Different radiomics annotation methods comparison in rectal cancer characterisation and prognosis prediction: a two-centre study.

OBJECTIVES: To explore the performance differences of multiple annotations in radiomics analysis and provide a reference for tumour annotation in large-scale medical image analysis. METHODS: A total of 342 patients from two centres who underwent radical resection for rectal cancer were retrospectively studied and divided into training, internal validation, and external validation cohorts. Three predictive tasks of tumour T-stage (pT), lymph node metastasis (pLNM), and disease-free survival (pDFS) were performed. Twelve radiomics models were constructed using Lasso-Logistic or Lasso-Cox to evaluate and four annotation methods, 2D detailed annotation along tumour boundaries (2D), 3D detailed annotation along tumour boundaries (3D), 2D bounding box (2DBB), and 3D bounding box (3DBB) on T2-weighted images, were compared. Radiomics models were used to establish combined models incorporating clinical risk factors. The DeLong test was performed to compare the performance of models using the receiver operating characteristic curves. RESULTS: For radiomics models, the area under the curve values ranged from 0.627 (0.518-0.728) to 0.811 (0.705-0.917) in the internal validation cohort and from 0.619 (0.469-0.754) to 0.824 (0.689-0.918) in the external validation cohort. Most radiomics models based on four annotations did not differ significantly, except between the 3D and 3DBB models for pLNM (p = 0.0188) in the internal validation cohort. For combined models, only the 2D model significantly differed from the 2DBB (p = 0.0372) and 3D models (p = 0.0380) for pDFS. CONCLUSION: Radiomics and combined models constructed with 2D and bounding box annotations showed comparable performances to those with 3D and detailed annotations along tumour boundaries in rectal cancer characterisation and prognosis prediction. CRITICAL RELEVANCE STATEMENT: For quantitative analysis of radiological images, the selection of 2D maximum tumour area or bounding box annotation is as representative and easy to operate as 3D whole tumour or detailed annotations along tumour boundaries. KEY POINTS: There is currently a lack of discussion on whether different annotation efforts in radiomics are predictively representative. No significant differences were observed in radiomics and combined models regardless of the annotations (2D, 3D, detailed, or bounding box). Prioritise selecting the more time and effort-saving 2D maximum area bounding box annotation.

Deep Learning-Based Multiparametric MRI Model for Preoperative T-Stage in Rectal Cancer.

BACKGROUND: Conventional MRI staging can be challenging in the preoperative assessment of rectal cancer. Deep learning methods based on MRI have shown promise in cancer diagnosis and prognostication. However, the value of deep learning in rectal cancer T-staging is unclear. PURPOSE: To develop a deep learning model based on preoperative multiparametric MRI for evaluation of rectal cancer and to investigate its potential to improve T-staging accuracy. STUDY TYPE: Retrospective. POPULATION: After cross-validation, 260 patients (123 with T-stage T1-2 and 134 with T-stage T3-4) with histopathologically confirmed rectal cancer were randomly divided to the training (N = 208) and test sets (N = 52). FIELD STRENGTH/SEQUENCE: 3.0 T/Dynamic contrast enhanced (DCE), T2-weighted imaging (T2W), and diffusion-weighted imaging (DWI). ASSESSMENT: The deep learning (DL) model of multiparametric (DCE, T2W, and DWI) convolutional neural network were constructed for evaluating preoperative diagnosis. The pathological findings served as the reference standard for T-stage. For comparison, the single parameter DL-model, a logistic regression model composed of clinical features and subjective assessment of radiologists were used. STATISTICAL TESTS: The receiver operating characteristic curve (ROC) was used to evaluate the models, the Fleiss' kappa for the intercorrelation coefficients, and DeLong test for compare the diagnostic performance of ROCs. P-values less than 0.05 were considered statistically significant. RESULTS: The Area Under Curve (AUC) of the multiparametric DL-model was 0.854, which was significantly higher than the radiologist's assessment (AUC = 0.678), clinical model (AUC = 0.747), and the single parameter DL-models including T2W-model (AUC = 0.735), DWI-model (AUC = 0.759), and DCE-model (AUC = 0.789). DATA CONCLUSION: In the evaluation of rectal cancer patients, the proposed multiparametric DL-model outperformed the radiologist's assessment, the clinical model as well as the single parameter models. The multiparametric DL-model has the potential to assist clinicians by providing more reliable and precise preoperative T staging diagnosis. EVIDENCE LEVEL: 3 TECHNICAL EFFICACY: Stage 2.

Uncertainty-guided transformer for brain tumor segmentation.

Multi-model data can enhance brain tumor segmentation for the rich information it provides. However, it also introduces some redundant information that interferes with the segmentation estimation, as some modalities may catch features irrelevant to the tissue of interest. Besides, the ambiguous boundaries and irregulate shapes of different grade tumors lead to a non-confidence estimate of segmentation quality. Given these concerns, we exploit an uncertainty-guided U-shaped transformer with multiple heads to construct drop-out format masks for robust training. Specifically, our drop-out masks are composed of boundary mask, prior probability mask, and conditional probability mask, which can help our approach focus more on uncertainty regions. Extensive experimental results show that our method achieves comparable or higher results than previous state-of-the-art brain tumor segmentation methods, achieving average dice coefficients of [Formula: see text] and Hausdorff distance of 4.91 on the BraTS2021 dataset. Our code is freely available at https://github.com/chaineypung/BTS-UGT.

Observation of Curative Effect of Lung Recruitment in Patients with Acute Respiratory Distress Syndrome after Cardiopulmonary Bypass Surgery.

Recruitment maneuver (RM) has become a routine supplementary maneuver for clinical rescue of severe ARDS with low tidal volume/pressure-limited mechanical ventilation. Recruitment of patients with ARDS mechanical ventilation can improve the lung compliance, promote the opening of collapsed alveoli, improve the ratio of ventilation to blood flow, reduce dead space, reduce shunt flow, and improve oxygenation function. In this paper, the patients were divided into lung recruitment group and conventional treatment group by the random number permutation table method. When the patient's percutaneous oxygen saturation is less than or equal to 88%, the partial pressure of oxygen in the arterial blood gas is less than or equal to 55 mmHg, or the ventilator tube is disconnected during sputum suction or other accidents, a CPAP × 60 - second lung recruitment maneuver is required. Then adjust the ventilator parameters in the same way. In the process of lung recruitment, the changes in invasive continuous arterial blood pressure will also be observed. If the blood pressure dropped to ≤90/60 mmHg, one recruitment maneuver was terminated in advance. And both groups of patients used the Dräger- or PB840-imported multifunctional ventilator. The treatment of primary disease and predisposing factors, fluid management strategies, antibiotics and glucocorticoids, nutrition, and metabolic support in the two groups of patients in the study were the same. The PaO2/FiO2 value improved by 51% 10 minutes after recruitment, and the median increased from 111 (IQR, 73-265) before recruitment to 170 (IQR, 102-340) (P < 0.01), the improvement of PaO2/FiO2 at 4 hours after recruitment and 12 hours after recruitment was 78% (P < 0.05) and 39% (P < 0.01), respectively, and the median PaO2/FiO2 at 4 hours after recruitment was 198 (IQR, 116-256). The median PaO2/FiO2 became 155 (IQR, 127-235) 12 hours after recruitment. Recruitment can reduce the accumulation of neutrophils in lung tissue, reduce the release of inflammatory factors, reduce pulmonary edema, and reduce pathological damage.

Integrative Analysis of 5-Hydroxymethylcytosine and Transcriptional Profiling Identified 5hmC-Modified lncRNA Panel as Non-Invasive Biomarkers for Diagnosis and Prognosis of Pancreatic Cancer.

Pancreatic adenocarcinoma (PAAD) is the fourth leading cause of cancer-related deaths worldwide. 5-Hydroxymethylcytosine (5hmC)-mediated epigenetic regulation has been reported to be involved in cancer pathobiology and has emerged to be promising biomarkers for cancer diagnosis and prognosis. However, 5hmC alterations at long non-coding RNA (lncRNA) genes and their clinical significance remained unknown. In this study, we performed the genome-wide investigation of lncRNA-associated plasma cfDNA 5hmC changes in PAAD by plotting 5hmC reads against lncRNA genes, and identified six PAAD-specific lncRNAs with abnormal 5hmC modifications compared with healthy individuals. Then we applied machine-learning and Cox regression approaches to develop predictive diagnostic (5hLRS) and prognostic (5hLPS) models using the 5hmC-modified lncRNAs. The 5hLRS demonstrated excellent performance in discriminating PAAD from healthy controls with an area under the curve (AUC) of 0.833 in the training cohort and 0.719 in the independent testing cohort. The 5hLPS also effectively divides PAAD patients into high-risk and low-risk groups with significantly different clinical outcomes in the training cohort (log-rank test p = 0.04) and independent testing cohort (log-rank test p = 0.0035). Functional analysis based on competitive endogenous RNA (ceRNA) and enrichment analysis suggested that these differentially regulated 5hmC modified lncRNAs were associated with angiogenesis, circulatory system process, leukocyte differentiation and metal ion homeostasis that are known important events in the development and progression of PAAD. In conclusion, our study indicated the potential clinical utility of 5hmC profiles at lncRNA loci as valuable biomarkers for non-invasive diagnosis and prognostication of cancers.

Mean apparent diffusion coefficient in a single slice may predict tumor response to whole-brain radiation therapy in non-small-cell lung cancer patients with brain metastases.

OBJECTIVES: This study aimed to access the performance of apparent diffusion coefficient (ADC) as a predictor for treatment response to whole-brain radiotherapy (WBRT) in patients with brain metastases (BMs) from non-small-cell lung cancer (NSCLC). METHODS: A retrospective analysis was conducted of 102 NSCLC patients with BMs who underwent WBRT between 2012 and 2016. Diffusion-weighted MRI were performed pre-WBRT and within 12 weeks after WBRT started. Mean single-plane ADC value of ROIs was evaluated by two radiologists blinded to results of each other. The treatment response rate, intracranial progression-free survival (PFS), and overall survival (OS) were analyzed based on the ADC value and ΔADC respectively. At last, we used COX and logistic regression to do the multivariate analysis. RESULTS: There was good inter-observer agreement of mean ADC value pre-WBRT, post-WBRT, and ΔADC between the 2 radiologists (Pearson correlation 0.915 [pre-WBRT], 0.950 [post-WBRT], 0.937 [ΔADC], p < 0.001, for each one). High mean ADC value were related with better response rate (72.2% vs 37.5%, p = 0.001) and iPFS (7.6 vs 6.4 months, p = 0.031). High ΔADC were related with better response rate (73.6% vs 36.7%, p < 0.001). Multivariate analysis shows that histopathology, BMs number, high ADC value pre-WBRT, and high ΔADC post-WBRT were related to better treatment response of WBRT, and KPS, BMs number, and low ADC value pre-WBRT increased the risk of developing intracranial relapse. CONCLUSIONS: The mean single-plane ADC value pre-WBRT and ΔADC post-WBRT were potential predictor for intracranial tumor response to WBRT in NSCLC patients with brain metastases. KEY POINTS: • ADC value is a potential predictor of intracranial treatment response to WBRT in NSCLC patients with brain metastases. • Higher mean ADC value pre-WBRT and ΔADC post-WBRT of brain metastases were related to better intracranial tumor response. • Prediction of response before WBRT using ADC value can help oncologists to make better therapy plans and avoid missing opportunities for rescue therapy.

Molecular Mechanism Exploration of Pancreatitis Based on miRNA Expression Profile.

BACKGROUND: Pancreatitis is a popular disease around the world, and can also lead to pancreatic cancer. Pancreatitis can be distinguished into two types, acute pancreatitis (AP) and chronic pancreatitis (CP). Every year, AP leads to approximately 275,000 new cases and is the most frequent gastrointestinal disease in American. METHODS: The miRNA expression profile of pancreatic cancer and pancreatitis was downloaded from GEO with accession id GSE24279. First, the differentially expressed miRNAs with |fold change| ≥ 2 and p-value ≤ 0.05 and then the target genes of significantly differentially expressed miRNAs in pancreatitis were identified and the interaction network was constructed. Also the biological functions of the target genes were explored based on GO and KEGG enrichment. Finally, the expression values of hsa-miR-373-5p and hsa-miR-374a-5p were validated using RT-PCR. RESULTS: A total of 40 and 13 differentially expressed miRNAs were screened out for pancreatic and pancreatitis, respectively. Two miRNAs, hsa-miR-373-5p and hsa-miR-374a-5p, had significantly down-regulated expression in pancreatitis. Target gene analysis showed that hsa-miR-373-5p probably participates in the development of pancreatitis by regulating MBL2, MAT2B, and BCL10. In addition, has-miR-374a-5p can regulate the expression of NCK1, MMP14. Those genes are involved in nuclear factor kappa B and p38 signaling in the early stage of pancreatitis. Also, NCK1 can regulate pancreatic ß-cell proinsulin content and participate in the progression of pancreatic cancer development. CONCLUSIONS: In summary, the findings in this study deciphered the potential miRNA regulation mechanism in pancreatitis, and identified valuable biomarkers for the diagnosis of pancreatitis.