Res-TransNet: A Hybrid deep Learning Network for Predicting Pathological Subtypes of lung Adenocarcinoma in CT Images.

This study aims to develop a CT-based hybrid deep learning network to predict pathological subtypes of early-stage lung adenocarcinoma by integrating residual network (ResNet) with Vision Transformer (ViT). A total of 1411 pathologically confirmed ground-glass nodules (GGNs) retrospectively collected from two centers were used as internal and external validation sets for model development. 3D ResNet and ViT were applied to investigate two deep learning frameworks to classify three subtypes of lung adenocarcinoma namely invasive adenocarcinoma (IAC), minimally invasive adenocarcinoma and adenocarcinoma in situ, respectively. To further improve the model performance, four Res-TransNet based models were proposed by integrating ResNet and ViT with different ensemble learning strategies. Two classification tasks involving predicting IAC from Non-IAC (Task1) and classifying three subtypes (Task2) were designed and conducted in this study. For Task 1, the optimal Res-TransNet model yielded area under the receiver operating characteristic curve (AUC) values of 0.986 and 0.933 on internal and external validation sets, which were significantly higher than that of ResNet and ViT models (p < 0.05). For Task 2, the optimal fusion model generated the accuracy and weighted F1 score of 68.3% and 66.1% on the external validation set. The experimental results demonstrate that Res-TransNet can significantly increase the classification performance compared with the two basic models and have the potential to assist radiologists in precision diagnosis.

Diagnostic value of combining preoperative inflammatory markers ratios with CA199 for patients with early-stage pancreatic cancer.

BACKGROUND: An early diagnosis of pancreatic cancer (PC) is extremely difficult because of the lack of sensitive liquid biopsy methods and effective biomarkers. We attempted to evaluate whether circulating inflammatory marker could complement CA199 for the detection of early-stage PC. METHODS: We enrolled 430 patients with early-stage PC, 287 patients with other pancreatic tumors (OPT), and 401 healthy controls (HC). The patients and HC were randomly divided into a training set (n = 872) and two testing sets (n1 = 218, n2 = 28). The receiver operating characteristic (ROC) curves were investigated to evaluate the diagnostic performance of circulating inflammatory markers ratios, CA199, and combinations of the markers ratios in the training set, which would then be validated in the two testing sets. RESULTS: Circulating fibrinogen, neutrophils, and monocytes in patients with PC were significantly higher while circulating albumin, prealbumin, lymphocytes, and platelets of patients with PC were significantly lower compared to those of HC and OPT (all P < 0.05). The fibrinogen-to-albumin (FAR), fibrinogen-to-prealbumin (FPR), neutrophil-to-lymphocyte (NLR), platelet-to-lymphocyte (PLR), monocyte-to-lymphocyte (MLR), and fibrinogen-to-lymphocyte (FLR) ratios were significantly higher while the prognostic nutrition index values (PNI) were lower in patients with PC than in HC and OPT (all P < 0.05). Combining the FAR, FPR, and FLR with CA199 exhibited the best diagnostic value for distinguishing patients with early-stage PC from HC with an area under the curve (AUC) of 0.964, and for distinguishing patients with early-stage PC from OPT with an AUC of 0.924 in the training sets. In the testing set, compared with HC, the combination markers had powerful efficiency for PC with an AUC 0.947 and AUC 0.942 when comparing PC with OPT. The AUC was 0.915 for the combination of CA199, FAR, FPR, and FLR for differentiating between patients with pancreatic head cancer (PHC) and other pancreatic head tumors (OPHT), and 0.894 for differentiating between patients with pancreatic body and tail cancer (PBTC) and other pancreatic body and tail tumors (OPBTT). CONCLUSION: A combination of FAR, FPR, FLR, and CA199 may serve as a potential non-invasive biomarker for differentiating early-stage PC from HC and OPT, especially early-stage PHC.

Design and development of novel fluorescence sensing material for exosome recognition.

Exosomes are extracellular vesicles with relatively specific expression of CD63 transmembrane protein. In this study, We designed and constructed a multisite-targeting polymer which has both fluorescence and targeting recognition. It can bond to the hydrophilic group of CD63 by connecting with hydrogen. The chemical structure and the ability to combine with CD63 of fluorescent monomer and polymer were characterized and confirmed by FTIR and 1H NMR. MTT assay was performed to detect the cytotoxicity and biocompatibility of this polymer. Then we found the cell viability was 80.64% and the hemolysis rate of erythrocyte was only 0.101% even at F concentration of 20 µM. In vitro, the proposed polymer showed better ability to enter cells after linking exosomes via CD63; in vivo, it showed the ability to bind stably to exosomes and target tumor implants.

Deep Learning for Differentiating Benign From Malignant Parotid Lesions on MR Images.

PURPOSE/OBJECTIVESS: Salivary gland tumors are a rare, histologically heterogeneous group of tumors. The distinction between malignant and benign tumors of the parotid gland is clinically important. This study aims to develop and evaluate a deep-learning network for diagnosing parotid gland tumors via the deep learning of MR images. MATERIALS/METHODS: Two hundred thirty-three patients with parotid gland tumors were enrolled in this study. Histology results were available for all tumors. All patients underwent MRI scans, including T1-weighted, CE-T1-weighted and T2-weighted imaging series. The parotid glands and tumors were segmented on all three MR image series by a radiologist with 10 years of clinical experience. A total of 3791 parotid gland region images were cropped from the MR images. A label (pleomorphic adenoma and Warthin tumor, malignant tumor or free of tumor), which was based on histology results, was assigned to each image. To train the deep-learning model, these data were randomly divided into a training dataset (90%, comprising 3035 MR images from 212 patients: 714 pleomorphic adenoma images, 558 Warthin tumor images, 861 malignant tumor images, and 902 images free of tumor) and a validation dataset (10%, comprising 275 images from 21 patients: 57 pleomorphic adenoma images, 36 Warthin tumor images, 93 malignant tumor images, and 89 images free of tumor). A modified ResNet model was developed to classify these images. The input images were resized to 224x224 pixels, including four channels (T1-weighted tumor images only, T2-weighted tumor images only, CE-T1-weighted tumor images only and parotid gland images). Random image flipping and contrast adjustment were used for data enhancement. The model was trained for 1200 epochs with a learning rate of 1e-6, and the Adam optimizer was implemented. It took approximately 2 hours to complete the whole training procedure. The whole program was developed with PyTorch (version 1.2). RESULTS: The model accuracy with the training dataset was 92.94% (95% CI [0.91, 0.93]). The micro-AUC was 0.98. The experimental results showed that the accuracy of the final algorithm in the diagnosis and staging of parotid cancer was 82.18% (95% CI [0.77, 0.86]). The micro-AUC was 0.93. CONCLUSION: The proposed model may be used to assist clinicians in the diagnosis of parotid tumors. However, future larger-scale multicenter studies are required for full validation.

[Budd-Chiari syndrome in children and adolescents: therapeutic radiological intervention].

OBJECTIVE: Due to its minimal-invasive approach, endovascular procedure had replaced surgery in treating Budd-Chiari syndrome (BCS). The interventional therapy was a safe and effective treatment in adults with BCS and the cure rate was high. However Budd-Chiari syndrome in children and adolescents is rare. Published literature on interventional procedure for Budd-Chiari syndrome in children and adolescents is scarce. The aim of the study was to present results of percutaneous transluminal angioplasty (PTA) and stents placement in children and adolescents with BCS and to evaluate the efficacy and safety in these patients of this approach. METHOD: Twenty-five patients [16 boys and 9 girls; average age of (14.5 ± 3.4) years old; age ranged from 5 to 17 years] with Budd-Chiari syndrome who were hospitalized from December 1990 to August 2012 were presented. All of them were diagnosed by color Doppler ultrasound scan while 12 of them had magnetic resonance venography (MRV) scan. All of the patients had undergone angiographic examination. Four cases with membranous obstruction of the inferior vena cava (IVC) were treated with PTA. One case with segmental block of IVC was treated with PTA and stent placement. Five cases with membranous obstruction of IVC and hepatic vein (/and accessory hepatic vein) were treated with PTA. Among 8 cases with membranous obstruction of hepatic veins, 6 cases were treated with PTA and the others with PTA and stent placement. Among 4 cases with blocks of 3 hepatic veins (HVs), one was treated with PTA, one with PTA plus catheter thrombolysis plus PTA, one with PTA and stent placement and the other one was unsuccessful. Three cases with obstruction of HV and accessory HV (AHV) were treated with PTA. Totally, 24 patients were treated with interventional approach and followed up. RESULT: The procedure was successful in 24 patients. The involved veins (hepatic veins or IVC) were patented after interventional procedure. The pressure of hepatic vein was (42.1 ± 4.2) cm H2O (37-50 cm H2O) (1 cm H2O = 0.098 kPa) before the interventional therapy, while it was (17.3 ± 3.3) cm H2O (14-26 cm H2O) after it. The pressure of IVC was (30.6 ± 2.9) cm H2O (26-36 cm H2O) before the interventional therapy, while it was (18.8 ± 4.2) cm H2O (15-26 cm H2O) after it. The symptoms and signs vanished instantly after interventional procedure. There were no procedure-related complications. The rate of overall initial cure was 96%. The patients were followed up for a mean of 25.8 months (range 6 months to 8 years). Seven cases developed restenosis after first procedure. Five of them were treated with PTA, one with PTA plus catheter thrombolysis plus PTA, one with PTA and stent placement. All of the involved veins were patented again. Clinical symptoms were relieved. There were no procedure-related complications as well. CONCLUSION: The interventional procedure in children and adolescents with BCS is the same as in adults. Radiological therapeutic intervention is efficacious and safe in children and adolescents with BCS.