Confocal laser endomicroscopy(CLE)has become an indispensable tool in the diagnosis and detection of gastrointestinal(GI)diseases due to its high-resolution and high-contrast imaging capabilities.However,the early-sta...Confocal laser endomicroscopy(CLE)has become an indispensable tool in the diagnosis and detection of gastrointestinal(GI)diseases due to its high-resolution and high-contrast imaging capabilities.However,the early-stage imaging changes of gastrointestinal disorders are often subtle,and traditional medical image analysis methods rely heavily on manual interpretation,which is time-consuming,subject to observer variability,and inefficient for accurate lesion identification across large-scale image datasets.With the introduction of artificial intelligence(AI)technologies,AI-driven CLE image analysis systems can automatically extract pathological features and have demonstrated significant clinical value in lesion recognition,classification diagnosis,and malignancy prediction of GI diseases.These systems greatly enhance diagnostic efficiency and early detection capabilities.This review summarizes the applications of AI-assisted CLE in GI diseases,analyzes the limitations of current technologies,and explores future research directions.It is expected that the deep integration of AI and confocal imaging technologies will provide strong support for precision diagnosis and personalized treatment in the field of gastrointestinal disorders.展开更多
Starting from the diffraction imaging process,we have discussed the relationship between optical imaging system and fractional Fourier transform, and proposed a specific system which can form an inverse amplified imag...Starting from the diffraction imaging process,we have discussed the relationship between optical imaging system and fractional Fourier transform, and proposed a specific system which can form an inverse amplified image of input function.展开更多
BACKGROUND Conventional plain X-ray images of rats,the most common animals used as degeneration models,exhibit unclear vertebral structure and blurry intervertebral disc spaces due to their small size,slender vertebra...BACKGROUND Conventional plain X-ray images of rats,the most common animals used as degeneration models,exhibit unclear vertebral structure and blurry intervertebral disc spaces due to their small size,slender vertebral bodies.AIM To apply molybdenum target X-ray photography in the evaluation of caudal intervertebral disc(IVD)degeneration in rat models.METHODS Two types of rat caudal IVD degeneration models(needle-punctured model and endplate-destructed model)were established,and their effectiveness was verified using nuclear magnetic resonance imaging.Molybdenum target inspection and routine plain X-ray were then performed on these models.Additionally,four observers were assigned to measure the intervertebral height of degenerated segments on molybdenum target plain X-ray images and routine plain X-ray images,respectively.The degeneration was evaluated and statistical analysis was subsequently conducted.RESULTS Nine rats in the needle-punctured model and 10 rats in the endplate-destructed model were effective.Compared with routine plain X-ray images,molybdenum target plain X-ray images showed higher clarity,stronger contrast,as well as clearer and more accurate structural development.The McNemar test confirmed that the difference was statistically significant(P=0.031).In the two models,the reliability of the intervertebral height measured by the four observers on routine plain X-ray images was poor(ICC<0.4),while the data obtained from the molybdenum target plain X-ray images were more reliable.CONCLUSIONMolybdenum target inspection can obtain clearer images and display fine calcification in the imaging evaluation of caudal IVD degeneration in rats,thus ensuring a more accurate evaluation of degeneration.展开更多
Imaging evaluation of lymph node metastasis and infiltration faces problems such as low artificial outline efficiency and insufficient consistency.Deep learning technology based on convolutional neural networks has gr...Imaging evaluation of lymph node metastasis and infiltration faces problems such as low artificial outline efficiency and insufficient consistency.Deep learning technology based on convolutional neural networks has greatly improved the technical effect of radiomics in lymph node pathological characteristics analysis and efficacy monitoring through automatic lymph node detection,precise segmentation and three-dimensional reconstruction algorithms.This review focuses on the automatic lymph node segmentation model,treatment response prediction algorithm and benign and malignant differential diagnosis system for multimodal imaging,in order to provide a basis for further research on artificial intelligence to assist lymph node disease management and clinical decision-making,and provide a reference for promoting the construction of a system for accurate diagnosis,personalized treatment and prognostic evaluation of lymph node-related diseases.展开更多
Integrating multiple medical imaging techniques,including Magnetic Resonance Imaging(MRI),Computed Tomography,Positron Emission Tomography(PET),and ultrasound,provides a comprehensive view of the patient health status...Integrating multiple medical imaging techniques,including Magnetic Resonance Imaging(MRI),Computed Tomography,Positron Emission Tomography(PET),and ultrasound,provides a comprehensive view of the patient health status.Each of these methods contributes unique diagnostic insights,enhancing the overall assessment of patient condition.Nevertheless,the amalgamation of data from multiple modalities presents difficulties due to disparities in resolution,data collection methods,and noise levels.While traditional models like Convolutional Neural Networks(CNNs)excel in single-modality tasks,they struggle to handle multi-modal complexities,lacking the capacity to model global relationships.This research presents a novel approach for examining multi-modal medical imagery using a transformer-based system.The framework employs self-attention and cross-attention mechanisms to synchronize and integrate features across various modalities.Additionally,it shows resilience to variations in noise and image quality,making it adaptable for real-time clinical use.To address the computational hurdles linked to transformer models,particularly in real-time clinical applications in resource-constrained environments,several optimization techniques have been integrated to boost scalability and efficiency.Initially,a streamlined transformer architecture was adopted to minimize the computational load while maintaining model effectiveness.Methods such as model pruning,quantization,and knowledge distillation have been applied to reduce the parameter count and enhance the inference speed.Furthermore,efficient attention mechanisms such as linear or sparse attention were employed to alleviate the substantial memory and processing requirements of traditional self-attention operations.For further deployment optimization,researchers have implemented hardware-aware acceleration strategies,including the use of TensorRT and ONNX-based model compression,to ensure efficient execution on edge devices.These optimizations allow the approach to function effectively in real-time clinical settings,ensuring viability even in environments with limited resources.Future research directions include integrating non-imaging data to facilitate personalized treatment and enhancing computational efficiency for implementation in resource-limited environments.This study highlights the transformative potential of transformer models in multi-modal medical imaging,offering improvements in diagnostic accuracy and patient care outcomes.展开更多
Artificial intelligence(AI),particularly deep learning algorithms utilizing convolutional neural networks,plays an increasingly pivotal role in enhancing medical image examination.It demonstrates the potential for imp...Artificial intelligence(AI),particularly deep learning algorithms utilizing convolutional neural networks,plays an increasingly pivotal role in enhancing medical image examination.It demonstrates the potential for improving diagnostic accuracy within dental care.Orthopantomograms(OPGs)are essential in dentistry;however,their manual interpretation is often inconsistent and tedious.To the best of our knowledge,this is the first comprehensive application of YOLOv5m for the simultaneous detection and classification of six distinct dental pathologies using panoramic OPG images.The model was trained and refined on a custom dataset that began with 232 panoramic radiographs and was later expanded to 604 samples.These included annotated subclasses representing Caries,Infection,Impacted Teeth,Fractured Teeth,Broken Crowns,and Healthy conditions.The training was performed using GPU resources alongside tuned hyperparameters of batch size,learning rate schedule,and early stopping tailored for generalization to prevent overfitting.Evaluation on a held-out test set showed strong performance in the detection and localization of various dental pathologies and robust overall accuracy.At an IoU of 0.5,the system obtained a mean precision of 94.22%and recall of 90.42%,with mAP being 93.71%.This research confirms the use of YOLOv5m as a robust,highly efficient AI technology for the analysis of dental pathologies using OPGs,providing a clinically useful solution to enhance workflow efficiency and aid in sustaining consistency in complex multi-dimensional case evaluations.展开更多
BACKGROUND Congestive hepatopathy,also known as nutmeg liver,is liver damage secondary to chronic heart failure(HF).Its morphological characteristics in terms of medical imaging are not defined and remain unclear.AIM ...BACKGROUND Congestive hepatopathy,also known as nutmeg liver,is liver damage secondary to chronic heart failure(HF).Its morphological characteristics in terms of medical imaging are not defined and remain unclear.AIM To leverage machine learning to capture imaging features of congestive hepatopathy using incidentally acquired computed tomography(CT)scans.METHODS We retrospectively analyzed 179 chronic HF patients who underwent echocardiography and CT within one year.Right HF severity was classified into three grades.Liver CT images at the paraumbilical vein level were used to develop a ResNet-based machine learning model to predict tricuspid regurgitation(TR)severity.Model accuracy was compared with that of six gastroenterology and four radiology experts.RESULTS In the included patients,120 were male(mean age:73.1±14.4 years).The accuracy of the results predicting TR severity from a single CT image for the machine learning model was significantly higher than the average accuracy of the experts.The model was found to be exceptionally reliable for predicting severe TR.CONCLUSION Deep learning models,particularly those using ResNet architectures,can help identify morphological changes associated with TR severity,aiding in early liver dysfunction detection in patients with HF,thereby improving outcomes.展开更多
Background:Accurate classification of normal blood cells is a critical foundation for automated hematological analysis,including the detection of pathological conditions like leukemia.While convolutional neural networ...Background:Accurate classification of normal blood cells is a critical foundation for automated hematological analysis,including the detection of pathological conditions like leukemia.While convolutional neural networks(CNNs)excel in local feature extraction,their ability to capture global contextual relationships in complex cellular morphologies is limited.This study introduces a hybrid CNN-Transformer framework to enhance normal blood cell classification,laying the groundwork for future leukemia diagnostics.Methods:The proposed architecture integrates pre-trained CNNs(ResNet50,EfficientNetB3,InceptionV3,CustomCNN)with Vision Transformer(ViT)layers to combine local and global feature modeling.Four hybrid models were evaluated on the publicly available Blood Cell Images dataset from Kaggle,comprising 17,092 annotated normal blood cell images across eight classes.The models were trained using transfer learning,fine-tuning,and computational optimizations,including cross-model parameter sharing to reduce redundancy by reusing weights across CNN backbones and attention-guided layer pruning to eliminate low-contribution layers based on attention scores,improving efficiency without sacrificing accuracy.Results:The InceptionV3-ViT model achieved a weighted accuracy of 97.66%(accounting for class imbalance by weighting each class’s contribution),a macro F1-score of 0.98,and a ROC-AUC of 0.998.The framework excelled in distinguishing morphologically similar cell types demonstrating robustness and reliable calibration(ECE of 0.019).The framework addresses generalization challenges,including class imbalance and morphological similarities,ensuring robust performance across diverse cell types.Conclusion:The hybrid CNN-Transformer framework significantly improves normal blood cell classification by capturing multi-scale features and long-range dependencies.Its high accuracy,efficiency,and generalization position it as a strong baseline for automated hematological analysis,with potential for extension to leukemia subtype classification through future validation on pathological samples.展开更多
This study explores the relationship between cardiac activity and biochemical indicators in Pacific oysters(Crassostrea gigas)during cold storage to develop a nondestructive vitality assessment method.Oysters were sto...This study explores the relationship between cardiac activity and biochemical indicators in Pacific oysters(Crassostrea gigas)during cold storage to develop a nondestructive vitality assessment method.Oysters were stored at−1℃ for 14 d,with cardiac patterns tracked via image analysis,and biochemical markers(pH,adenosine triphosphate(ATP)-related compounds,and adenylate energy charge(AEC))were assessed.Five cardiac patterns were identified,with regular alternating contractions common early but decreasing over time,aligning with declines in AEC(44.11%-35.52%)and pH(6.98-6.55).The intervals between ventricular and atrial contractions rose from 4.2 to 5.6 s,strongly correlating with biochemical signs of vitality loss.Image analysis revealed characteristic waveforms for each cardiac pattern,despite amplitude variations caused by optical artifacts.These findings indicate that cardiac pattern analysis via image processing could be an effective nondestructive indicator of oyster vitality,offering a novel approach to quality control in shellfish storage and distribution.展开更多
Medical image analysis has become a cornerstone of modern healthcare,driven by the exponential growth of data from imaging modalities such as MRI,CT,PET,ultrasound,and X-ray.Traditional machine learning methods have m...Medical image analysis has become a cornerstone of modern healthcare,driven by the exponential growth of data from imaging modalities such as MRI,CT,PET,ultrasound,and X-ray.Traditional machine learning methods have made early contributions;however,recent advancements in deep learning(DL)have revolutionized the field,offering state-of-the-art performance in image classification,segmentation,detection,fusion,registration,and enhancement.This comprehensive review presents an in-depth analysis of deep learning methodologies applied across medical image analysis tasks,highlighting both foundational models and recent innovations.The article begins by introducing conventional techniques and their limitations,setting the stage for DL-based solutions.Core DL architectures,including Convolutional Neural Networks(CNNs),Recurrent Neural Networks(RNNs),Generative Adversarial Networks(GANs),Vision Transformers(ViTs),and hybrid models,are discussed in detail,including their advantages and domain-specific adaptations.Advanced learning paradigms such as semi-supervised learning,selfsupervised learning,and few-shot learning are explored for their potential to mitigate data annotation challenges in clinical datasets.This review further categorizes major tasks in medical image analysis,elaborating on how DL techniques have enabled precise tumor segmentation,lesion detection,modality fusion,super-resolution,and robust classification across diverse clinical settings.Emphasis is placed on applications in oncology,cardiology,neurology,and infectious diseases,including COVID-19.Challenges such as data scarcity,label imbalance,model generalizability,interpretability,and integration into clinical workflows are critically examined.Ethical considerations,explainable AI(XAI),federated learning,and regulatory compliance are discussed as essential components of real-world deployment.Benchmark datasets,evaluation metrics,and comparative performance analyses are presented to support future research.The article concludes with a forward-looking perspective on the role of foundation models,multimodal learning,edge AI,and bio-inspired computing in the future of medical imaging.Overall,this review serves as a valuable resource for researchers,clinicians,and developers aiming to harness deep learning for intelligent,efficient,and clinically viable medical image analysis.展开更多
Given the importance of sentiment analysis in diverse environments,various methods are used for image sentiment analysis,including contextual sentiment analysis that utilizes character and scene relationships.However,...Given the importance of sentiment analysis in diverse environments,various methods are used for image sentiment analysis,including contextual sentiment analysis that utilizes character and scene relationships.However,most existing works employ character faces in conjunction with context,yet lack the capacity to analyze the emotions of characters in unconstrained environments,such as when their faces are obscured or blurred.Accordingly,this article presents the Adaptive Multi-Channel Sentiment Analysis Network(AMSA),a contextual image sentiment analysis framework,which consists of three channels:body,face,and context.AMSA employs Multi-task Cascaded Convolutional Networks(MTCNN)to detect faces within body frames;if detected,facial features are extracted and fused with body and context information for emotion recognition.If not,the model leverages body and context features alone.Meanwhile,to address class imbalance in the EMOTIC dataset,Focal Loss is introduced to improve classification performance,especially for minority emotion categories.Experimental results have shown that certain sentiment categories with lower representation in the dataset demonstrate leading classification accuracy,the AMSA yields a 2.53%increase compared with state-of-the-art methods.展开更多
Subcutaneous vein network plays important roles to maintain microcirculation that is related to some diagnostic aspects.Despite developments of optical imaging technologies,still the difficulties about deep skin vascu...Subcutaneous vein network plays important roles to maintain microcirculation that is related to some diagnostic aspects.Despite developments of optical imaging technologies,still the difficulties about deep skin vascular imaging have been continued.On the other hand,since hemoglobin con-centration of human blood has key role in the veins imaging by optical manner,the used wavelength in vascular imaging,must be chosen considering absorption of hemoglobin.In this research,we constructed a near infrared(NIR)light source because of lower absorption of hemoglobin in this optical region.To obtain vascular image,reflectance geometry was used.Next,from recorded images,vascular network analysis,such as calculation of width of vascular of interest and complexity of selected region were implemented.By comparing with other modalities,we observed that proposed imaging system has great advantages including nonionized radiation,moderate penetration depth of 0.5-3 mm and diameter of 1 mm,cost-effective and algorit hmic simplicity for analysis.展开更多
Facial beauty analysis is an important topic in human society.It may be used as a guidance for face beautification applications such as cosmetic surgery.Deep neural networks(DNNs)have recently been adopted for facial ...Facial beauty analysis is an important topic in human society.It may be used as a guidance for face beautification applications such as cosmetic surgery.Deep neural networks(DNNs)have recently been adopted for facial beauty analysis and have achieved remarkable performance.However,most existing DNN-based models regard facial beauty analysis as a normal classification task.They ignore important prior knowledge in traditional machine learning models which illustrate the significant contribution of the geometric features in facial beauty analysis.To be specific,landmarks of the whole face and facial organs are introduced to extract geometric features to make the decision.Inspired by this,we introduce a novel dual-branch network for facial beauty analysis:one branch takes the Swin Transformer as the backbone to model the full face and global patterns,and another branch focuses on the masked facial organs with the residual network to model the local patterns of certain facial parts.Additionally,the designed multi-scale feature fusion module can further facilitate our network to learn complementary semantic information between the two branches.In model optimisation,we propose a hybrid loss function,where especially geometric regulation is introduced by regressing the facial landmarks and it can force the extracted features to convey facial geometric features.Experiments performed on the SCUT-FBP5500 dataset and the SCUT-FBP dataset demonstrate that our model outperforms the state-of-the-art convolutional neural networks models,which proves the effectiveness of the proposed geometric regularisation and dual-branch structure with the hybrid network.To the best of our knowledge,this is the first study to introduce a Vision Transformer into the facial beauty analysis task.展开更多
The Ki67 index (KI) is a standard clinical marker for tumor proliferation;however, its application is hindered by intratumoral heterogeneity. In this study, we used digital image analysis to comprehensively analyze Ki...The Ki67 index (KI) is a standard clinical marker for tumor proliferation;however, its application is hindered by intratumoral heterogeneity. In this study, we used digital image analysis to comprehensively analyze Ki67 heterogeneity and distribution patterns in breast carcinoma. Using Smart Pathology software, we digitized and analyzed 42 excised breast carcinoma Ki67 slides. Boxplots, histograms, and heat maps were generated to illustrate the KI distribution. We found that 30% of cases (13/42) exhibited discrepancies between global and hotspot KI when using a 14% KI threshold for classification. Patients with higher global or hotspot KI values displayed greater heterogenicity. Ki67 distribution patterns were categorized as randomly distributed (52%, 22/42), peripheral (43%, 18/42), and centered (5%, 2/42). Our sampling simulator indicated analyzing more than 10 high-power fields was typically required to accurately estimate global KI, with sampling size being correlated with heterogeneity. In conclusion, using digital image analysis in whole-slide images allows for comprehensive Ki67 profile assessment, shedding light on heterogeneity and distribution patterns. This spatial information can facilitate KI surveys of breast cancer and other malignancies.展开更多
Recognizing the variation of genetic resources is the first step in selection.One of the most important variations in grain crops is the uniformity of seed grain weight,which can be converted into seed size.However,it...Recognizing the variation of genetic resources is the first step in selection.One of the most important variations in grain crops is the uniformity of seed grain weight,which can be converted into seed size.However,it has been challenging since it needs high labor costs and time to measure it on a large scale.The current study used an image analysis technique to measure the grain seed area of about 100 seeds per accession with 64 germplasm of Tartary buckwheat(Fagopyrum tataricum)to study variation among and within them.To understand the nature of variation,skewness and kurtosis analysis of probability density function curve for seed area were used.As a result,a large variation among and within accessions was found.This means that the seed sizes within an accession are not uniform in this given cleistogamous species due to its non-uniform flowering time.This implies that the seed size should be considered an important factor for the germplasm enhancement program.展开更多
Methods and procedures of three-dimensional （3D） characterization of the pore structure features in the packed ore particle bed are focused. X-ray computed tomography was applied to deriving the cross-sectional imag...Methods and procedures of three-dimensional （3D） characterization of the pore structure features in the packed ore particle bed are focused. X-ray computed tomography was applied to deriving the cross-sectional images of specimens with single particle size of 1-2, 2-3, 3-4, 4-5, 5-6, 6-7, 7-8, 8-9, 9-10 ram. Based on the in-house developed 3D image analysis programs using Matlab, the volume porosity, pore size distribution and degree of connectivity were calculated and analyzed in detail. The results indicate that the volume porosity, the mean diameter of pores and the effective pore size （d50） increase with the increasing of particle size. Lognormal distribution or Gauss distribution is mostly suitable to model the pore size distribution. The degree of connectivity investigated on the basis of cluster-labeling algorithm also increases with increasing the particle size approximately.展开更多
It is critical to establish a direct and precise method with a high sensitivity and selectivity in analytical chemistry. In this research, making use of a well known phenomenon of capillary flow, we have proposed an...It is critical to establish a direct and precise method with a high sensitivity and selectivity in analytical chemistry. In this research, making use of a well known phenomenon of capillary flow, we have proposed an image analysis method of nucleic acids at the price of a small amount of sample. When a droplet of the supramolecular complex solution, formed by neutral red and nucleic acids(NA) under an approximate neutral condition, was placed on the hydrophobic surface of dimethyl dichlorosilane pretreated glass slides, and it was evaporated, the supramolecular complex exhibited the periphery of the droplet due to the capillary effect, and accumulated there to form a red capillary flow directed assembly ring(CFDAR). A typical CFDAR has an outer diameter of (2 r ) about 1.18 mm and a ring width(2 δ ) of about 41 μm. Depending on the experimental conditions, a variety of CFDAR can be assembled. The experimental results are in agreement with our former theoretical discussion. It was found that when a droplet volume is 0.1 μL, the fluorescence intensity of the CFDAR formed by the NR NA is in proportion to the content of calf thymus DNA in the range of 0-0.28 ng, fish sperm DNA of 0-0.24 ng and yeast RNA of 0-0.16 ng with the limit of detection(3 σ ) of 1 7, 1.4 and 0.9 pg, respectively for the three nucleic acids.展开更多
The integration of image analysis through deep learning(DL)into rock classification represents a significant leap forward in geological research.While traditional methods remain invaluable for their expertise and hist...The integration of image analysis through deep learning(DL)into rock classification represents a significant leap forward in geological research.While traditional methods remain invaluable for their expertise and historical context,DL offers a powerful complement by enhancing the speed,objectivity,and precision of the classification process.This research explores the significance of image data augmentation techniques in optimizing the performance of convolutional neural networks(CNNs)for geological image analysis,particularly in the classification of igneous,metamorphic,and sedimentary rock types from rock thin section(RTS)images.This study primarily focuses on classic image augmentation techniques and evaluates their impact on model accuracy and precision.Results demonstrate that augmentation techniques like Equalize significantly enhance the model's classification capabilities,achieving an F1-Score of 0.9869 for igneous rocks,0.9884 for metamorphic rocks,and 0.9929 for sedimentary rocks,representing improvements compared to the baseline original results.Moreover,the weighted average F1-Score across all classes and techniques is 0.9886,indicating an enhancement.Conversely,methods like Distort lead to decreased accuracy and F1-Score,with an F1-Score of 0.949 for igneous rocks,0.954 for metamorphic rocks,and 0.9416 for sedimentary rocks,exacerbating the performance compared to the baseline.The study underscores the practicality of image data augmentation in geological image classification and advocates for the adoption of DL methods in this domain for automation and improved results.The findings of this study can benefit various fields,including remote sensing,mineral exploration,and environmental monitoring,by enhancing the accuracy of geological image analysis both for scientific research and industrial applications.展开更多
In traditional trial-and-error method,enhancing the Young's modulus of magnesium alloys while maintaining a favorable ductility has consistently been a challenge.It is a need to explore more efficient and expedite...In traditional trial-and-error method,enhancing the Young's modulus of magnesium alloys while maintaining a favorable ductility has consistently been a challenge.It is a need to explore more efficient and expedited methods to design magnesium alloys with high modulus and ductility.In this study,machine learning(ML)and assisted microstructure control methods are used to design high modulus magnesium alloys.Six key features that influence stiffness and ductility have been extracted in this ML model based on abundant data from literature sources.As a result,predictive models for Young's modulus and elongation are established,with errors<2.4%and 4.5%through XGBoost machine learning model,respectively.Within the given range of six features,the magnesium alloys can be fabricated with the Young's modulus exceeding 50 GPa and an elongation surpassing 6%.As a validation,Mg-Al-Y alloys were experimentally prepared to meet the criteria of six features,achieving Young's modulus of 51.5 GPa,and the elongation of 7%.Moreover,the SHapley Additive exPlanation(SHAP)is introduced to boost the model interpretability.This indicates that balancing the volume fraction of reinforcement,the most important feature,is key to achieve Mg-Al-Y alloys with high Young's modulus and favorable elongation through the two models.Enhancing reinforcement dispersion and reducing the size of reinforcement and grain can further improve the elongation of high-stiffness Mg alloy.展开更多
基金Supported by Interdisciplinary Program of Shanghai Jiao Tong University,No.YG2024 LC01National Natural Science Foundation of China,No.62406190.
文摘Confocal laser endomicroscopy(CLE)has become an indispensable tool in the diagnosis and detection of gastrointestinal(GI)diseases due to its high-resolution and high-contrast imaging capabilities.However,the early-stage imaging changes of gastrointestinal disorders are often subtle,and traditional medical image analysis methods rely heavily on manual interpretation,which is time-consuming,subject to observer variability,and inefficient for accurate lesion identification across large-scale image datasets.With the introduction of artificial intelligence(AI)technologies,AI-driven CLE image analysis systems can automatically extract pathological features and have demonstrated significant clinical value in lesion recognition,classification diagnosis,and malignancy prediction of GI diseases.These systems greatly enhance diagnostic efficiency and early detection capabilities.This review summarizes the applications of AI-assisted CLE in GI diseases,analyzes the limitations of current technologies,and explores future research directions.It is expected that the deep integration of AI and confocal imaging technologies will provide strong support for precision diagnosis and personalized treatment in the field of gastrointestinal disorders.
文摘Starting from the diffraction imaging process,we have discussed the relationship between optical imaging system and fractional Fourier transform, and proposed a specific system which can form an inverse amplified image of input function.
基金Supported by the National Key Research and Development Program of China,No.2017YFA0105404。
文摘BACKGROUND Conventional plain X-ray images of rats,the most common animals used as degeneration models,exhibit unclear vertebral structure and blurry intervertebral disc spaces due to their small size,slender vertebral bodies.AIM To apply molybdenum target X-ray photography in the evaluation of caudal intervertebral disc(IVD)degeneration in rat models.METHODS Two types of rat caudal IVD degeneration models(needle-punctured model and endplate-destructed model)were established,and their effectiveness was verified using nuclear magnetic resonance imaging.Molybdenum target inspection and routine plain X-ray were then performed on these models.Additionally,four observers were assigned to measure the intervertebral height of degenerated segments on molybdenum target plain X-ray images and routine plain X-ray images,respectively.The degeneration was evaluated and statistical analysis was subsequently conducted.RESULTS Nine rats in the needle-punctured model and 10 rats in the endplate-destructed model were effective.Compared with routine plain X-ray images,molybdenum target plain X-ray images showed higher clarity,stronger contrast,as well as clearer and more accurate structural development.The McNemar test confirmed that the difference was statistically significant(P=0.031).In the two models,the reliability of the intervertebral height measured by the four observers on routine plain X-ray images was poor(ICC<0.4),while the data obtained from the molybdenum target plain X-ray images were more reliable.CONCLUSIONMolybdenum target inspection can obtain clearer images and display fine calcification in the imaging evaluation of caudal IVD degeneration in rats,thus ensuring a more accurate evaluation of degeneration.
基金Supported by Clinical Trials from the Nanjing Drum Tower Hospital,Affiliated Hospital of Medical School,Nanjing University,No.2021-LCYJ-MS-11Nanjing Drum Tower Hospital National Natural Science Foundation Youth Cultivation Project,No.2024-JCYJQP-15.
文摘Imaging evaluation of lymph node metastasis and infiltration faces problems such as low artificial outline efficiency and insufficient consistency.Deep learning technology based on convolutional neural networks has greatly improved the technical effect of radiomics in lymph node pathological characteristics analysis and efficacy monitoring through automatic lymph node detection,precise segmentation and three-dimensional reconstruction algorithms.This review focuses on the automatic lymph node segmentation model,treatment response prediction algorithm and benign and malignant differential diagnosis system for multimodal imaging,in order to provide a basis for further research on artificial intelligence to assist lymph node disease management and clinical decision-making,and provide a reference for promoting the construction of a system for accurate diagnosis,personalized treatment and prognostic evaluation of lymph node-related diseases.
基金supported by the Deanship of Research and Graduate Studies at King Khalid University under Small Research Project grant number RGP1/139/45.
文摘Integrating multiple medical imaging techniques,including Magnetic Resonance Imaging(MRI),Computed Tomography,Positron Emission Tomography(PET),and ultrasound,provides a comprehensive view of the patient health status.Each of these methods contributes unique diagnostic insights,enhancing the overall assessment of patient condition.Nevertheless,the amalgamation of data from multiple modalities presents difficulties due to disparities in resolution,data collection methods,and noise levels.While traditional models like Convolutional Neural Networks(CNNs)excel in single-modality tasks,they struggle to handle multi-modal complexities,lacking the capacity to model global relationships.This research presents a novel approach for examining multi-modal medical imagery using a transformer-based system.The framework employs self-attention and cross-attention mechanisms to synchronize and integrate features across various modalities.Additionally,it shows resilience to variations in noise and image quality,making it adaptable for real-time clinical use.To address the computational hurdles linked to transformer models,particularly in real-time clinical applications in resource-constrained environments,several optimization techniques have been integrated to boost scalability and efficiency.Initially,a streamlined transformer architecture was adopted to minimize the computational load while maintaining model effectiveness.Methods such as model pruning,quantization,and knowledge distillation have been applied to reduce the parameter count and enhance the inference speed.Furthermore,efficient attention mechanisms such as linear or sparse attention were employed to alleviate the substantial memory and processing requirements of traditional self-attention operations.For further deployment optimization,researchers have implemented hardware-aware acceleration strategies,including the use of TensorRT and ONNX-based model compression,to ensure efficient execution on edge devices.These optimizations allow the approach to function effectively in real-time clinical settings,ensuring viability even in environments with limited resources.Future research directions include integrating non-imaging data to facilitate personalized treatment and enhancing computational efficiency for implementation in resource-limited environments.This study highlights the transformative potential of transformer models in multi-modal medical imaging,offering improvements in diagnostic accuracy and patient care outcomes.
基金funding from the Princess Nourah bint Abdulrahman University Researchers Supporting Project(PNURSP2025R195)the University of Bisha through its Fast-Track Research Support Program.
文摘Artificial intelligence(AI),particularly deep learning algorithms utilizing convolutional neural networks,plays an increasingly pivotal role in enhancing medical image examination.It demonstrates the potential for improving diagnostic accuracy within dental care.Orthopantomograms(OPGs)are essential in dentistry;however,their manual interpretation is often inconsistent and tedious.To the best of our knowledge,this is the first comprehensive application of YOLOv5m for the simultaneous detection and classification of six distinct dental pathologies using panoramic OPG images.The model was trained and refined on a custom dataset that began with 232 panoramic radiographs and was later expanded to 604 samples.These included annotated subclasses representing Caries,Infection,Impacted Teeth,Fractured Teeth,Broken Crowns,and Healthy conditions.The training was performed using GPU resources alongside tuned hyperparameters of batch size,learning rate schedule,and early stopping tailored for generalization to prevent overfitting.Evaluation on a held-out test set showed strong performance in the detection and localization of various dental pathologies and robust overall accuracy.At an IoU of 0.5,the system obtained a mean precision of 94.22%and recall of 90.42%,with mAP being 93.71%.This research confirms the use of YOLOv5m as a robust,highly efficient AI technology for the analysis of dental pathologies using OPGs,providing a clinically useful solution to enhance workflow efficiency and aid in sustaining consistency in complex multi-dimensional case evaluations.
基金Supported by Grant-in-Aid for Research on Hepatitis from the Japan Agency for Medical Research and Development,No.24fk0210128h0002Grant-in-Aid for Scientific Research,No.KAKENHI-23K07372.
文摘BACKGROUND Congestive hepatopathy,also known as nutmeg liver,is liver damage secondary to chronic heart failure(HF).Its morphological characteristics in terms of medical imaging are not defined and remain unclear.AIM To leverage machine learning to capture imaging features of congestive hepatopathy using incidentally acquired computed tomography(CT)scans.METHODS We retrospectively analyzed 179 chronic HF patients who underwent echocardiography and CT within one year.Right HF severity was classified into three grades.Liver CT images at the paraumbilical vein level were used to develop a ResNet-based machine learning model to predict tricuspid regurgitation(TR)severity.Model accuracy was compared with that of six gastroenterology and four radiology experts.RESULTS In the included patients,120 were male(mean age:73.1±14.4 years).The accuracy of the results predicting TR severity from a single CT image for the machine learning model was significantly higher than the average accuracy of the experts.The model was found to be exceptionally reliable for predicting severe TR.CONCLUSION Deep learning models,particularly those using ResNet architectures,can help identify morphological changes associated with TR severity,aiding in early liver dysfunction detection in patients with HF,thereby improving outcomes.
基金the Deanship of Graduate Studies and Scientific Research at Najran University,Saudi Arabia,for their financial support through the Easy Track Research program,grant code(NU/EFP/MRC/13).
文摘Background:Accurate classification of normal blood cells is a critical foundation for automated hematological analysis,including the detection of pathological conditions like leukemia.While convolutional neural networks(CNNs)excel in local feature extraction,their ability to capture global contextual relationships in complex cellular morphologies is limited.This study introduces a hybrid CNN-Transformer framework to enhance normal blood cell classification,laying the groundwork for future leukemia diagnostics.Methods:The proposed architecture integrates pre-trained CNNs(ResNet50,EfficientNetB3,InceptionV3,CustomCNN)with Vision Transformer(ViT)layers to combine local and global feature modeling.Four hybrid models were evaluated on the publicly available Blood Cell Images dataset from Kaggle,comprising 17,092 annotated normal blood cell images across eight classes.The models were trained using transfer learning,fine-tuning,and computational optimizations,including cross-model parameter sharing to reduce redundancy by reusing weights across CNN backbones and attention-guided layer pruning to eliminate low-contribution layers based on attention scores,improving efficiency without sacrificing accuracy.Results:The InceptionV3-ViT model achieved a weighted accuracy of 97.66%(accounting for class imbalance by weighting each class’s contribution),a macro F1-score of 0.98,and a ROC-AUC of 0.998.The framework excelled in distinguishing morphologically similar cell types demonstrating robustness and reliable calibration(ECE of 0.019).The framework addresses generalization challenges,including class imbalance and morphological similarities,ensuring robust performance across diverse cell types.Conclusion:The hybrid CNN-Transformer framework significantly improves normal blood cell classification by capturing multi-scale features and long-range dependencies.Its high accuracy,efficiency,and generalization position it as a strong baseline for automated hematological analysis,with potential for extension to leukemia subtype classification through future validation on pathological samples.
基金supported by the Japan Society for the Promotion of Science(No.19H05611).
文摘This study explores the relationship between cardiac activity and biochemical indicators in Pacific oysters(Crassostrea gigas)during cold storage to develop a nondestructive vitality assessment method.Oysters were stored at−1℃ for 14 d,with cardiac patterns tracked via image analysis,and biochemical markers(pH,adenosine triphosphate(ATP)-related compounds,and adenylate energy charge(AEC))were assessed.Five cardiac patterns were identified,with regular alternating contractions common early but decreasing over time,aligning with declines in AEC(44.11%-35.52%)and pH(6.98-6.55).The intervals between ventricular and atrial contractions rose from 4.2 to 5.6 s,strongly correlating with biochemical signs of vitality loss.Image analysis revealed characteristic waveforms for each cardiac pattern,despite amplitude variations caused by optical artifacts.These findings indicate that cardiac pattern analysis via image processing could be an effective nondestructive indicator of oyster vitality,offering a novel approach to quality control in shellfish storage and distribution.
文摘Medical image analysis has become a cornerstone of modern healthcare,driven by the exponential growth of data from imaging modalities such as MRI,CT,PET,ultrasound,and X-ray.Traditional machine learning methods have made early contributions;however,recent advancements in deep learning(DL)have revolutionized the field,offering state-of-the-art performance in image classification,segmentation,detection,fusion,registration,and enhancement.This comprehensive review presents an in-depth analysis of deep learning methodologies applied across medical image analysis tasks,highlighting both foundational models and recent innovations.The article begins by introducing conventional techniques and their limitations,setting the stage for DL-based solutions.Core DL architectures,including Convolutional Neural Networks(CNNs),Recurrent Neural Networks(RNNs),Generative Adversarial Networks(GANs),Vision Transformers(ViTs),and hybrid models,are discussed in detail,including their advantages and domain-specific adaptations.Advanced learning paradigms such as semi-supervised learning,selfsupervised learning,and few-shot learning are explored for their potential to mitigate data annotation challenges in clinical datasets.This review further categorizes major tasks in medical image analysis,elaborating on how DL techniques have enabled precise tumor segmentation,lesion detection,modality fusion,super-resolution,and robust classification across diverse clinical settings.Emphasis is placed on applications in oncology,cardiology,neurology,and infectious diseases,including COVID-19.Challenges such as data scarcity,label imbalance,model generalizability,interpretability,and integration into clinical workflows are critically examined.Ethical considerations,explainable AI(XAI),federated learning,and regulatory compliance are discussed as essential components of real-world deployment.Benchmark datasets,evaluation metrics,and comparative performance analyses are presented to support future research.The article concludes with a forward-looking perspective on the role of foundation models,multimodal learning,edge AI,and bio-inspired computing in the future of medical imaging.Overall,this review serves as a valuable resource for researchers,clinicians,and developers aiming to harness deep learning for intelligent,efficient,and clinically viable medical image analysis.
文摘Given the importance of sentiment analysis in diverse environments,various methods are used for image sentiment analysis,including contextual sentiment analysis that utilizes character and scene relationships.However,most existing works employ character faces in conjunction with context,yet lack the capacity to analyze the emotions of characters in unconstrained environments,such as when their faces are obscured or blurred.Accordingly,this article presents the Adaptive Multi-Channel Sentiment Analysis Network(AMSA),a contextual image sentiment analysis framework,which consists of three channels:body,face,and context.AMSA employs Multi-task Cascaded Convolutional Networks(MTCNN)to detect faces within body frames;if detected,facial features are extracted and fused with body and context information for emotion recognition.If not,the model leverages body and context features alone.Meanwhile,to address class imbalance in the EMOTIC dataset,Focal Loss is introduced to improve classification performance,especially for minority emotion categories.Experimental results have shown that certain sentiment categories with lower representation in the dataset demonstrate leading classification accuracy,the AMSA yields a 2.53%increase compared with state-of-the-art methods.
基金Scientic and Technological Research Council of Turkey(TUBITAK),under grand,No:113E771.
文摘Subcutaneous vein network plays important roles to maintain microcirculation that is related to some diagnostic aspects.Despite developments of optical imaging technologies,still the difficulties about deep skin vascular imaging have been continued.On the other hand,since hemoglobin con-centration of human blood has key role in the veins imaging by optical manner,the used wavelength in vascular imaging,must be chosen considering absorption of hemoglobin.In this research,we constructed a near infrared(NIR)light source because of lower absorption of hemoglobin in this optical region.To obtain vascular image,reflectance geometry was used.Next,from recorded images,vascular network analysis,such as calculation of width of vascular of interest and complexity of selected region were implemented.By comparing with other modalities,we observed that proposed imaging system has great advantages including nonionized radiation,moderate penetration depth of 0.5-3 mm and diameter of 1 mm,cost-effective and algorit hmic simplicity for analysis.
基金Shenzhen Science and Technology Program,Grant/Award Number:ZDSYS20211021111415025Shenzhen Institute of Artificial Intelligence and Robotics for SocietyYouth Science and Technology Talents Development Project of Guizhou Education Department,Grant/Award Number:QianJiaoheKYZi[2018]459。
文摘Facial beauty analysis is an important topic in human society.It may be used as a guidance for face beautification applications such as cosmetic surgery.Deep neural networks(DNNs)have recently been adopted for facial beauty analysis and have achieved remarkable performance.However,most existing DNN-based models regard facial beauty analysis as a normal classification task.They ignore important prior knowledge in traditional machine learning models which illustrate the significant contribution of the geometric features in facial beauty analysis.To be specific,landmarks of the whole face and facial organs are introduced to extract geometric features to make the decision.Inspired by this,we introduce a novel dual-branch network for facial beauty analysis:one branch takes the Swin Transformer as the backbone to model the full face and global patterns,and another branch focuses on the masked facial organs with the residual network to model the local patterns of certain facial parts.Additionally,the designed multi-scale feature fusion module can further facilitate our network to learn complementary semantic information between the two branches.In model optimisation,we propose a hybrid loss function,where especially geometric regulation is introduced by regressing the facial landmarks and it can force the extracted features to convey facial geometric features.Experiments performed on the SCUT-FBP5500 dataset and the SCUT-FBP dataset demonstrate that our model outperforms the state-of-the-art convolutional neural networks models,which proves the effectiveness of the proposed geometric regularisation and dual-branch structure with the hybrid network.To the best of our knowledge,this is the first study to introduce a Vision Transformer into the facial beauty analysis task.
文摘The Ki67 index (KI) is a standard clinical marker for tumor proliferation;however, its application is hindered by intratumoral heterogeneity. In this study, we used digital image analysis to comprehensively analyze Ki67 heterogeneity and distribution patterns in breast carcinoma. Using Smart Pathology software, we digitized and analyzed 42 excised breast carcinoma Ki67 slides. Boxplots, histograms, and heat maps were generated to illustrate the KI distribution. We found that 30% of cases (13/42) exhibited discrepancies between global and hotspot KI when using a 14% KI threshold for classification. Patients with higher global or hotspot KI values displayed greater heterogenicity. Ki67 distribution patterns were categorized as randomly distributed (52%, 22/42), peripheral (43%, 18/42), and centered (5%, 2/42). Our sampling simulator indicated analyzing more than 10 high-power fields was typically required to accurately estimate global KI, with sampling size being correlated with heterogeneity. In conclusion, using digital image analysis in whole-slide images allows for comprehensive Ki67 profile assessment, shedding light on heterogeneity and distribution patterns. This spatial information can facilitate KI surveys of breast cancer and other malignancies.
基金supported by a grant from the Standardization and Integration of Resources Information for Seed-Cluster in Hub-Spoke Material Bank Program(Project No.PJ01587004),Rural Development Administration,Republic of Korea.
文摘Recognizing the variation of genetic resources is the first step in selection.One of the most important variations in grain crops is the uniformity of seed grain weight,which can be converted into seed size.However,it has been challenging since it needs high labor costs and time to measure it on a large scale.The current study used an image analysis technique to measure the grain seed area of about 100 seeds per accession with 64 germplasm of Tartary buckwheat(Fagopyrum tataricum)to study variation among and within them.To understand the nature of variation,skewness and kurtosis analysis of probability density function curve for seed area were used.As a result,a large variation among and within accessions was found.This means that the seed sizes within an accession are not uniform in this given cleistogamous species due to its non-uniform flowering time.This implies that the seed size should be considered an important factor for the germplasm enhancement program.
基金Projects(50934002,51074013,51304076,51104100)supported by the National Natural Science Foundation of ChinaProject(IRT0950)supported by the Program for Changjiang Scholars Innovative Research Team in Universities,ChinaProject(2012M510007)supported by China Postdoctoral Science Foundation
文摘Methods and procedures of three-dimensional （3D） characterization of the pore structure features in the packed ore particle bed are focused. X-ray computed tomography was applied to deriving the cross-sectional images of specimens with single particle size of 1-2, 2-3, 3-4, 4-5, 5-6, 6-7, 7-8, 8-9, 9-10 ram. Based on the in-house developed 3D image analysis programs using Matlab, the volume porosity, pore size distribution and degree of connectivity were calculated and analyzed in detail. The results indicate that the volume porosity, the mean diameter of pores and the effective pore size （d50） increase with the increasing of particle size. Lognormal distribution or Gauss distribution is mostly suitable to model the pore size distribution. The degree of connectivity investigated on the basis of cluster-labeling algorithm also increases with increasing the particle size approximately.
基金Supported by the NationalNaturalScience Foundation of China( No. 2 0 175 0 1) and U niversity Key Teachers Programdirected under the Ministry of Education ofP.R.China( No. 2 0 0 0 - 6 5 )
文摘It is critical to establish a direct and precise method with a high sensitivity and selectivity in analytical chemistry. In this research, making use of a well known phenomenon of capillary flow, we have proposed an image analysis method of nucleic acids at the price of a small amount of sample. When a droplet of the supramolecular complex solution, formed by neutral red and nucleic acids(NA) under an approximate neutral condition, was placed on the hydrophobic surface of dimethyl dichlorosilane pretreated glass slides, and it was evaporated, the supramolecular complex exhibited the periphery of the droplet due to the capillary effect, and accumulated there to form a red capillary flow directed assembly ring(CFDAR). A typical CFDAR has an outer diameter of (2 r ) about 1.18 mm and a ring width(2 δ ) of about 41 μm. Depending on the experimental conditions, a variety of CFDAR can be assembled. The experimental results are in agreement with our former theoretical discussion. It was found that when a droplet volume is 0.1 μL, the fluorescence intensity of the CFDAR formed by the NR NA is in proportion to the content of calf thymus DNA in the range of 0-0.28 ng, fish sperm DNA of 0-0.24 ng and yeast RNA of 0-0.16 ng with the limit of detection(3 σ ) of 1 7, 1.4 and 0.9 pg, respectively for the three nucleic acids.
文摘The integration of image analysis through deep learning(DL)into rock classification represents a significant leap forward in geological research.While traditional methods remain invaluable for their expertise and historical context,DL offers a powerful complement by enhancing the speed,objectivity,and precision of the classification process.This research explores the significance of image data augmentation techniques in optimizing the performance of convolutional neural networks(CNNs)for geological image analysis,particularly in the classification of igneous,metamorphic,and sedimentary rock types from rock thin section(RTS)images.This study primarily focuses on classic image augmentation techniques and evaluates their impact on model accuracy and precision.Results demonstrate that augmentation techniques like Equalize significantly enhance the model's classification capabilities,achieving an F1-Score of 0.9869 for igneous rocks,0.9884 for metamorphic rocks,and 0.9929 for sedimentary rocks,representing improvements compared to the baseline original results.Moreover,the weighted average F1-Score across all classes and techniques is 0.9886,indicating an enhancement.Conversely,methods like Distort lead to decreased accuracy and F1-Score,with an F1-Score of 0.949 for igneous rocks,0.954 for metamorphic rocks,and 0.9416 for sedimentary rocks,exacerbating the performance compared to the baseline.The study underscores the practicality of image data augmentation in geological image classification and advocates for the adoption of DL methods in this domain for automation and improved results.The findings of this study can benefit various fields,including remote sensing,mineral exploration,and environmental monitoring,by enhancing the accuracy of geological image analysis both for scientific research and industrial applications.
基金financially supported by the National Natural Science Foundation of China(Nos.U2241231 and 52071206).
文摘In traditional trial-and-error method,enhancing the Young's modulus of magnesium alloys while maintaining a favorable ductility has consistently been a challenge.It is a need to explore more efficient and expedited methods to design magnesium alloys with high modulus and ductility.In this study,machine learning(ML)and assisted microstructure control methods are used to design high modulus magnesium alloys.Six key features that influence stiffness and ductility have been extracted in this ML model based on abundant data from literature sources.As a result,predictive models for Young's modulus and elongation are established,with errors<2.4%and 4.5%through XGBoost machine learning model,respectively.Within the given range of six features,the magnesium alloys can be fabricated with the Young's modulus exceeding 50 GPa and an elongation surpassing 6%.As a validation,Mg-Al-Y alloys were experimentally prepared to meet the criteria of six features,achieving Young's modulus of 51.5 GPa,and the elongation of 7%.Moreover,the SHapley Additive exPlanation(SHAP)is introduced to boost the model interpretability.This indicates that balancing the volume fraction of reinforcement,the most important feature,is key to achieve Mg-Al-Y alloys with high Young's modulus and favorable elongation through the two models.Enhancing reinforcement dispersion and reducing the size of reinforcement and grain can further improve the elongation of high-stiffness Mg alloy.
