Deep causal learning for pancreatic cancer segmentation in CT sequences.

Segmenting the irregular pancreas and inconspicuous tumor simultaneously is an essential but challenging step in diagnosing pancreatic cancer. Current deep-learning (DL) methods usually segment the pancreas or tumor independently using mixed image features, which are disrupted by surrounding complex and low-contrast background tissues. Here, we proposed a deep causal learning framework named CausegNet for pancreas and tumor co-segmentation in 3D CT sequences. Specifically, a causality-aware module and a counterfactual loss are employed to enhance the DL network's comprehension of the anatomical causal relationship between the foreground elements (pancreas and tumor) and the background. By integrating causality into CausegNet, the network focuses solely on extracting intrinsic foreground causal features while effectively learning the potential causality between the pancreas and the tumor. Then based on the extracted causal features, CausegNet applies a counterfactual inference to significantly reduce the background interference and sequentially search for pancreas and tumor from the foreground. Consequently, our approach can handle deformable pancreas and obscure tumors, resulting in superior co-segmentation performance in both public and real clinical datasets, achieving the highest pancreas/tumor Dice coefficients of 86.67%/84.28%. The visualized features and anti-noise experiments further demonstrate the causal interpretability and stability of our method. Furthermore, our approach improves the accuracy and sensitivity of downstream pancreatic cancer risk assessment task by 12.50% and 50.00%, respectively, compared to experienced clinicians, indicating promising clinical applications.

Resolving fluorescence spectra of Maillard reaction products formed on bovine serum albumin using parallel factor analysis.

Formation of Maillard reaction products (MRPs) is increasingly studied by the use of fluorescence spectroscopy, and most often, by measuring single excitation/emission pairs or use of unresolved spectra. However, due to the matrix complexity and potential co-formation of fluorescent oxidation products on tryptophan and tyrosine residues, this practice will often introduce errors in both identification and quantification. The present study investigates the combination of fluorescence excitation emission matrix (EEM) spectroscopy and parallel factor analysis (PARAFAC) to resolve the EEMs into its underlying fluorescent signals, allowing for better identification and quantification of MRPs. EEMs were recorded on a sample system of bovine serum albumin incubated at 40 °C for up to one week with either glucose, methylglyoxal or glyoxal added. Ten unique PARAFAC components were resolved, and assignment was attempted based on similarity with fluorescence of pure standards of MRPs and oxidation products and reported data from literature. Of the ten fluorescent PARAFAC components, tyrosine and buried and exposed tryptophan were resolved and identified, as well as the formation of specific MRPs (argpyrimidine and Nα-acetyl-Nδ-(5-methyl-4-imidazolon-2-yl)ornithine) and tryptophan oxidation products (kynurenine and dioxindolylalanine). The formation of the PARAFAC resolved protein modifications were qualitatively validated by liquid chromatography-mass spectrometry.

Cysteine residues are responsible for the sulfurous off-flavor formed in heated whey protein solutions.

Odor-active volatile sulfur compounds are formed in heated food protein systems. In the present study, hydrogen sulfide (H2S) was found to be the most abundant sulfur volatile in whey protein solutions (whey protein isolate [WPI], a whey model system and single whey proteins) by gas chromatography-flame photometric detector (GC-FPD) analysis after heat treatments (60-90 °C for 10 min, 90 °C for 120 min and UHT-like treatment). H2S was detected in WPI after heating at 90 °C for 10 min, and was significantly increased at higher heat load (90 °C for 120 min and the UHT-like treatment). Site-specific LC-MS/MS-based proteomic analysis was conducted, monitoring desulfurization reactions in these protein systems to investigate the mechanism of H2S formation in heated WPI. Cysteine residues from beta-lactoglobulin were found to be responsible for the formation of H2S in heated WPI, presumably via beta-elimination.

Multi-Dimensional Cascaded Net with Uncertain Probability Reduction for Abdominal Multi-Organ Segmentation in CT Sequences.

BACKGROUND AND OBJECTIVE: Deep learning abdominal multi-organ segmentation provides preoperative guidance for abdominal surgery. However, due to the large volume of 3D CT sequences, the existing methods cannot balance complete semantic features and high-resolution detail information, which leads to uncertain, rough, and inaccurate segmentation, especially in small and irregular organs. In this paper, we propose a two-stage algorithm named multi-dimensional cascaded net (MDCNet) to solve the above problems and segment multi-organs in CT images, including the spleen, kidney, gallbladder, esophagus, liver, stomach, pancreas, and duodenum. METHODS: MDCNet combines the powerful semantic encoder ability of a 3D net and the rich high-resolution information of a 2.5D net. In stage1, a prior-guided shallow-layer-enhanced 3D location net extracts entire semantic features from a downsampled CT volume to perform rough segmentation. Additionally, we use circular inference and parameter Dice loss to alleviate uncertain boundary. The inputs of stage2 are high-resolution slices, which are obtained by the original image and coarse segmentation of stage1. Stage2 offsets the details lost during downsampling, resulting in smooth and accurate refined contours. The 2.5D net from the axial, coronal, and sagittal views also compensates for the missing spatial information of a single view. RESULTS: The experiments on the two datasets both obtained the best performance, particularly a higher Dice on small gallbladders and irregular duodenums, which reached 0.85±0.12 and 0.77±0.07 respectively, increasing by 0.02 and 0.03 compared to the state-of-the-art method. CONCLUSION: Our method can extract all semantic and high-resolution detail information from a large-volume CT image. It reduces the boundary uncertainty while yielding smoother segmentation edges, indicating good clinical application prospects.

Oxidation of Whey Proteins during Thermal Treatment Characterized by a Site-Specific LC-MS/MS-Based Proteomic Approach.

Thermal treatment is often employed in food processing to tailor product properties by manipulating the ingredient functionality, but these elevated temperatures may accelerate oxidation and nutrient loss. Here, oxidation of different whey protein systems [α-lactalbumin (α-LA), ß-lactoglobulin (ß-LG), a mix of α-LA and ß-LG (whey model), and a commercial whey protein isolate (WPI)] was investigated during heat treatment at 60-90 °C and a UHT-like treatment by LC-MS-based proteomic analysis. The relative modification levels of each oxidation site were calculated and compared among different heat treatments and sample systems. Oxidation increased significantly in protein systems after heating at ≥90 °C but decreased in systems with higher complexity [pure protein (α-LA > ß-LG) > whey model > WPI]. In α-LA, Cys, Met, and Trp residues were found to be most prone to oxidation. In ß-LG-containing protein systems, Cys residues were suggested to scavenge most of the reactive oxidants and undergo an oxidation-mediated disulfide rearrangement. The rearranged disulfide bonds contributed to protein aggregation, which was suggested to provide physical protection against oxidation. Overall, limited loss of amino acid residues was detected after acidic hydrolysis followed by UHPLC analysis, which showed only a minor effect of heat treatment on protein oxidation in these protein systems.

Site-Specific Characterization of Heat-Induced Disulfide Rearrangement in Beta-Lactoglobulin by Liquid Chromatography-Mass Spectrometry.

Disulfides are important for maintaining the protein native structure, but they may undergo rearrangement in the presence of free Cys residues, especially under elevated temperatures. Disulfide rearrangement may result in protein aggregation, which is associated with in vivo pathologies in organisms and in vitro protein functionality in food systems. In a food context, it is therefore important to understand the process of disulfide rearrangement on a site-specific level in order to control aggregation. In the present study, a liquid chromatography-mass spectrometry (LC-MS)-based bottom-up site-specific proteomic approach was optimized to study disulfide rearrangements in beta-lactoglobulin (ß-LG) under different heat treatments (60-90 °C). Artifactual disulfide rearrangement observed during sample preparation using a conventional protocol was detected and minimized by blocking the remaining free Cys residues with iodoacetamide in the presence of urea after heat treatment. Use of endoproteinase Glu-C for enzymatic hydrolysis allowed, for the first time, identification and comparison of the relative intensity of all theoretically possible ß-LG disulfide cross-links formed by the heat treatments. Non-native disulfides were formed from heat treatment at approx. 70 °C where ß-LG started to unfold, while higher levels of inter-molecular disulfide links were formed at ≥80 °C, in agreement with ß-LG aggregation detected by size exclusion chromatography analysis. Collectively, the Cys residues of the surface-located native disulfide Cys66-Cys160 were proposed to be more reactive, participating in heat-induced disulfide rearrangement, compared to other Cys residues. The abundant signal of non-native disulfide bonds containing Cys66, especially Cys66-Cys66, observed after heating suggested that Cys66 is a key disulfide-linked Cys residue in ß-LG participating in heat-induced inter-molecular disulfide bonds and the corresponding protein aggregation.

Antibiotic Resistance Characterization of Bacteria Isolated from Traditional Chinese Paocai.

In this work, the antibiotic resistance of 218 isolates to 9 different antibiotics was analyzed with minimum inhibitory concentration method. All Lactobacillus pentosus strains were found to be resistant to streptomycin sulfate and ciprofloxacin hydrochloride. Lactococcus lactis strains were resistant to streptomycin sulfate. Specifically, 90% Klebsiella oxytoca and all Citrobacter freundii strains were resistant to ampicillin sodium. 30% K. oxytoca strains were resistant to ciprofloxacin hydrochloride. All Bacillus albus strains were resistant to erythromycin and 80% strains were resistant to ampicillin sodium. Results from PCR analysis revealed that 90 isolates carried the aadE gene. The tetM gene was detected in four L. pentosus isolates. And the streptomycin resistant gene aadA was detected in one L. pentosus isolate. Metagenome analysis revealed that 74.7% genes associated with antibiotic resistance were antibiotic resistance genes. The tetM and aadA genes, detected in PCR analysis, were also retrieved from the paocai metagenome. In brief, this study generated the antibiotic resistance profile of some paocai-originated bacteria strains. L. pentosus found in the final edible paocai were inherently resistant to antibiotics, such as streptomycin and ciprofloxacin. Results in this work reminds us to carefully choose the LAB strains for traditional Chinese paocai production to avoid potential spreading of antibiotic resistant genes.

Computer-aided differentiates benign from malignant IPMN and MCN with a novel feature selection algorithm.

In clinical practice, differentiating benign from malignant intraductal papillary mucinous neoplasm (IPMN) and mucinous cystic neoplasm (MCN) preoperatively is crucial for deciding future treating algorithm. However, it remains challenging as benign and malignant lesions usually show similarities in both imaging appearances and clinical indices. Therefore, a robust and accurate computer-aided diagnosis (CAD) system based on radiomics and clinical indices was proposed in this paper to solve this dilemma. In the proposed CAD system, 107 patients were enrolled, where 90 cases were randomly selected for the training set with 5-fold cross validation to build the diagnostic model, while 17 cases were remained for an independent testing set to validate the performance. 436 high-throughput radiomics features while 9 clinical indices were designed and extracted. A novel feature selection algorithm named BLR (Bootstrapping repeated LASSO with Random selections) was proposed to select the most effective features. Then the selected features were sent to Support Vector Machine (SVM) to differentiate the benign or malignant. In the cross-validation cohort and independent testing cohort, the area under receiver operating characteristic curve (AUC) of CAD scheme were 0.83 and 0.92, respectively. The results fully prove the proposed CAD system achieves significant effect in tumors diagnosis.

Breast DCE-MRI radiomics: a robust computer-aided system based on reproducible BI-RADS features across the influence of datasets bias and segmentation methods.

PURPOSE: A highly accurate and robust computer-aided system based on quantitative high-throughput Breast Imaging Reporting and Data System (BI-RADS) features from dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) can drive the success of radiomic applications in breast cancer diagnosis. We aim to build a stable system with highly reproducible radiomics features, which can make diagnostic performance independent of datasets bias and segmentation methods. METHOD: We applied a dataset of 267 patients including 136 malignant and 131 benign tumors from two MRI manufacturers, where 211 cases from a Philips system and 55 cases from a GE system. First, manual annotations, 3D-Unet and 2D-Unet were applied as different segmentation methods. Second, we designed and extracted 3172 features from six modalities of DCE-MRI based on BI-RADS. Third, the feature selection was conducted. Between-class distance was utilized to eliminate the effect of dataset bias caused by two machines. Concordance correlation coefficient, intraclass correlation coefficient and deviation were employed to evaluate the influence of three segmentation methods. We further eliminated features redundancy using genetic algorithm. Finally, three classifiers including support vector machine (SVM), the bagged trees and K-Nearest Neighbor were evaluated by their performance for diagnosing malignant and benign tumors. RESULTS: A total of 246 features were preserved to have high stability and reproducibility. The final feature set showed the robust performance under these factors and achieved the area under curve of 0.88, the accuracy of 0.824, the sensitivity of 0.844, the specificity of 0.807 in differentiating benign and malignant tumors with the SVM classifier using manually segmentation results. CONCLUSION: The final selected 246 features are reproducible and show little dependence on segmentation methods and data perturbation. The high stability and effectiveness of diagnosis across these factors illustrate that the preserved features can be used for prognostic analysis and help radiologists in the diagnosis of breast cancer.

Effect of heat treatment on bacteriostatic activity and protein profile of bovine whey proteins.

Bovine milk shows bacteriostatic activity mainly due to the presence of antibacterial proteins, like lactoferrin, lactoperoxidase and immunoglobulins. Heat treatment is applied to kill bacteria and thereby extend shelf life of dairy products. Such heat treatment may, however, impair the activity of native antibacterial proteins in milk. The aim of this study was to investigate bacteriostatic capacity and retention of antibacterial proteins in unheated and heated bovine milk. Skim milk samples were heated at 65 °C, 70 °C, 75 °C, 80 °C and 85 °C, for 30 min. Whey was isolated from the heat-treated skim milk and the bacteriostatic capacity of this whey was tested against Streptococcus thermophilus, Escherichia coli, Lactococcus lactis and Pseudomonas fluorescens. The proteomic profile of native whey was determined using LC-MS/MS-based proteomics. Results showed that the bacteriostatic activity of whey negatively correlated with intensity of heat treatment, which was also reflected in the reduced level of native antibacterial proteins. There is a significant difference between milk samples treated for 30 min at <75 °C and milk samples treated at ≥75 °C in both bacteriostatic capacity and native antibacterial proteins. Growth rates of Streptococcus thermophilus, Lactococcus lactis and Pseudomonas fluorescens were negatively correlated with retention of lactoferrin and lactoperoxidase. In conclusion, our study shows that the bacteriostatic capacity of whey decreases with increasing heating intensity, which is strongly correlated with the denaturation of antibacterial proteins. Bacteriostatic activity can be a biomarker for loss of function of antibacterial proteins, and can thereby be used as an indicator for the extent of heat processing of dairy products including antibacterial proteins in a mild heat treatment.