MEET: A Multi-Band EEG Transformer for Brain States Decoding.

OBJECTIVE: Electroencephalography (EEG) is among the most widely used and inexpensive neuroimaging techniques. Compared to the CNN or RNN based models, Transformer can better capture the temporal information in EEG signals and focus more on global features of the brain's functional activities. Importantly, according to the multiscale nature of EEG signals, it is crucial to consider the multi-band concept into the design of EEG Transformer architecture. METHODS: We propose a novel Multi-band EEG Transformer (MEET) to represent and analyze the multiscale temporal time series of human brain EEG signals. MEET mainly includes three parts: 1) transform the EEG signals into multi-band images, and preserve the 3D spatial information between electrodes; 2) design a Band Attention Block to compute the attention maps of the stacked multi-band images and infer the fused feature maps; 3) apply the Temporal Self-Attention and Spatial Self-Attention modules to extract the spatiotemporal features for the characterization and differentiation of multi-frame dynamic brain states. RESULTS: The experimental results show that: 1) MEET outperforms state-of-the-art methods on multiple open EEG datasets (SEED, SEED-IV, WM) for brain states classification; 2) MEET demonstrates that 5-bands fusion is the best integration strategy; and 3) MEET identifies interpretable brain attention regions. SIGNIFICANCE: MEET is an interpretable and universal model based on the multiband-multiscale characteristics of EEG. CONCLUSION: The innovative combination of band attention and temporal/spatial self-attention mechanisms in MEET achieves promising data-driven learning of the temporal dependencies and spatial relationships of EEG signals across the entire brain in a holistic and comprehensive fashion.

Differentiate preterm and term infant brains and characterize the corresponding biomarkers via DICCCOL-based multi-modality graph neural networks.

Preterm birth is a worldwide problem that affects infants throughout their lives significantly. Therefore, differentiating brain disorders, and further identifying and characterizing the corresponding biomarkers are key issues to investigate the effects of preterm birth, which facilitates the interventions for neuroprotection and improves outcomes of prematurity. Until now, many efforts have been made to study the effects of preterm birth; however, most of the studies merely focus on either functional or structural perspective. In addition, an effective framework not only jointly studies the brain function and structure at a group-level, but also retains the individual differences among the subjects. In this study, a novel dense individualized and common connectivity-based cortical landmarks (DICCCOL)-based multi-modality graph neural networks (DM-GNN) framework is proposed to differentiate preterm and term infant brains and characterize the corresponding biomarkers. This framework adopts the DICCCOL system as the initialized graph node of GNN for each subject, utilizing both functional and structural profiles and effectively retaining the individual differences. To be specific, functional magnetic resonance imaging (fMRI) of the brain provides the features for the graph nodes, and brain fiber connectivity is utilized as the structural representation of the graph edges. Self-attention graph pooling (SAGPOOL)-based GNN is then applied to jointly study the function and structure of the brain and identify the biomarkers. Our results successfully demonstrate that the proposed framework can effectively differentiate the preterm and term infant brains. Furthermore, the self-attention-based mechanism can accurately calculate the attention score and recognize the most significant biomarkers. In this study, not only 87.6% classification accuracy is observed for the developing Human Connectome Project (dHCP) dataset, but also distinguishing features are explored and extracted. Our study provides a novel and uniform framework to differentiate brain disorders and characterize the corresponding biomarkers.

Pure large cell neuroendocrine carcinoma originating from the endometrium: A case report.

BACKGROUND: Large cell neuroendocrine carcinoma (LCNEC) of the endometrium is an uncommon and highly aggressive tumor that has not been comprehensively characterized. We report a case of pure endometrial LCNEC and review the current literature of similar cases to raise awareness of the histological features, treatment, and prognosis of this tumor. CASE SUMMARY: We report the case of a 73-year-old woman who presented with irregular postmenopausal vaginal bleeding. Ultrasonography showed an enlarged uterus and a 5.1 cm × 3.3 cm area of medium and low echogenicity in the uterine cavity. Biopsy by dilatation and curettage suggested poorly differentiated carcinoma. Magnetic resonance imaging revealed a heterogeneously enhanced uterine tumor with diffuse infiltration of the posterior wall of the uterine myometrium and enlarged pelvic lymph nodes. The patient underwent a hysterectomy and bilateral adnexal resection. Gross observation revealed an ill-defined white solid mass of the posterior wall of the uterus infiltrating into the serosa with multiple solid nodules on the serous surface. Microscopically, the tumor cells showed neuroendocrine morphology (organoid nesting). Immunohistochemistry revealed the tumor cells were diffusely positive for the neuroendocrine markers CD56, chromogranin A, and synaptophysin. Thus, the tumor was diagnosed as stage IIIC endometrial LCNEC. CONCLUSION: Pathologic findings and immunohistochemistry are essential in making a diagnosis of endometrial LCNEC.

Chondrosarcoma of the para-acetabulum: correlation of imaging features with histopathological grade.

OBJECTIVE: To evaluate the computed tomography (CT) and magnetic resonance (MR) imaging features of para-acetabular chondrosarcoma (CS) and assess the difference between low-grade CS (LGCS) and high-grade CS (HGCS). MATERIALS AND METHODS: Thirty-one patients with histopathologically confirmed central para-acetabular CSs (6 LGCS and 25 HGCS) were retrospectively reviewed. Image features were evaluated for the following: cortical destruction, tumor border and pattern, calcification mode, soft-tissue mass, density/signal intensity, peritumoral edema, acetabular (cartilage) destruction, diffuse signal changes in acetabulum, mass inside hip joint, femoral head involvement, enhancement manifestations and the maximum length of the tumor. These image features between LGCS and HGCS were also assessed. RESULTS: The most common CT and/or MR findings included cortical destruction, punctate, ring-and-arc and linear calcification, soft-tissue mass, lobulated border, high signal intensity with low signal septa on T2-weighted image, peritumoral edema, hip joint infiltration, peripheral and septal enhancement on post-enhanced MR image. Statistical analysis showed that the image features, such as cortical destruction, soft-tissue mass, hip joint infiltration and tumor size were significantly different between LGCS and HGCS (p < 0.05). CONCLUSION: The characteristic radiological features of para-acetabular CSs are osteolytic lesions with cortical destruction, soft-tissue mass, lobulated border, calcification, and high signal intensity with low signal septa on T2-weighted MR image, peripheral and septal enhancement on post-enhanced MR image. Cortical destruction, soft-tissue mass, hip joint infiltration and tumor size can differentiate HGCS from LGCS.

Clinical manifestations of dermatomyositis and clinically amyopathic dermatomyositis patients with positive expression of anti-melanoma differentiation-associated gene 5 antibody.

OBJECTIVE: To investigate the clinical features of dermatomyositis (DM) and clinically amyopathic DM (CADM) patients with the presence of anti-melanoma differentiation-associated gene 5 (anti-MDA-5) antibodies. METHODS: We screened the serum anti-MDA-5 antibody levels of 140 patients with various connective tissue diseases (CTDs), including 32 with DM and 32 with CADM, or idiopathic pulmonary fibrosis (IPF). The clinical courses of DM/CADM patients with a positive expression of anti-MDA-5 antibodies were delineated. RESULTS: Anti-MDA-5 antibodies were detected at a significantly higher frequency in CADM patients than in DM patients (12 of 32 versus 3 of 32; P = 0.016), but were not detected in patients with other CTDs or IPF and healthy controls. Patients with a positive expression of anti-MDA-5 antibodies developed significantly more skin ulcerations (12 of 15 versus 4 of 49; P < 0.001) and interstitial lung disease (ILD; 15 of 15 versus 31 of 49 [P = 0.003]) than those without anti-MDA-5 antibodies. High-resolution computed tomography scores of the MDA-5-positive subset were increased compared with the MDA-5-negative group (mean ± SD 117.7 ± 76.3 versus 54.4 ± 50.7; P = 0.004), and the scores correlated well with anti-MDA-5 antibody levels (r(2) = 0.582, P = 0.029). The respiratory symptoms as well as skin ulcerations were dramatically improved in patients with anti-MDA-5 antibody levels <500 units/ml after treatment, whereas patients with anti-MDA-5 antibody levels >500 units/ml were resistant to the treatment and died of respiratory failure in a short period of time. CONCLUSION: Anti-MDA-5 antibody levels closely correlate with the severity of skin ulcerations, ILD, and the prognosis of the disease. Dynamic observation of serum anti-MDA-5 antibody levels would be helpful in predicting the course of ILD and facilitating better therapeutic targeting.