analyzed. Computed tomography imaging and pathologic features were compared. RESULTS: Abdominal unenhanced CT revealed a well-defined hypodense mass in the spleen with complex internal architecture with focal necrosis and/or speckle-strip calcification. On postcontrast CT, slightly delayed enhancement was observed in 5 cases. Four patients had a normalized spleen. The fourth patient had lung metastasis. The fifth patient had 2 relatively small lesions as well as metastases to the spine. CONCLUSIONS: Computed tomography imaging features of IPT-like FDCS of the spleen are distinctly different from other hypovascular splenic neoplasm; however, the definitive diagnosis requires further confirmation with needle biopsy or surgery. Inflammatory pseudotumor-like FDCS of the spleen should be suggested by using the CT imaging features of the splenic mass with evidence of metastatic disease.

Clinical and computed tomography imaging features of renal medullary carcinoma: A report of six cases.

Patients with renal medullary carcinoma (RMC) have a poor prognosis, usually due to late diagnosis. Computed tomography (CT) analysis may aid the differentiation between RMC and other types of renal cell carcinoma, in order to establish an accurate early diagnosis. There is a limited number of reports in the literature focusing on clinical and multi-slice CT (MSCT) imaging findings of RMC. Consequently, the present study aimed to characterize the clinical and MSCT imaging features of RMC. For this purpose, the MSCT imaging findings of 6 patients with RMC were retrospectively studied. The patients were subjected to MSCT in order to investigate the characteristics of the tumors, including location, size, density, calcification, cystic or solid appearance, capsule sign, enhancement pattern and presence of retroperitoneal lymph node metastasis. The tumors in the current study presented a mean diameter of 7.48±3.25 cm, and were observed to be solitary and heterogeneous with necrotic components. The majority of the tumors did not contain calcifications (5/6); displayed an ill-defined margin (4/6); were centered in the medulla; extended into the renal pelvis or peripelvic tissues (6/6); and did not exhibit a fibrous capsule. Localized caliectasis was observed in 3 of the 6 cases. The attenuation of the solid region of the RMC on unenhanced CT was equal to that of the renal cortex or medulla (42.3±2.7 vs. 40.7±3.6 and 41.2±3.9 Hounsfield units, respectively; P>0.05) while, on enhanced CT, the enhancement of the tumor was lower than that of the normal renal cortex and medulla during all phases (cortical phase, 52.6±4.8 vs. l99.5±9.7 and 72.7±6.4; medullary phase, 58.6±5.7 vs. 184.6±10.8 and 93.5±7.8; delayed phase, 56.8±6.1 vs. 175.7±8.5 and 96.5±7.9, respectively; P<0.05). In conclusion, RMC tends to be an infiltrative, ill-defined heterogeneous mass with intratumoral necrosis, which arises from the renal medulla, and displays lower enhancement than the renal cortex and medulla during all phases on enhanced CT. Despite its rarity in adults, RMC should be included in a differential diagnosis when CT imaging reveals these features.

Intramastoid solitary neurofibroma mimicking middle ear carcinoma.

We report an extremely rare case of intramastoid neurofibroma. A mass with destruction of the mastoid bone of a 51-year-old woman was examined with computed tomography. Subsequent magnetic resonance imaging demonstrated an ill-defined soft tissue mass with the opacification of mastoid air cells that had a mass effect in the same area. The patient underwent left subtotal temporal bone resection, and histological and immunohistochemical findings confirmed the lesion to be a neurofibroma. Given that similar imaging features of neurofibroma have been reported previously elsewhere in the head/neck and extremities, we suggest that it may be possible to include this tumor in the preoperative differential diagnosis.

Computed tomography imaging characteristics of synchronous gastrointestinal stromal tumors in patients with gastric cancer and correlation with clinicopathological findings.

This study was conducted to analyze the computed tomography (CT) imaging and clinicopathological characteristics of synchronous gastrointestinal stromal tumors (GISTs) in patients with gastric cancer. We retrospectively reviewed the CT images in 18 cases of immunohistochemically confirmed synchronous GISTs in patients with primary gastric cancer. Relevant histology and clinical data were also obtained. Multi-slice CT (MSCT) identified suspected synchronous GISTs in 26 patients, of which 18 were finally confirmed. Of the 18 patients, 12 were male and 6 female, with a mean age of 69.2 years. All 18 lesions were discovered as a solitary mass, ranging in size from 1.0 to 6.5 cm (mean, 2.2 cm). The lesions were predominantly located in the gastric fundus and were characterized by an ovoid shape, well-defined margins and isodensity, with slight to moderate gradual enhancement on MSCT. There was no evidence of necrosis, hemorrhage or cystic degeneration, irrespective of the tumor size. Synchronous GISTs exhibited spindle cell morphological characteristics and specific immunohistochemical properties, and were classified as being of low or very low malignant potential. In conclusion, since gastric cancer patients with a synchronous GIST are rare, a solitary, ovoid and well-defined mass with slight to moderate gradual enhancement following contrast agent injection may be suggestive of this diagnosis, although detection of synchronous GISTs does not appear to significantly affect therapy and outcome.