"Hot Cross Bun" Sign in a Patient with Glutamic Acid Decarboxylase 65-KDa Isoform Associated Cerebellar Ataxia: Case Report and Review of the Literature.

The "hot cross bun" sign (HCBs) is a cruciform hyperintensity on T2-weighted imaging within the pons initially found in patients diagnosed as multiple system atrophy. However, recent findings have broadened the disease spectrum presented with HCBs. Here is a case report at an academic medical center. Cerebral magnetic resonance imaging (MRI), electroneuromyography, serum, and CSF analysis were performed. Literature is comprehensively reviewed. We investigated a woman presented with blurred speech and cerebellar ataxia. Her MRI showed the vertical line of HCBs 2 weeks after disease onset and gradually enhanced, presenting as an intact HCBs in a year. Glutamic acid decarboxylase 65-kDa isoform (GAD65) antibody IgG was detected in serum and CSF. The patient was diagnosed as GAD65 associated cerebellar ataxia and treated with corticosteroid and rituximab. We found 6 previously reported autoimmune cerebellar ataxia patients with HCBs. Anti-KLHL-11, anti-Homer-3, anti-Ri, and anti-Amphiphysin were associated. All patients had cerebellar ataxia with other neurological symptoms. Five patients were diagnosed with tumor. First-line immunotherapy including corticosteroid, intravenous immunoglobulin, and plasma exchange for most patients was unsatisfied. This case highlights the importance of considering GAD65 IgG evaluation in patients with progressive cerebellar syndrome and HCBs. Early diagnosis and therapy are challenging but imperative. Further studies are required in regard to therapeutic management.

An interpretable transformer network for the retinal disease classification using optical coherence tomography.

Retinal illnesses such as age-related macular degeneration and diabetic macular edema will lead to irreversible blindness. With optical coherence tomography (OCT), doctors are able to see cross-sections of the retinal layers and provide patients with a diagnosis. Manual reading of OCT images is time-consuming, labor-intensive and even error-prone. Computer-aided diagnosis algorithms improve efficiency by automatically analyzing and diagnosing retinal OCT images. However, the accuracy and interpretability of these algorithms can be further improved through effective feature extraction, loss optimization and visualization analysis. In this paper, we propose an interpretable Swin-Poly Transformer network for performing automatically retinal OCT image classification. By shifting the window partition, the Swin-Poly Transformer constructs connections between neighboring non-overlapping windows in the previous layer and thus has the flexibility to model multi-scale features. Besides, the Swin-Poly Transformer modifies the importance of polynomial bases to refine cross entropy for better retinal OCT image classification. In addition, the proposed method also provides confidence score maps, assisting medical practitioners to understand the models' decision-making process. Experiments in OCT2017 and OCT-C8 reveal that the proposed method outperforms both the convolutional neural network approach and ViT, with an accuracy of 99.80% and an AUC of 99.99%.

Case report: Intracranial lesions of cat-scratch disease mimicking an atypical meningioma.

Objectives: Cat-scratch disease (CSD) is an infectious disease caused by Bartonella henselae. The most typical symptom of patients with CSD is regional lymphadenopathy, while central nervous system lesions related to CSD are rare. Here, we present a case of an aged woman with CSD involving the dura mater with a manifestation similar to that of an atypical meningioma. Methods: The patient was followed up by our neurosurgery and radiology teams. Clinical information was recorded, and the pre- and post-operation CT results and magnetic resonance imaging (MRI) changes were collected. The paraffin-embedded tissue was sampled for the polymerase chain reaction (PCR) test. Results: In this study, we present the details of a 54 year-old Chinese woman admitted to our hospital with a paroxysmal headache for 2 years that had worsened in the past 3 months. Brain CT and MRI showed a meningioma-like lesion below the occipital plate. En bloc resection of the sinus junction area was performed. A pathological examination showed granulation tissue and fibrosis with acute and chronic inflammation, granuloma, and central stellate microabscess, which was suspected as the cat-scratch disease. The paraffin-embedded tissue was sampled for a polymerase chain reaction (PCR) test to amplify the corresponding pathogen gene sequence, which was Bartonella henselae. Conclusion: The case in our study underscores the fact that the incubation period of CSD may be very long. On the contrary, CSD can involve the meninges, resulting in tumor-like lesions.

Cancer detection for small-size and ambiguous tumors based on semantic FPN and transformer.

Early detection of tumors has great significance for formative detection and determination of treatment plans. However, cancer detection remains a challenging task due to the interference of diseased tissue, the diversity of mass scales, and the ambiguity of tumor boundaries. It is difficult to extract the features of small-sized tumors and tumor boundaries, so semantic information of high-level feature maps is needed to enrich the regional features and local attention features of tumors. To solve the problems of small tumor objects and lack of contextual features, this paper proposes a novel Semantic Pyramid Network with a Transformer Self-attention, named SPN-TS, for tumor detection. Specifically, the paper first designs a new Feature Pyramid Network in the feature extraction stage. It changes the traditional cross-layer connection scheme and focuses on enriching the features of small-sized tumor regions. Then, we introduce the transformer attention mechanism into the framework to learn the local feature of tumor boundaries. Extensive experimental evaluations were performed on the publicly available CBIS-DDSM dataset, which is a Curated Breast Imaging Subset of the Digital Database for Screening Mammography. The proposed method achieved better performance in these models, achieving 93.26% sensitivity, 95.26% specificity, 96.78% accuracy, and 87.27% Matthews Correlation Coefficient (MCC) value, respectively. The method can achieve the best detection performance by effectively solving the difficulties of small objects and boundaries ambiguity. The algorithm can further promote the detection of other diseases in the future, and also provide algorithmic references for the general object detection field.

Liver function indicators and risk of hepatocellular carcinoma: a bidirectional mendelian randomization study.

Observational studies have shown an association between liver dysfunction and hepatocellular carcinoma (HCC), but the causality relationship between them is unclear. We aimed to determine whether there is a bidirectional causal relationship between liver function indicators (alanine aminotransferase, ALT; aspartate aminotransferase, AST; alkaline phosphatase, ALP; γ-glutamyltransferase, GGT) and HCC. Our two-sample Mendelian randomization (MR) study acquired single nucleotide polymorphisms (SNPs) associated with liver function indicators (ALT, n = 134,182; AST, n = 134,154; GGT, n = 118,309; ALP, n = 105,030) and with HCC (n = 197,611) from publicly available genome-wide association studies (GWAS) of East Asian ancestry in Japan (BioBank Japan, BBJ). Univariable MR analyses were performed to identify whether the genetic evidence of exposure was significantly associated with outcome. Multivariable MR analysis was conducted to estimate the independent effects of exposures on outcome. Univariable MR analysis indicated that the level of ALT, AST, and GGT was the risk factor for HCC incidence. Meanwhile, multivariable MR analysis revealed that AST was an independent risk factor for HCC. The hazard ratio (HR) of the probability of HCC was 3.045 [95% confidence interval (95%CI), 1.697-5.463, p = 0.003] for AST. The results of reverse MR analyses showed that gene-predictive HCC incidence could increase the levels of AST (HR = 1.031, 95%CI: 1.009-1.054, p = 2.52 × 10-4) and ALT (HR = 1.040, 95%CI: 1.019-1.063, p = 0.005). Meanwhile, HCC may be negatively correlated with ALP levels (HR = 0.971, 95%CI: 0.947-0.995, p = 0.018). This study provides evidence to support that genetically predicted higher levels of AST are related to increased risk of HCC, with no strong evidence of a causal effect of genetically predicted ALP, ALP, and GGT on HCC. In addition, genetic predisposition to HCC could influence blood concentration of ALT, AST, and ALP. Thus, this may create a vicious cycle.

Predicting the grade of meningiomas by clinical-radiological features: A comparison of precontrast and postcontrast MRI.

Objectives: Postcontrast magnetic resonance imaging (MRI) is important for the differentiation between low-grade (WHO I) and high-grade (WHO II/III) meningiomas. However, nephrogenic systemic fibrosis and cerebral gadolinium deposition are major concerns for postcontrast MRI. This study aimed to develop and validate an accessible risk-scoring model for this differential diagnosis using the clinical characteristics and radiological features of precontrast MRI. Methods: From January 2019 to October 2021, a total of 231 meningioma patients (development cohort n = 137, low grade/high grade, 85/52; external validation cohort n = 94, low-grade/high-grade, 60/34) were retrospectively included. Fourteen types of demographic and radiological characteristics were evaluated by logistic regression analyses in the development cohort. The selected characteristics were applied to develop two distinguishing models using nomograms, based on full MRI and precontrast MRI. Their distinguishing performances were validated and compared using the external validation cohort. Results: One demographic characteristic (male), three precontrast MRI features (intratumoral cystic changes, lobulated and irregular shape, and peritumoral edema), and one postcontrast MRI feature (absence of a dural tail sign) were independent predictive factors for high-grade meningiomas. The area under the receiver operating characteristic (ROC) curve (AUC) values of the two distinguishing models (precontrast-postcontrast nomogram vs. precontrast nomogram) in the development cohort were 0.919 and 0.898 and in the validation cohort were 0.922 and 0.878. DeLong's test showed no statistical difference between the AUC values of the two distinguishing models (p = 0.101). Conclusions: An accessible risk-scoring model based on the demographic characteristics and radiological features of precontrast MRI is sufficient to distinguish between low-grade and high-grade meningiomas, with a performance equal to that of a full MRI, based on radiological features.

Evaluation of histopathological changes in the microstructure at the center and periphery of glioma tumors using diffusional kurtosis imaging.

OBJECTIVE: To explore the relationship between alterations in gliomas revealed by diffusional kurtosis imaging (DKI) and the histopathological microstructural changes. METHODS: Thirty-seven patients with cerebral gliomas underwent conventional MRI and DKI at 3.0T. Normalized fractional anisotropy (FA), mean diffusivity (MD) and mean kurtosis (MK) were compared in different regions of glioma tumors. Parameters with a high sensitivity and specificity regarding the discrimination of glioma grade were evaluated using receiver operating characteristic (ROC) curve analysis. Correlations between normalized FA, MD, and MK and histopathological findings (tumor cell density, total vascular area [TVA], and Ki-67 labeling index [LI]) were assessed using Pearson correlation analyses. RESULTS: Normalized FA, MD, and MK differed significantly between low-grade gliomas (LGGs) and high-grade gliomas (HGGs) (P=0.02, P=0.001 and P<0.001, respectively) at the center of the tumor. Normalized MK exhibited the highest sensitivity (80%) and specificity (100%) in distinguishing HGGs from LGGs. Relative to the tumor center, normalized MK was significantly increased in the tumor periphery (P<0.001) in LGGs and significantly decreased (P=0.002) in HGGs. The significant correlations were found between normalized MK and all histopathological findings (tumor cell density: r=0.596, P=0.006; TVA: r=0.764, P<0.001; and Ki-67 LI: r=0.766, P<0.001) among samples from the center of the tumor. CONCLUSION: DKI, especially concerning the MK parameter, demonstrated high sensitivity in the detection of microstructural changes in patients with brain gliomas.

MDCT Evaluation of Costal Bone Lesions: Comparison of Axial, Multiplanar, and 3D Volume-Rendered Images: A Retrospective Study.

The aim of this retrospective study was to compare accuracies of axial, multiplanar, and volume-rendered 3-dimensional (3D) images in the diagnosis of costal bone lesions.Forty-one patients, aged from 10 to 72-years old, with costal bone lesions underwent multidetector CT (MDCT). Axial, multiplanar, and 3D-volume-rendered images were reviewed by 3 reviewers for the property of the lesions (fracture, tumor, and tumor-like lesions or inflammation). In case of fracture, the diagnosis was demonstrated with the location of the fracture and the amounts of the costal bone involved. In case of a tumor or tumor-like lesions, the diagnosis was demonstrated pathological property. Final diagnosis was determined by biopsy or surgery. Diagnostic accuracy and interreviewers agreement were evaluated.For the diagnosis of fractures, average accuracy was 77%, 100%, and 100% for axial, multiplanar, and 3D-volume-rendered images, respectively. For the diagnosis of tumor and tumor-like lesions, average accuracy was 90% for axial, 96% for multiplanar, and 99% for 3D-volume-rendered images. For the diagnosis of inflammation lesions, average accuracy was 100% for all the 3 image formats. Interobserver agreement independence of imaging formats was high.Multiplanar and 3D-volume-rendered images were superior to axial images in diagnosis of fracture, tumor, and tumor-like lesions; however, for the evaluation of inflammation lesions, there were no difference by 3 image formats.

Improving depiction of temporal bone anatomy with low-radiation dose CT by an integrated circuit detector in pediatric patients: a preliminary study.

The purpose of this study was to determine the performance of low-dose computed tomography (CT) scanning with integrated circuit (IC) detector in defining fine structures of temporal bone in children by comparing with the conventional detector. The study was performed with the approval of our institutional review board and the patients' anonymity was maintained. A total of 86 children<3 years of age underwent imaging of temporal bone with low-dose CT (80 kV/150 mAs) equipped with either IC detector or conventional discrete circuit (DC) detector. The image noise was measured for quantitative analysis. Thirty-five structures of temporal bone were further assessed and rated by 2 radiologists for qualitative analysis. κ Statistics were performed to determine the agreement reached between the 2 radiologists on each image. Mann-Whitney U test was used to determine the difference in image quality between the 2 detector systems. Objective analysis showed that the image noise was significantly lower (P<0.001) with the IC detector than with the DC detector. The κ values for qualitative assessment of the 35 fine anatomical structures revealed high interobserver agreement. The delineation for 30 of the 35 landmarks (86%) with the IC detector was superior to that with the conventional DC detector (P<0.05) although there were no differences in the delineation of the remaining 5 structures (P>0.05). The low-dose CT images acquired with the IC detector provide better depiction of fine osseous structures of temporal bone than that with the conventional DC detector.

Preoperative evaluation of collateral venous anastomoses in meningioma involving cerebral venous sinus by susceptibility weighted imaging.

Precise preoperative identification of the collateral venous anastomoses is critical for proper surgical management of patients with meningioma involving sinus. This study was to assess the feasibility of susceptibility weighted imaging (SWI) to delineate the collateral venous anastomoses before surgery. Twenty-five patients with meningiomas that were involved in sinuses underwent surgery and the collateral anastomoses were evaluated with SWI and phase-contrast magnetic resonance venography (MRV) before surgery. The results obtained with SWI were compared with those obtained with MRV. Intraoperative findings were used as the gold standard. By surgery, a total of 98 collateral anastomotic veins were identified in the 25 patients. SWI depicted 85 collateral anastomotic veins close to the meningioma with a sensitivity of 87%, whereas MRV showed 57 collateral anastomotic veins with a sensitivity of 58%. The detectability of collateral anastomotic veins in SWI images was superior to MRV. The results suggest that SWI is superior to MRV and could provide more reliable information on the collateral venous anastomoses in patients with meningioma.

Oculomotor paralysis: 3D-CISS MR imaging with MPR in the evaluation of neuralgic manifestation and the adjacent structures.

PURPOSE: To evaluate the value of three-dimensional (3D) constructive interference in steady-state (CISS) magnetic resonance (MR) imaging with multi-planar reconstruction (MPR) in displaying the relationship between the oculomotor nerve and its adjacent structures for patients with oculomotor paralysis. MATERIALS AND METHODS: 17 consecutive patients with oculomotor paralysis were examined with 3D-CISS and conventional spin-echo (SE) sequences on a 1.5-Tesla MR system. Original transverse and MPR images were used for image interpretation. The features of the oculomotor nerve and its adjacent structures were identified. The diagnosis was surgically confirmed in all patients. RESULTS: Through 3D-CISS with MPR images, obvious relationship of the oculomotor nerve and its adjacent structures was demonstrated on 17 patients. Of those oculomotor nerves, 15 were compressed by the arteries (n=15), one by the craniopharyngioma (n=1), and another one by the neurofibroma (n=1). CONCLUSION: 3D-CISS MR imaging with MPR provides an excellent way to characterize the relationship between the nerve and its adjacent structures in the cisternal segment of the oculomotor nerve in the patients with oculomotor paralysis. Moreover, this method shows anatomical details for imaging diagnosis and surgical procedure.

Preoperative evaluation of venous systems with computed tomography venography in parasagittal meningiomas.

OBJECTIVE: Parasagittal meningiomas (PSM) may pose a difficult surgical challenge because venous patency and collateral anastomoses have to be clearly defined for correct surgical planning. The aim of this study was to study the diagnostic accuracy of computed tomography venography (CTV) in assessing the superior sagittal sinus (SSS) obstruction and venous collaterals. METHODS: Twenty-four patients undergoing surgery for meningiomas located adjacent to the SSS were prospectively evaluated. All the patients underwent both conventional computed tomography examination and CTV. Computed tomography venography was evaluated by 2 expert neuroradiologists to assess (1) patency of the sinus (patent/occluded), (2) the extent of occlusion (in centimeters), and (3) the number of collateral anastomoses close to the insertion of the meningioma. Computed tomography venography assessment was compared with surgery. All patients were operated on, and intraoperative findings were taken as the criterion standard. RESULTS: Computed tomography venography showed the sinus to be occluded in 6 cases. Computed tomography venography data were confirmed by surgery, showing 6 patients to have have an occlusion of the SSS. The CTV sensitivity was thus 100%. Computed tomography venography depicted 83% of collateral venous anastomoses close to the meningioma as subsequently confirmed by surgery. CONCLUSIONS: In the preoperative planning for patients with meningiomas located close to the SSS, CTV can provide additional and more reliable information concerning venous infiltration and the presence of collateral anastomoses compared with noncontrast computed tomography.