Search Results (11)

Background: In clinical practice, identifying the location and extent of tumors and lesions is crucial for disease diagnosis and treatment. Artificial intelligence, particularly deep neural networks, offers precise and automated segmentation, yet limited data and high computational demands often hinder its application. Transfer learning helps mitigate these challenges by significantly reducing computational costs, although applying these models can still be resource intensive. This study aims to present flexible and computationally efficient architecture that leverages transfer learning and delivers highly accurate results across various medical imaging problems. Methods: We evaluated three datasets with varying similarities to ImageNet: ISIC 2018 (skin lesions), CBIS-DDSM (breast masses), and the Shenzhen and Montgomery CXR Set (lung segmentation). An ablation study on ISIC 2018 tested various pre-trained backbones, architectures, and loss functions. Results: The optimal configuration—DeepLabV3+ with a pre-trained ResNet50 backbone and Log-Cosh Dice loss—was validated on the remaining datasets, achieving state-of-the-art results. Conclusion: Computationally simpler architectures can deliver robust performance without extensive resources, establishing DeepLabV3+ with the ResNet50 as a baseline for future studies. In the medical domain, enhancing data quality is more critical for improving segmentation accuracy than increasing model complexity. Full article

Drying is one of the most effective preservation methods for extending the shelf-life of perishable foods. The microbial safety of low-moisture food products had not been recognized as a concern until outbreaks reported over the past decade in products contaminated with bacterial pathogens, in particular Salmonella. There is now an urgent need to understand the influence of process conditions on the thermal inactivation of pathogens in various drying operations. This study aimed to develop a predictive model for Salmonella inactivation in diced apples during hot air drying and in high-humidity heating in closed environments. Fresh-cut apple cubes (6 mm) inoculated with a cocktail of Salmonella enterica strains (Enteritidis PT30, Montevideo 488275, and Agona 447967) were placed in a customized box inside an oven for three different treatments: (1) open-box drying at oven temperature 90 °C (Drying-90); (2) close-box pre-drying heating at 90 °C (PD heating-90); and (3) close-box pre-drying heating at 70 °C (PD heating-70). Air temperature, relative humidity (RH), and sample temperatures were monitored, and Salmonella survival was measured at multiple time intervals. After 10 min, the air RH reached 66% in PD heating-90 and 74% in PD heating-70, versus 30% in Drying-90. A 5-log reduction in Salmonella was achieved in 8.5 min in PD heating-90, and 14 min in PD heating-70, compared to 28.7 min in Drying-90. A mathematical model using sample surface RH and sample temperature profiles accurately predicted Salmonella inactivation across all treatments (RMSE = 0.92 log CFU/g, R² = 0.86), with thermal death parameters comparable to isothermal studies. This study underscores the role of humidity in enhancing microbial reduction during drying and proposes high-humidity pre-drying heating as an effective control step. The developed model shows promise for real-time prediction of microbial inactivation in complex drying environments with dynamic temperature and humidity conditions. Full article

The symmetry between production efficiency and safety is a crucial aspect of industrial operations. To enhance the identification of proper safety harness use by workers at height, this study introduces a machine vision approach as a substitute for manual supervision. By focusing on the safety rope that connects the worker to an anchor point, we propose a semantic segmentation mask annotation principle to evaluate proper harness use. We introduce CEMFormer, a novel semantic segmentation model utilizing ConvNeXt as the backbone, which surpasses the traditional ResNet in accuracy. Efficient Multi-Scale Attention (EMA) is incorporated to optimize channel weights and integrate spatial information. Mask2Former serves as the segmentation head, enhanced by Poly Loss for classification and Log-Cosh Dice Loss for mask loss, thereby improving training efficiency. Experimental results indicate that CEMFormer achieves a mean accuracy of 92.31%, surpassing the baseline and five state-of-the-art models. Ablation studies underscore the contribution of each component to the model’s accuracy, demonstrating the effectiveness of the proposed approach in ensuring worker safety. Full article

Purpose: The accuracy of target delineation in radiation treatment planning of high-grade gliomas (HGGs) is crucial to achieve high tumor control, while minimizing treatment-related toxicity. Magnetic resonance imaging (MRI) represents the standard imaging modality for delineation of gliomas with inherent limitations in accurately determining the microscopic extent of tumors. The purpose of this study was to assess the survival impact of multi-observer delineation variability of multiparametric MRI (mpMRI) and [¹⁸F]-FET PET/CT. Materials and Methods: Thirty prospectively included patients with histologically confirmed HGGs underwent a PET/CT and mpMRI including diffusion-weighted imaging (DWI: b0, b1000, ADC), contrast-enhanced T1-weighted imaging (T1-Gado), T2-weighted fluid-attenuated inversion recovery (T2Flair), and perfusion-weighted imaging with computation of relative cerebral blood volume (rCBV) and K2 maps. Nine radiation oncologists delineated the PET/CT and MRI sequences. Spatial similarity (Dice similarity coefficient: DSC) was calculated between the readers for each sequence. Impact of the DSC on progression-free survival (PFS) and overall survival (OS) was assessed using Kaplan–Meier curves and the log-rank test. Results: The highest DSC mean values were reached for morphological sequences, ranging from 0.71 +/− 0.18 to 0.84 +/− 0.09 for T2Flair and T1Gado, respectively, while metabolic volumes defined by PET/CT achieved a mean DSC of 0.75 +/− 0.11. rCBV variability (mean DSC0.32 +/− 0.20) significantly impacted PFS (p = 0.02) and OS (p = 0.002). Conclusions: Our data suggest that the T1-Gado and T2Flair sequences were the most reproducible sequences, followed by PET/CT. Reproducibility for functional sequences was low, but rCBV inter-reader similarity significantly impacted PFS and OS. Full article

This study assessed the extent of L. monocytogenes transfer from onions to the surface of a commercial dicer, from inoculated onions to uninoculated onions, and the efficacy of various sanitizers during the subsequent flume washing of diced onions. Spanish yellow onions (Allium cepa L.) were dip-inoculated in a 3-strain avirulent L. monocytogenes cocktail (5.9 or 4.2 log CFU/50 g) and air-dried. After dicing one 2.2 kg batch of onions inoculated at ~5.9 log CFU/50 g followed by ten uninoculated batches of 2.2 kg each, L. monocytogenes progressively decreased from 4.6 to 2.6 log CFU/50 g in baches 1 through 10, respectively. After onions inoculated at ~4.0 log CFU/g were diced and flume washed for 2 min in tap water, electrolyzed water containing 55 ppm free chlorine, 80 ppm free chlorine from a commercial sanitizer, or 80 ppm peroxyacetic acid and dewatered on a mechanical shaker table, L. monocytogenes populations decreased 0.4, 0.3, 1.4, and 1.0 log, respectively, with populations of ~1.2 log CFU/mL in water for all three sanitizers. These findings should be useful in future risk assessments and aid in the development of improved industry guidelines to better enhance the safety of diced onions. Full article

Crop classification of large-scale agricultural land is crucial for crop monitoring and yield estimation. Hyperspectral image classification has proven to be an effective method for this task. Most current popular hyperspectral image classification methods are based on image classification, specifically on convolutional neural networks (CNNs) and recurrent neural networks (RNNs). In contrast, this paper focuses on methods based on semantic segmentation and proposes a new transformer-based approach called HyperSFormer for crop hyperspectral image classification. The key enhancement of the proposed method is the replacement of the encoder in SegFormer with an improved Swin Transformer while keeping the SegFormer decoder. The entire model adopts a simple and uniform transformer architecture. Additionally, the paper introduces the hyper patch embedding (HPE) module to extract spectral and local spatial information from the hyperspectral images, which enhances the effectiveness of the features used as input for the model. To ensure detailed model processing and achieve end-to-end hyperspectral image classification, the transpose padding upsample (TPU) module is proposed for the model’s output. In order to address the problem of insufficient and imbalanced samples in hyperspectral image classification, the paper designs an adaptive min log sampling (AMLS) strategy and a loss function that incorporates dice loss and focal loss to assist model training. Experimental results using three public hyperspectral image datasets demonstrate the strong performance of HyperSFormer, particularly in the presence of imbalanced sample data, complex negative samples, and mixed sample classes. HyperSFormer outperforms state-of-the-art methods, including fast patch-free global learning (FPGA), a spectral–spatial-dependent global learning framework (SSDGL), and SegFormer, by at least 2.7% in the mean intersection over union (mIoU). It also improves the overall accuracy and average accuracy values by at least 0.9% and 0.3%, respectively, and the kappa coefficient by at least 0.011. Furthermore, ablation experiments were conducted to determine the optimal hyperparameter and loss function settings for the proposed method, validating the rationality of these settings and the fusion loss function. Full article

Accurate radiation therapy (RT) targeting is crucial for glioblastoma treatment but may be challenging using clinical imaging alone due to the infiltrative nature of glioblastomas. Precise targeting by whole-brain spectroscopic MRI, which maps tumor metabolites including choline (Cho) and N-acetylaspartate (NAA), can quantify early treatment-induced molecular changes that other traditional modalities cannot measure. We developed a pipeline to determine how spectroscopic MRI changes during early RT are associated with patient outcomes to provide insight into the utility of adaptive RT planning. Data were obtained from a study (NCT03137888) where glioblastoma patients received high-dose RT guided by the pre-RT Cho/NAA twice normal (Cho/NAA ≥ 2x) volume, and received spectroscopic MRI scans pre- and mid-RT. Overlap statistics between pre- and mid-RT scans were used to quantify metabolic activity changes after two weeks of RT. Log-rank tests were used to quantify the relationship between imaging metrics and patient overall and progression-free survival (OS/PFS). Patients with lower Jaccard/Dice coefficients had longer PFS (p = 0.045 for both), and patients with lower Jaccard/Dice coefficients had higher OS trending towards significance (p = 0.060 for both). Cho/NAA ≥ 2x volumes changed significantly during early RT, putting healthy tissue at risk of irradiation, and warranting further study into using adaptive RT planning. Full article

Monitoring the health and safety of forests has become a rising problem with the advent of global wildfires, rampant logging, and reforestation efforts. This paper proposes a model for the automatic segmentation and classification of aerial forest imagery. The model is based on U-net architecture and relies on dice coefficients, binary cross-entropy, and accuracy as loss functions. While models without autoencoder-based structures can only reach a dice coefficient of 45%, the proposed model can achieve a dice coefficient of 79.85%. In addition, for barren adn dense forestry image classification, the proposed model can achieve 82.51%. This paper demonstrates how complex convolutional neural networks can be applied to aerial forest images to help preserve and save the forest environment. Full article

Due to the complexity of ocean environments, inhomogeneous phenomenon always exist in SAR images of oil spills on the sea surface. In order to address this issue, a universal parameter adaptive Gamma-Log net for detecting oil spills in inhomogeneous SAR images is proposed in this paper. The Gamma-Log net consists of an image feature division module, a correction parameter extraction module, a Gamma-Log correction module and a feature integration module. The normalized input image features are divided into four blocks for correction in the image feature division module. According to the input characteristics, the Gamma-Log correction input parameters are obtained in the correction parameter extraction module. Subsequently, an adaptive method is introduced to adjust the parameters independently by the network to improve efficiency. Then, the input features are corrected in the Gamma-Log correction module by Gamma correction and logarithmic correction. Both correction methods can adjust the gray imbalance in the image and change the overall gray value and contrast. The separated feature blocks are finally reunited together by the feature integration module. In order to avoid information loss, an attention mechanism is added to this module. In the experiments, by adding Gamma-Log Net to multiple semantic segmentation networks, the MIoU and dice indicators increased to some extent, and the HD distance(Hausdorff-95) decreased. Our work demonstrates that the Gamma-Log net can be helpful for oil spill detection in inhomogeneous SAR images. Full article

Problem. Image biomarker analysis, also known as radiomics, is a tool for tissue characterization and treatment prognosis that relies on routinely acquired clinical images and delineations. Due to the uncertainty in image acquisition, processing, and segmentation (delineation) protocols, radiomics often lack reproducibility. Radiomics harmonization techniques have been proposed as a solution to reduce these sources of uncertainty and/or their influence on the prognostic model performance. A relevant question is how to estimate the protocol-induced uncertainty of a specific image biomarker, what the effect is on the model performance, and how to optimize the model given the uncertainty. Methods. Two non-small cell lung cancer (NSCLC) cohorts, composed of 421 and 240 patients, respectively, were used for training and testing. Per patient, a Monte Carlo algorithm was used to generate three hundred synthetic contours with a surface dice tolerance measure of less than 1.18 mm with respect to the original GTV. These contours were subsequently used to derive 104 radiomic features, which were ranked on their relative sensitivity to contour perturbation, expressed in the parameter η. The top four (low η) and the bottom four (high η) features were selected for two models based on the Cox proportional hazards model. To investigate the influence of segmentation uncertainty on the prognostic model, we trained and tested the setup in 5000 augmented realizations (using a Monte Carlo sampling method); the log-rank test was used to assess the stratification performance and stability of segmentation uncertainty. Results. Although both low and high η setup showed significant testing set log-rank p-values (p = 0.01) in the original GTV delineations (without segmentation uncertainty introduced), in the model with high uncertainty, to effect ratio, only around 30% of the augmented realizations resulted in model performance with p < 0.05 in the test set. In contrast, the low η setup performed with a log-rank p < 0.05 in 90% of the augmented realizations. Moreover, the high η setup classification was uncertain in its predictions for 50% of the subjects in the testing set (for 80% agreement rate), whereas the low η setup was uncertain only in 10% of the cases. Discussion. Estimating image biomarker model performance based only on the original GTV segmentation, without considering segmentation, uncertainty may be deceiving. The model might result in a significant stratification performance, but can be unstable for delineation variations, which are inherent to manual segmentation. Simulating segmentation uncertainty using the method described allows for more stable image biomarker estimation, selection, and model development. The segmentation uncertainty estimation method described here is universal and can be extended to estimate other protocol uncertainties (such as image acquisition and pre-processing). Full article

The purpose of the current study was to explore if and how additional-language learners may show changes in phraseological patterns over the course of a stay in a target-language environment. In particular, we focused on noun+adjective combinations produced by a group of additional-language speakers of French at three points in time, spanning 21 months and including an academic year in France. We extracted each combination from a longitudinal corpus and determined frequency counts and two strength-of-association measures (Mutual information [MI] score and Log Dice) for each combination. Separate analyses were conducted for frequency and the strength-of-association measures, revealing that phraseological patterns are significantly predicted by adjective position in the case of all three measures, and that MI scores showed significant change over time. We interpret the results in light of past research that has reported contradictory findings concerning change in phraseological patterns following an immersion experience. Full article