Development and validation of deep learning models for bowel obstruction on plain abdominal radiograph.

OBJECTIVE: Artificial intelligence (AI) could help medical practitioners in analyzing radiological images to determine the presence and site of bowel obstruction. This retrospective diagnostic study proposed a series of deep learning (DL) models for diagnosing bowel obstruction on abdominal radiograph. METHODS: A total of 2082 upright plain abdominal radiographs were retrospectively collected from four hospitals. The images were labeled as normal, small bowel obstruction and large bowel obstruction by three senior radiologists based on comprehensive examinations and interventions within 48 hours after admission. Gradient-weighted class activation mapping was used to visualize the inferential explanation. RESULTS: In the validation set, the Xception-backboned model achieved the highest accuracy (0.863), surpassing the VGG16 (0.847) and ResNet models (0.836). In the test set, the Xception model (accuracy: 0.807) outperformed other models and a junior radiologist (0.780) but not a senior radiologist (0.840). In the AI-aided diagnostic framework, the junior and senior radiologists made their judgements while aware of the Xception model predictions. Their accuracy significantly improved to 0.887 and 0.913, respectively. CONCLUSIONS: We developed and validated DL-based computer vision models for diagnosing bowel obstruction on plain abdominal radiograph. DL-based computer-aided diagnostic systems could reduce medical practitioners' workloads and improve diagnostic accuracy.

A few-shot learning framework for the diagnosis of osteopenia and osteoporosis using knee X-ray images.

OBJECTIVE: We developed a few-shot learning (FSL) framework for the diagnosis of osteopenia and osteoporosis in knee X-ray images. METHODS: Computer vision models containing deep convolutional neural networks were fine-tuned to enable generalization from natural images (ImageNet) to chest X-ray images (normal vs. pneumonia, base images). Then, a series of automated machine learning classifiers based on the Euclidean distances of base images were developed to make predictions for novel images (normal vs. osteopenia vs. osteoporosis). The performance of the FSL framework was compared with that of junior and senior radiologists. In addition, the gradient-weighted class activation mapping algorithm was used for visual interpretation. RESULTS: In Cohort #1, the mean accuracy (0.728) and sensitivity (0.774) of the FSL models were higher than those of the radiologists (0.512 and 0.448). A diagnostic pipeline of FSL model (first)-radiologists (second) achieved better performance (0.653 accuracy, 0.582 sensitivity, and 0.816 specificity) than radiologists alone. In Cohort #2, the diagnostic pipeline also showed improved performance. CONCLUSIONS: The FSL framework yielded practical performance with respect to the diagnosis of osteopenia and osteoporosis in comparison with radiologists. This retrospective study supports the use of promising FSL methods in computer-aided diagnosis tasks involving limited samples.

Effects of black bean cell wall pectin by exogenous calcium ions: Insight into the metabolomics, physicochemical properties and anti-digestive capacity.

In this work, the metabolomics, physicochemical and in vitro digestion properties of black beans influenced by different calcium ion solutions (0, 0.5 %, 1 %, and 2 %) were explored. The addition of calcium ions had a significant effect on the metabolic processing of black beans, including 16 differential metabolites and 4 metabolic pathways related to the cell wall. From the results of FT-IR and ICP-OES, it was confirmed that calcium ions can interact with COO- in non-methylated galacturonic acid in pectin to form calcium carboxylate strengthening the middle lamellae of the cell wall. Based on this mechanism, the soaked beans with an intact and dense cell structure were verified by the analyses of SEM and CLSM. Compared with other soaked beans, BB-2 exhibited lower cell permeability with electrical conductivity value decreased to 0.60 µs·cm-1. Additionally, BB-2 demonstrated slower digestion properties with digestion rate coefficient at 0.0020 min-1 and digestion extent only at 30.83 %, which is attributed to its increasingly compact cell wall and densely cellular matrix. This study illustrates the effect of calcium ions on the cellular structure of black beans, providing an effective process method for low glycemic index diets.

Effect of Fucoidan on Structure and Bioactivity of Chinese Steamed Bread.

Fucoidan refers to a group of sulphated polysaccharides obtained from brown seaweed, with numerous biological activities. In this study, fucoidan was fortified into Chinese steamed bread (CSB) at different concentrations (0, 1%, 3% and 5%) and the effect of fucoidan on the dough properties, structure properties and bioactivity were investigated. The results showed that fucoidan could change the viscosity of unfermented dough, and a high concentration of fucoidan could remove the free radicals produced by the SH-SS exchange reaction (GS-) in the dough, which significantly reduced the content of disulfide bond and reduced the expanded volume of fermented dough (p < 0.05). In addition, fucoidan forms a physical barrier on the surface of starch particles and hinders the reaction between protein-to-protein; therefore, fucoidan increased the hardness, gumminess and chewiness in CSB, and reduced the specific volume in CSB. Furthermore, the fucoidan-fortified CSB samples were found to have both the ability to significantly reduce the predicted glycemic index (pGI) (p < 0.05) and improve antioxidant activity (p < 0.05). Collectively, these findings could provide a theoretical basis for the applications of fucoidan as a functional component in fermented foods.

The mediation mechanism of calcium ions on black bean type 3 resistant starch: Metabolomics, structure characteristics and digestibility.

The type 3 resistant starch (RS3) is beneficial for blood glucose management. A high quality RS3 was provided and its formation mechanism after calcium ion (Ca2+) treatment was investigated in this study. The metabolomics, structure and digestion properties were evaluated. Metabolomics was performed by untargeted UHPLC-Q-TOF/MS, and a total of 11 significantly different metabolites was found. The NMR, ATR-FTIR, and XRD results showed that the degree of double helix decreased from 5.34 to 1.07, crystallinity decreased from 33.58 % to 19.88 %, and the amorphous region increased from 69.76 % to 78.33 %. Large particle polymers were observed by SEM on the granule surface of starch with Ca2+ treatment. Digestion test showed that Ca2+ increased the RS3 from 9.70 % to 22.26 %. The result indicated that Ca2+ induced the formation of chelates between Ca2+ and -OH, promoted the RS3 content and regulated carbohydrate metabolism. The study provided theoretical basis for producing low-glycemic black bean foods.

Development and validation of a multimodal model in predicting severe acute pancreatitis based on radiomics and deep learning.

OBJECTIVE: Aim to establish a multimodal model for predicting severe acute pancreatitis (SAP) using machine learning (ML) and deep learning (DL). METHODS: In this multicentre retrospective study, patients diagnosed with acute pancreatitis at admission were enrolled from January 2017 to December 2021. Clinical information within 24 h and CT scans within 72 h of admission were collected. First, we trained Model α based on clinical features selected by least absolute shrinkage and selection operator analysis. Second, radiomics features were extracted from 3D-CT scans and Model ß was developed on the features after dimensionality reduction using principal component analysis. Third, Model γ was trained on 2D-CT images. Lastly, a multimodal model, namely PrismSAP, was constructed based on aforementioned features in the training set. The predictive accuracy of PrismSAP was verified in the validation and internal test sets and further validated in the external test set. Model performance was evaluated using area under the curve (AUC), accuracy, sensitivity, specificity, recall, precision and F1-score. RESULTS: A total of 1,221 eligible patients were randomly split into a training set (n = 864), a validation set (n = 209) and an internal test set (n = 148). Data of 266 patients were for external testing. In the external test set, PrismSAP performed best with the highest AUC of 0.916 (0.873-0.960) among all models [Model α: 0.709 (0.618-0.800); Model ß: 0.749 (0.675-0.824); Model γ: 0.687 (0.592-0.782); MCTSI: 0.778 (0.698-0.857); RANSON: 0.642 (0.559-0.725); BISAP: 0.751 (0.668-0.833); SABP: 0.710 (0.621-0.798)]. CONCLUSION: The proposed multimodal model outperformed any single-modality models and traditional scoring systems.

Development and Validation of Multimodal Models to Predict the 30-Day Mortality of ICU Patients Based on Clinical Parameters and Chest X-Rays.

We aimed to develop and validate multimodal ICU patient prognosis models that combine clinical parameters data and chest X-ray (CXR) images. A total of 3798 subjects with clinical parameters and CXR images were extracted from the Medical Information Mart for Intensive Care IV (MIMIC-IV) database and an external hospital (the test set). The primary outcome was 30-day mortality after ICU admission. Automated machine learning (AutoML) and convolutional neural networks (CNNs) were used to construct single-modal models based on clinical parameters and CXR separately. An early fusion approach was used to integrate both modalities (clinical parameters and CXR) into a multimodal model named PrismICU. Compared to the single-modal models, i.e., the clinical parameter model (AUC = 0.80, F1-score = 0.43) and the CXR model (AUC = 0.76, F1-score = 0.45) and the scoring system APACHE II (AUC = 0.83, F1-score = 0.77), PrismICU (AUC = 0.95, F1 score = 0.95) showed improved performance in predicting the 30-day mortality in the validation set. In the test set, PrismICU (AUC = 0.82, F1-score = 0.61) was also better than the clinical parameters model (AUC = 0.72, F1-score = 0.50), CXR model (AUC = 0.71, F1-score = 0.36), and APACHE II (AUC = 0.62, F1-score = 0.50). PrismICU, which integrated clinical parameters data and CXR images, performed better than single-modal models and the existing scoring system. It supports the potential of multimodal models based on structured data and imaging in clinical management.

Preparation and Physico-Chemical Characterization of OSA-Modified Starches from Different Botanical Origins and Study on the Properties of Pickering Emulsions Stabilized by These Starches.

Native starch (NS) from different botanical origins (native rice/tapioca/oat starch, NRS/NTS/NOS) were hydrophobically modified by octenyl succinic anhydride (OSA), and the octenyl succinic (OS) groups were successfully introduced in the starch molecules which obtained OS-starch (OSRS, OSTS and OSOS) with different levels of modification (0.5%, 1.0%, 1.5%, 2.0%, 2.5%, 3.0%) and degree of substitution (DS). The structural properties of the OS-starch, such as granule size, crystal, wettability and morphology were studied, and the OS-starch was used as particulate stabilizers to produce oil-in-water (O/W) Pickering emulsions. The emulsion index, droplet size distribution and microstructures of Pickering emulsions produced by different OS-starches were compared. OSA modification had almost no effect on the morphology or crystal structure types of three kinds of NS and OS-starch but markedly increased the contact angle and particle size distribution of OSRS, OSTS and OSOS. Esterification reaction of OSA and starch mainly occurred in amorphous regions of starch, and the OSA significantly improved the emulsifying capacity of OSRS, OSTS and OSOS granules and thus stabilized emulsions formed at higher levels (2.5% and 3.0%) of modification of OS-Starch exhibited better stability; the ability of OS-starch to stabilize Pickering emulsion was 3.0% OSRS > 3.0% OSOS > 3.0% OSTS, respectively. Observation and structural properties analysis of OS-starch granules and Pickering emulsion droplets showed that the number and thickness of the starch granules on the oil-water interface of the emulsion droplets increased with improvement of the OSA modification level, and an aggregation state was formed between the OS-starch granules, which was also enhanced with the OSA modification levels. These were all necessary for the Pickering emulsion stabilized by starch granules to remain in a steady state.

Clinical analysis of subxiphoid vs. lateral approaches for treating early anterior mediastinal thymoma.

Objective: To investigate the clinical efficacy of the subxiphoid approach for early anterior mediastinal thymoma and evaluate its advantages over the lateral intercostal approach. Methods: A total of 345 patients with early anterior mediastinal thymoma were retrospectively analyzed from January 2016 to December 2020 in the First Affiliated Hospital of Soochow University. Out of these, 99 patients underwent subxiphoid video-assisted thoracoscopic thymectomy and 246 patients underwent transthoracic video-assisted thoracoscopic thymectomy. We compared the intraoperative conditions (such as operation time and intraoperative blood loss), postoperative conditions [such as postoperative pleural drainage volume, extubation time, postoperative hospital stay, and postoperative visual analogue scale (VAS) pain score], and postoperative complications (such as death, pneumonia, delayed wound healing, cardiac arrhythmia, and phrenic nerve injury) of the two groups and analyzed the clinical advantages of the subxiphoid approach for treating early anterior mediastinal thymoma. Results: There was no significant difference between the two groups in terms of general clinical features, operation time, and postoperative complications (P > 0.05).However, there was a significant difference in terms of intraoperative blood loss, postoperative pleural drainage volume, tube extubation time, postoperative hospital stay, postoperative VAS pain score, and postoperative analgesics (a significantly decreased flurbiprofen axetil amount) (P < 0.05). Conclusion: Compared with the lateral intercostal thoracic approach, the subxiphoid approach had advantages in terms of intraoperative blood loss, postoperative hospital stay, tube extubation time, postoperative pleural drainage volume, postoperative VAS pain score, and analgesics dosage. It could provide a better view of the bilateral pleural cavities and more thorough thymectomy and superior cosmesis, and it proved to be a safe and feasible minimally invasive surgical method.