Search

Prognostic assessment remains a critical challenge in medical research, often limited by the lack of well-labeled data. In this work, we introduce ContraSurv, a weakly-supervised learning framework based on contrastive learning, designed to enhance prognostic predictions in 3D medical images. ContraSurv utilizes both the self-supervised information inherent in unlabeled data and the weakly-supervised cues present in censored data, refining its capacity to extract prognostic representations. For this purpose, we establish a Vision Transformer architecture optimized for our medical image datasets and introduce novel methodologies for both self-supervised and supervised contrastive learning for prognostic assessment. Additionally, we propose a specialized supervised contrastive loss function and introduce SurvMix, a novel data augmentation technique for survival analysis. Evaluations were conducted across three cancer types and two imaging modalities on three real-world datasets. The results confirmed the enhanced performance of ContraSurv over competing methods, particularly in data with a high censoring rate.

Prostate cancer (PCa) is one of the leading causes of cancer death among men worldwide. Circular RNAs (circRNAs) have been implicated in the pathogenesis of PCa. However, the precise action of circ_0001326 in PCa malignant progression is still unknown. The levels of circ_0001326, miR-577 and voltage dependent anion channel 1 (VDAC1) were determined by quantitative real-time polymerase chain reaction (qRT-PCR) and western blot. Cell proliferation, colony formation, apoptosis, migration and invasion were evaluated by the Cell Counting Kit-8 (CCK-8), EdU staining, colony formation, flow cytometry, wound-healing and transwell assays, respectively. Targeted relationships among circ_0001326, miR-577 and VDAC1 were confirmed by dual-luciferase reporter assays. Xenograft experiments were performed to detect the role of circ_0001326 in tumor growth. Our data revealed that circ_0001326 was overexpressed in PCa tissues and cells. Circ_0001326 depletion repressed PCa cell proliferation, migration, and invasion and enhanced apoptosis in vitro, as well as hampered tumor growth in vivo. Mechanistically, circ_0001326 directly targeted miR-577, and VDAC1 was directly targeted and suppressed by miR-577. Moreover, the effects of circ_0001326 knockdown on PCa cell functional behaviors were mediated by miR-577. VDAC1 silencing phenocopied miR-577 overexpression in regulating PCa cell functional behaviors in vitro. Furthermore, circ_0001326 regulated VDAC1 expression through sponging miR-577. Our findings showed that circ_0001326 regulated PCa cell functional behaviors at least partly through targeting the miR-577/VDAC1 axis., (© 2024 Wiley Periodicals LLC.)

Background: Clinical observations indicate a correlation between the gut microbiota and overactive bladder (OAB) symptoms. Nevertheless, the causal relationship and mechanisms between gut microbiota and OAB symptoms remain elusive., Methods: Two-sample Mendelian randomization (MR) analyses were performed to assess the association between gut microbiota and OAB symptoms, including urinary incontinence (UI). Data were obtained from the MiBioGen International Consortium genome-wide association studies (GWAS) dataset and the IEU GWAS database. The inverse variance weighted method was used as the primary approach in the MR analysis, with the weighted median, MR-Egger, and weighted mode methods as supplementary approaches. Sensitivity analyses were employed to assess potential violations of the MR assumptions., Results: Our analysis identified seven gut bacterial taxa with a causal relationship to OAB and nine gut bacterial taxa associated with UI. Genera Eubacteriumfissicatenumgroup , LachnospiraceaeNK4A136group , and Romboutsia were identified as protective factors against OAB, while genera Barnesiella , FamilyXIIIAD3011group , Odoribacter, and RuminococcaceaeUCG005 were associated with an increased risk of OAB. A higher abundance of the genus Coprococcus3 , order Burkholderiales, and phylum Verrucomicrobia predicted a lower risk of UI. Conversely, the class Mollicutes, genus Ruminococcus gauvreauii group, order Mollicutes RF9, and phylum Firmicutes and Tenericutes were positively correlated with UI risk. The sensitivity analysis excluded the influence of potential heterogeneity and horizontal pleiotropy., Conclusion: This study revealed a causal relationship between gut microbiota and OAB symptoms, providing new insights and a theoretical foundation to identify biomarkers and therapeutic targets for patients with OAB symptoms., Competing Interests: The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest., (Copyright © 2024 Shen, Fang, Zhang, Zhu, Chen and Tang.)

Objective. In the realm of utilizing artificial intelligence (AI) for medical image analysis, the paradigm of 'signal-image-knowledge' has remained unchanged. However, the process of 'signal to image' inevitably introduces information distortion, ultimately leading to irrecoverable biases in the 'image to knowledge' process. Our goal is to skip reconstruction and build a diagnostic model directly from the raw data (signal). Approach . This study focuses on computed tomography (CT) and its raw data (sinogram) as the research subjects. We simulate the real-world process of 'human-signal-image' using the workflow 'CT-simulated data- reconstructed CT,' and we develop a novel AI predictive model directly targeting raw data (RCTM). This model comprises orientation, spatial, and global analysis modules, embodying the fusion of local to global information extraction from raw data. We selected 1994 patients with retrospective cases of solid lung nodules and modeled different types of data. Main results . We employed predefined radiomic features to assess the diagnostic feature differences caused by reconstruction. The results indicated that approximately 14% of the features had Spearman correlation coefficients below 0.8. These findings suggest that despite the increasing maturity of CT reconstruction algorithms, they still introduce perturbations to diagnostic features. Moreover, our proposed RCTM achieved an area under the curve (AUC) of 0.863 in the diagnosis task, showcasing a comprehensive superiority over models constructed from secondary reconstructed CTs (0.840, 0.822, and 0.825). Additionally, the performance of RCTM closely resembled that of models constructed from original CT scans (0.868, 0.878, and 0.866). Significance . The diagnostic and therapeutic approach directly based on CT raw data can enhance the precision of AI models and the concept of 'signal-to-image' can be extended to other types of imaging. AI diagnostic models tailored to raw data offer the potential to disrupt the traditional paradigm of 'signal-image-knowledge', opening up new avenues for more accurate medical diagnostics., (Creative Commons Attribution license.)

Importance: Digestive system neoplasms (DSNs) are the leading cause of cancer-related mortality with a 5-year survival rate of less than 20%. Subjective evaluation of medical images including endoscopic images, whole slide images, computed tomography images, and magnetic resonance images plays a vital role in the clinical practice of DSNs, but with limited performance and increased workload of radiologists or pathologists. The application of artificial intelligence (AI) in medical image analysis holds promise to augment the visual interpretation of medical images, which could not only automate the complicated evaluation process but also convert medical images into quantitative imaging features that associated with tumor heterogeneity., Highlights: We briefly introduce the methodology of AI for medical image analysis and then review its clinical applications including clinical auxiliary diagnosis, assessment of treatment response, and prognosis prediction on 4 typical DSNs including esophageal cancer, gastric cancer, colorectal cancer, and hepatocellular carcinoma., Conclusion: AI technology has great potential in supporting the clinical diagnosis and treatment decision-making of DSNs. Several technical issues should be overcome before its application into clinical practice of DSNs., Competing Interests: Competing interests: The authors declare that they have no competing interests., (Copyright © 2023 Shuaitong Zhang et al.)

OBJECTIVE. The purpose of this study was to investigate the performance of quantitative parameters derived from dual-energy CT (DECT) in the preoperative diagnosis of regional metastatic lymph nodes (LNs) in patients with colorectal cancer. SUBJECTS AND METHODS. Triphasic contrast-enhanced DECT was performed for 178 patients with colon or high rectal cancer. The morphologic criteria, short-axis diameter, and quantitative DECT parameters of the largest regional LN were measured and compared between pathologically metastatic and nonmetastatic LNs. Univariate and multivariable logistic regression analyses were used to determine the independent DECT parameters for predicting LN metastasis. Diagnostic performance measures were assessed by ROC curve analysis and compared by McNemar test. RESULTS. A total of 178 largest LNs (72 metastatic, 106 nonmetastatic) were identified in 178 patients. The best single DECT parameter for differentiation between metastatic and nonmetastatic LNs was normalized effective atomic number (Z eff ) in the portal venous phase (AUC, 0.871; accuracy, 84.8%). These values were higher than those of morphologic criteria (AUC, 0.505-0.624; accuracy, 47.8-62.4%) and short-axis diameter (AUC, 0.647; accuracy, 66.3%) ( p < 0.05). The diagnostic accuracy of combined normalized iodine concentration in the arterial phase and normalized effective atomic number in the portal venous phase was further improved to 87.1% (AUC, 0.916). CONCLUSION. Quantitative parameters derived from DECT can be used to improve preoperative diagnostic accuracy in evaluation for regional metastatic LNs in patients with colorectal cancer.

An integrated segmentation method for 3D ultrasound carotid artery was proposed. 3D ultrasound image was sliced into transverse, coronal and sagittal 2D images on the carotid bifurcation point. Then, the three images were processed respectively, and the carotid artery contours and thickness were obtained finally. This paper tries to overcome the disadvantages of current computer aided diagnosis method, such as high computational complexity, easily introduced subjective errors et al. The proposed method could get the carotid artery overall information rapidly, accurately and completely. It could be transplanted into clinical usage for atherosclerosis diagnosis and prevention.