Anatomical Partition-Based Deep Learning: An Automatic Nasopharyngeal MRI Recognition Scheme.

BACKGROUND: Training deep learning (DL) models to automatically recognize diseases in nasopharyngeal MRI is a challenging task, and optimizing the performance of DL models is difficult. PURPOSE: To develop a method of training anatomical partition-based DL model which integrates knowledge of clinical anatomical regions in otorhinolaryngology to automatically recognize diseases in nasopharyngeal MRI. STUDY TYPE: Single-center retrospective study. POPULATION: A total of 2485 patients with nasopharyngeal diseases (age range 14-82 years, female, 779[31.3%]) and 600 people with normal nasopharynx (age range 18-78 years, female, 281[46.8%]) were included. SEQUENCE: 3.0 T; T2WI fast spin-echo sequence. ASSESSMENT: Full images (512 × 512) of 3085 patients constituted 100% of the dataset, 50% and 25% of which were randomly retained as two new datasets. Two new series of images (seg112 image [112 × 112] and seg224 image [224 × 224]) were automatically generated by a segmentation model. Four pretrained neural networks for nasopharyngeal diseases classification were trained under the nine datasets (full image, seg112 image, and seg224 image, each with 100% dataset, 50% dataset, and 25% dataset). STATISTICAL TESTS: The receiver operating characteristic curve was used to evaluate the performance of the models. Analysis of variance was used to compare the performance of the models built with different datasets. Statistical significance was set at P < 0.05. RESULTS: When the 100% dataset was used for training, the performances of the models trained with the seg112 images (average area under the curve [aAUC] 0.949 ± 0.052), seg224 images (aAUC 0.948 ± 0.053), and full images (aAUC 0.935 ± 0.053) were similar (P = 0.611). When the 25% dataset was used for training, the mean aAUC of the models that were trained with seg112 images (0.823 ± 0.116) and seg224 images (0.765 ± 0.155) was significantly higher than the models that were trained with full images (0.640 ± 0.154). DATA CONCLUSION: The proposed method can potentially improve the performance of the DL model for automatic recognition of diseases in nasopharyngeal MRI. LEVEL OF EVIDENCE: 4 TECHNICAL EFFICACY STAGE: 1.

Research progress on alternative non-classical mechanisms of PCSK9 in atherosclerosis in patients with and without diabetes.

The proprotein convertase subtilisin/kexin type 9 (PCSK9) acts via a canonical pathway to regulate circulating low-density lipoprotein-cholesterol (LDL-C) via degradation of the LDL receptor (LDLR) on the liver cell surface. Published research has shown that PCSK9 is involved in atherosclerosis via a variety of non-classical mechanisms that involve lysosomal, inflammatory, apoptotic, mitochondrial, and immune pathways. In this review paper, we summarized these additional mechanisms and described how anti-PCSK9 therapy exerts effects through these mechanisms. These additional pathways further illustrate the regulatory role of PCSK9 in atherosclerosis and offer an in-depth interpretation of how the PCSK9 inhibitor exerts effects on the treatment of atherosclerosis.

[Distribution Characteristics of Rh Phenotype and Feasibility of Compatible Blood Transfusion in Pregnant and Postpartum Women].

OBJECTIVE: To analyze the distribution characteristics of Rh phenotype in pregnant and postpartum women in Chongqing area, and to explore the clinical significance of Rh phenotype in pregnant and postpartum women and the feasibility of Rh phenotype compatible blood transfusion. METHODS: The ABO blood group and Rh phenotype of 65 161 pregnant and postpartum women were detected by microcolumn gel method, and 48 122 males in the same period were taken as controls. The data were analyzed by Chi-square test. RESULTS: There were 112 870 cases (99.64%) of RhD+ in 113 283 samples. In RhD+ cases, CCDee (48.39%) and CcDEe (32.88%) were the main phenotypes. The first case of D-- phenotype in Chongqing area was detected. 413 cases (0.36%) of RhD- were detected, with ccdee (52.78%) and Ccdee (33.41%) as the main phenotypes. Compared with RhD- group, RhD+ group showed statistically significant difference in Rh phenotype distribution (P < 0.01). Among 65 161 maternal samples, the positive rate of 5 antigens of Rh blood group from high to low was D > e > C > c > E, and there was no significant difference compared with male samples (P >0.05). There was no significant difference in the distribution of Rh phenotype between males and pregnant/postpartum women, as well as between pregnant/postpartum women with different ABO blood groups (P >0.05). In pregnant and postpartum women, there was no significant difference in distribution of Rh phenotype among the normal pregnancy population, the population with adverse pregnancy history, the population using human assisted reproductive technology (ART) and the population with infertility (P >0.05). There was no significant difference in the distribution of Rh phenotype between the 4 populations mentioned above and the inpatients in the local general Grade A hospitals and the blood donors (P >0.05). In RhD positive pregnant and postpartum women, the probability of finding compatible blood for CcDEe phenotype was 100%, the probability of finding compatible blood for CCDee, CcDee and CCDEe phenotypes was 45%-60%, the probability of finding compatible blood for ccDEE, ccDEe and CcDEE phenotypes was 5%-10%, and the probability of finding compatible blood for other phenotypes was lower than 0.5%. The supply of blood with CCDee and ccDEE phenotypes can meet the compatible transfusions requirements of 7 Rh phenotypes in more than 99% of patients. CONCLUSION: Rh phenotype detection should be carried out for pregnant and postpartum women, and it is feasible to carry out Rh phenotype-matched or compatible blood transfusion for pregnant and postpartum women who need blood transfusion.

[Serological Characteristics and Clinical Significance of Irregular Antibodies in Pregnant Women].

OBJECTIVE: To understand the serological characteristics of irregular antibodies in pregnant women and explore their clinical significance. METHODS: From January 2017 to March 2022, 151 471 pregnant women in Women and Children's Hospital of Chongqing Medical University were enrolled in this study, microcolumn gel card test was used for irregular antibody screening, and antibody specificity identification was further performed in some antibody-positive subjects. RESULTS: The positive rate of irregular antibody screening in the enrolled pregnant women was 0.91% (1 375/151 471), 0.23% (355/151 471) was detected in the first trimester, 0.05% (71/151 471) in the second trimester, and 0.63% (949/151 471) in the third trimester. The positive rate of irregular antibody screening in the third trimester was significantly higher than that in the first and second trimester, and a significant increase in the number of positive cases was found in the third trimester than that in the second trimester. The analysis of agglutination intensity of 1 375 irregular antibody screening positive results showed that the weakly positive agglutination intensity accounted for 50.11% (689/ 1 375), which was the highest, the suspicious positive was 18.69% (257/1 375), and the positive was 31.20% (429/1 375). The significant difference in distribution of agglutination intensity was not observed between the first trimester group and the second trimester group, however, in the third trimester, the proportion of suspicious positive and weakly positive was lower than the first trimester, while, the proportion of positive was higher than the first trimester, and the difference was statistically significant (P < 0.001). Among the irregular antibody screening positive pregnant women, the proportion of pregnant women with pregnancy number ≥ 2 was significantly higher than that with pregnancy ≤ 1. Among 60 pregnant women who underwent antibody identification, the distributions of the antibodies were as follows: Rh blood group system accounted for 23.33% (14/60), Lewis system 43.33% (26/60), Kidd system 3.33% (2/60), MNS system 16.67% (10/60), P1PK system 1.67% (1/60), autoantibodies 1.67% (1/60), and 4 cases was unable to identify (6.67%, 4/60). Among specific antibodies, the anti-Lea was the most common (30.00%), followed by anti-E (16.67%) and anti-M (16.67%). CONCLUSION: The differences of irregular antibody serological characteristics exist in pregnant women from different regions with different genetic backgrounds, understanding the characteristics of irregular antibody in local pregnant women is of great significance for ensuring transfusion safety in pregnant women and preventing hemolytic disease of newborn.

Predicting prognosis of nasopharyngeal carcinoma based on deep learning: peritumoral region should be valued.

BACKGROUND: The purpose of this study was to explore whether incorporating the peritumoral region to train deep neural networks could improve the performance of the models for predicting the prognosis of NPC. METHODS: A total of 381 NPC patients who were divided into high- and low-risk groups according to progression-free survival were retrospectively included. Deeplab v3 and U-Net were trained to build segmentation models for the automatic segmentation of the tumor and suspicious lymph nodes. Five datasets were constructed by expanding 5, 10, 20, 40, and 60 pixels outward from the edge of the automatically segmented region. Inception-Resnet-V2, ECA-ResNet50t, EfficientNet-B3, and EfficientNet-B0 were trained with the original, segmented, and the five new constructed datasets to establish the classification models. The receiver operating characteristic curve was used to evaluate the performance of each model. RESULTS: The Dice coefficients of Deeplab v3 and U-Net were 0.741(95%CI:0.722-0.760) and 0.737(95%CI:0.720-0.754), respectively. The average areas under the curve (aAUCs) of deep learning models for classification trained with the original and segmented images and with images expanded by 5, 10, 20, 40, and 60 pixels were 0.717 ± 0.043, 0.739 ± 0.016, 0.760 ± 0.010, 0.768 ± 0.018, 0.802 ± 0.013, 0.782 ± 0.039, and 0.753 ± 0.014, respectively. The models trained with the images expanded by 20 pixels obtained the best performance. CONCLUSIONS: The peritumoral region NPC contains information related to prognosis, and the incorporation of this region could improve the performance of deep learning models for prognosis prediction.

Giant Cell Tumor and Giant Cell Reparative Granuloma of Bone of the Head: CT and MR Imaging Findings.

BACKGROUND: This study aimed to determine the features and differentiation of Giant Cell Reparative Granuloma (GCRG) and Giant Cell Tumor (GCT) of the head on CT and MRI. METHODS: This retrospective study included six patients with histopathology-confirmed head GCRG and 5 patients with histopathology-confirmed head GCT. All images were independently reviewed by two radiologists. The growth pattern, bone changes, MRI signal intensity, enhancement patterns and other image features were recorded. All patients received CT scans and MR images. RESULTS: All the lesions were located centrally in the bone. Osteolytic bone destruction and expansive growth patterns were observed on CT images. Four of six cases broke the cortical bone with residual cortical bone, and the last two showed a thin cortex in GCRG. Five cases broke the cortical bone with residual cortical bone in GCT. There were enhancing septations in GCT lesions on contrast- enhanced T1-Weighted Images (T1WI) while enhancing septations were not present in GCRG cases. The size of GCT lesions was larger than that of GRCG. GCRG and GCT showed iso-low signals on T1WI and iso-high signals on T2-Weighted Images (T2WI). There was a case with cystic or necrotic lesions in each of the two types of lesions. Osteolytic bone destruction and expansive growth patterns were observed in GCTs and GCRGs. CONCLUSION: The size of the GRCG lesion was smaller than that of the GCT. The presence of enhancing septations and the size of the lesion may distinguish GCTs from GCRG.

Deep learning for the prediction of residual tumor after radiotherapy and treatment decision-making in patients with nasopharyngeal carcinoma based on magnetic resonance imaging.

Background: Concurrent chemoradiotherapy (CCRT) and induction chemotherapy (IC) plus CCRT (IC + CCRT) are the main treatments for patients with advanced nasopharyngeal carcinoma (NPC). We aimed to develop deep learning (DL) models using magnetic resonance (MR) imaging to predict the risk of residual tumor after each of the 2 treatments and to provide a reference for patients to select the best treatment option. Methods: A retrospective study was conducted on 424 patients with locoregionally advanced NPC who underwent CCRT or IC + CCRT between June 2012 and June 2019 in the Renmin Hospital of Wuhan University. According to the evaluation of MR images taken 3 to 6 months after radiotherapy, patients were divided into 2 categories: residual tumor and non-residual tumor. Transferred U-net and Deeplabv3 neural networks were trained, and the better-performance segmentation model was used to segment the tumor area on axial T1-weighted enhanced MR images. Then, 4 pretrained neural networks for prediction of residual tumors were trained with CCRT and IC + CCRT datasets, and the performances of the models trained using each image and each patient as a unit were evaluated. Patients in the test cohort of CCRT and IC + CCRT datasets were successively classified by the trained CCRT and IC + CCRT models. Model recommendations were formed according to the classification and compared with the treatment decisions of physicians. Results: The Dice coefficient of Deeplabv3 (0.752) was higher than that of U-net (0.689). The average area under the curve (aAUC) of the 4 networks was 0.728 for the CCRT and 0.828 for the IC + CCRT models trained using a single image as a unit, whereas the aAUC for models trained using each patient as a unit was 0.928 for the CCRT and 0.915 for the IC + CCRT models, respectively. The accuracy of the model recommendation and the decision of physicians was 84.06% and 60.00%, respectively. Conclusions: The proposed method can effectively predict the residual tumor status of patients after CCRT and IC + CCRT. Recommendations based on the model prediction results can protect some patients from receiving additional IC and improve the survival rate of patients with NPC.

Deep learning for locally advanced nasopharyngeal carcinoma prognostication based on pre- and post-treatment MRI.

PURPOSE: We aimed to predict the prognosis of advanced nasopharyngeal carcinoma (stage Ⅲ-Ⅳa) using Pre- and Post-treatment MR images based on deep learning (DL). METHODS: A total of 206 patients with primary nasopharyngeal carcinoma who were diagnosed and treated at the Renmin Hospital of Wuhan University between June 2012 and January 2018 were retrospectively selected. A rectangular region of interest (ROI), which included the tumor area, surrounding tissues and organs, was delineated on each Pre- and Post-treatment MR image. Two Inception-Resnet-V2 based transfer learning models, named Pre-model and Post-model, were trained with the Pre-treatment images and the Post-treatment images, respectively. In addition, an ensemble learning model based on the Pre-model and Post-models was established. The three established models were evaluated by receiver operating characteristic curve (ROC), confusion matrix, and Harrell's concordance indices (C-index). High-risk-related gradient-weighted class activation mapping (Grad-CAM) images were developed according to the DL models. RESULTS: The Pre-model, Post-model, and ensemble model displayed a C-index of 0.717 (95% CI: 0.639 to 0.795), 0.811 (95% CI: 0.745-0.877), 0.830 (95% CI: 0.767-0.893), and AUC of 0.741 (95% CI: 0.584-0.900), 0.806 (95% CI: 0.670-0.942), and 0.842 (95% CI: 0.718-0.967) for the test cohort, respectively. In comparison with the models, the performance of Post-model was better than the performance of Pre-model, which indicated the importance of Post-treatment images for prognosis prediction. All three DL models performed better than the TNM staging system (0.723, 95% CI: 0.567-0.879). The captured features presented on Grad-CAM images suggested that the areas around the tumor and lymph nodes were related to the prognosis of the tumor. CONCLUSIONS: The three established DL models based on Pre- and Post-treatment MR images have a better performance than TNM staging. Post-treatment MR images are of great significance for prognosis prediction and could contribute to clinical decision-making.

Factors affecting recurrent positive RT-PCR results in clinically cured COVID-19 patients: a multicenter study.

The aim of this study was to evaluate factors affecting the recurrence of positive RT-PCR results. By performing a retrospective analysis, we evaluated the clinical data of recurrent positive coronavirus disease 2019 (COVID-19) patients in multiple medical institutions in Wuhan. We recruited COVID-19 patients who were hospitalized from January 1 to March 10, 2020, in three tertiary hospitals in Wuhan, met the discharge criteria and received at least one additional nucleic acid test before leaving the hospital. According to the RT-PCR results, patients were split into a recurrent positive group (RPos group) and a nonrecurrent positive group (non-RPos group). Clinical characteristics, therapeutic schedules and antibody titers were compared between the two groups. AI-assisted chest high-resolution computed tomography (HRCT) technology was applied to investigate pulmonary inflammatory exudation and compare the extent of lung areas with different densities. This study involved 122 COVID-19 patients. There were no significant differences in age, sex, preexisting diseases, clinical symptoms, clinical classification, course of disease, therapeutic schedules or serum-specific antibodies between the two groups. A higher proportion of patients who showed pulmonary inflammatory exudation on HRCT scans were recurrent positive at the time of discharge than other patients (81.6% vs 13.7%, P < 0.01). In addition, the degree of pulmonary fibrosis was higher in the RPos group than in the non-RPos group (P < 0.05). Subpleural exudation at the peripheral edge of the lung and extensive pulmonary fibrosis at the time of discharge represent risk factors for the recurrence of COVID-19.

Two kinds of common prenatal screening tests for Down's syndrome: a systematic review and meta-analysis.

As the chromosomal examination of foetal cells for the prenatal diagnosis of Down's syndrome (DS) carries a risk of inducing miscarriage, serum screening tests are commonly used before invasive procedures. In this study, a total of 374 records from PubMed, EMBASE, and the ISI Science Citation Index databases were reviewed. As a result of duplication, insufficient data, and inappropriate article types, 18 independent articles containing 183,998 samples were used in the final systematic review and meta-analysis of the diagnostic performance of the serum triple screening test (STS) and the integrated screening test (INS). Data extracted from the selected studies were statistically analysed, and the presence of heterogeneity and publication bias was assessed using specific software. The overall sensitivity, specificity, positive likelihood ratio, negative likelihood ratio, diagnostic odds ratio, and the area under the curve for the STS were 0.77 (95% confidence interval = 0.73-0.81), 0.94 (0.94-0.94), 9.78 (6.87-13.93), 0.26 (0.22-0.31), 44.72 (30.77-65.01), and 0.9064, respectively. For the INS, these values were 0.93 (0.90-0.95), 0.93 (0.93-0.93), 22.38 (12.47-40.14), 0.08 (0.05-0.11), 289.81 (169.08-496.76), and 0.9781, respectively. These results indicate that the INS exhibits better diagnostic value for DS. However, further research is needed to identify other biomarkers to improve prenatal screening tests.

[Relationship between bacterial biofilm and chronic rhinosinusitis: a systematic review].

OBJECTIVE: To evaluate the relationship between bacterial biofilm (BBF) and chronic rhinosinusitis (CRS). METHODS: The database on line was searched to collect the studies on BBF and CRS. The method of meta analysis was used to analyze the data of suitable studies. RESULTS: Fourteen studies were included. System evaluation indicated that the BBF detection rate in CRS group was significantly higher than that in the control group (OR = 17.01, P < 0.01), and the nasal surgery's rate of BBF positive group was significantly higher than the negative group (OR = 3.99, P < 0.01). Preoperative Lund-Kennedy endoscopic score, Lund-MacKay CT score, symptom severity score, postoperative Lund-Kennedy score after six months showed no difference between BBF positive group and negative group. CONCLUSION: The presence of BBF is related to the pathogenesis of CRS and the history of nasal surgery.