Browse Articles
Discover research articles across all indexed journals
Epidemiology and prevention of HPV-driven cervical cancer in urban–rural China: a population-based study of 60,280 women
Popeye Sign in Transthyretin Amyloidosis
Exceeding 1.5 °C requires rethinking accountability in climate policy
Cancer classification with radiomics in controlled preclinical models
Abstract The premise of radiomics involves extracting high-dimensional quantitative features from medical images to aid clinical decision-making. While radiomics has shown promise in predicting disease characteristics, concerns regarding confounders, reproducibility, and interpretability limit its clinical adoption. In this study, we assessed the ability of radiomic features extracted from contoured CT images to classify two distinct tumour models, CT26 colorectal cancer (CRC) and 4T1 breast cancer (BC), in a highly controlled murine setting. We aimed to provide compelling data for the role of radiomics as a reliable cancer biomarker. We benchmarked radiomics-based classification against previously established blood-based biomarkers, including leukocyte populations and plasma proteins. Feature filtering reduced the original 1409 radiomic features to 18 non-redundant, high-importance predictors, primarily texture-based transformations. Unsupervised clustering via UMAP revealed that radiomics-based features did not segregate tumour types as effectively as blood biomarkers, suggesting potential confounding factors. Supervised machine learning using Random Forest showed that radiomic features achieved a classification accuracy of 0.87, lower than the 0.96 and 0.99 accuracies obtained with cell and plasma biomarkers, respectively. Furthermore, integrating radiomics with blood biomarkers did not enhance classification performance, and feature importance analysis using SHAP identified blood-based markers as the dominant predictors. These findings suggest that while radiomics retains some predictive value, it is less effective than blood biomarkers in this classification task and does not significantly contribute to multimodal tumour classification models. Our study underscores the need for further standardization and validation of radiomics before its clinical implementation.
Peripheral Artery Disease in the Legs
Dynamics of brain connectivity across the Alzheimer’s disease spectrum through magnetoencephalography
The Essential Role of States in Protecting Immunization Access
Technology is changing how we write — and how we think about writing
TIPE2 serves as a favorable prognostic biomarker and suppresses cholangiocarcinoma progression by targeting RAC1-mediated integrin αvβ6 trafficking
Toxic Erythema of Chemotherapy
Defending endangered trees against climate change and hungry goats
Numerical modeling of coupled stress-fracture evolution in water-resisting key strata during longwall mining
Final Analysis of a Study of Etranacogene Dezaparvovec for Hemophilia B
A lightweight YOLO11n seg framework for real time surface crack detection with segmentation
Abstract The recognition of superficial cracks is essential to ensure the safety, durability, and longevity of civil infrastructure such as bridges, pavements, tunnels, and buildings. Traditional crack detection methods have been largely based on manual inspections and classical image processing techniques, including edge detection, thresholding, and morphological operations. With the rapid advancement of computer vision and deep learning, significant progress has been made in automating crack detection. To gain insight into previous research, we reviewed some studies from the past few years and identified YOLO11 as the most suitable model for crack detection tasks. In this study, we propose a deep learning-based framework for surface crack detection using the Crack-Seg dataset and the YOLO11n-seg architecture. Experimental results demonstrate that YOLO11n-seg achieves strong performance on the Crack-Seg dataset. The suggested model reaches a Precision of 78.8%, which is comparable to heavy baselines. Our results show that the suggested lightweight model, with just 2.8 million parameters, has a Box mAP@50 of 76.2% with a Mask mAP@50 of 58.7%. Most importantly, the model reaches an inference rate of 3.6ms for each image (on Tesla T4), allowing for ultra-fast processing in highly automated inspection systems. These findings establish a new benchmark for edge-deployable crack recognition, demonstrating the possibility that the YOLO11n-seg architecture may provide acceptable segmentation performance with lower computational cost than large, traditional methods.