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
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
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
Optimizing root system architecture(RSA)is essential for plants because of its critical role in acquiring water and nutrients from the soil.However,the subterranean nature of roots complicates the measurement of RSA t...Optimizing root system architecture(RSA)is essential for plants because of its critical role in acquiring water and nutrients from the soil.However,the subterranean nature of roots complicates the measurement of RSA traits.Recently developed rhizobox methods allow for the rapid acquisition of root images.Nevertheless,effective and precise approaches for extracting RSA features from these images remain underdeveloped.Deep learning(DL)technology can enhance image segmentation and facilitate RSA trait extraction.However,comprehensive pipelines that integrate DL technologies into image-based root phenotyping techniques are still scarce,hampering their implementation.To address this challenge,we present a reproducible pipeline(faCRSA)for automated RSA traits analysis,consisting of three modules:(1)the RSA traits extraction module functions to segment soil-root images and calculate RSA traits.A lightweight convolutional neural network(CNN)named RootSeg was proposed for efficient and accurate segmentation;(2)the data storage module,which stores image and text data from other modules;and(3)the web application module,which allows researchers to analyze data online in a user-friendly manner.The correlation coefficients(R^(2))of total root length,root surface area,and root volume calculated from faCRSA and manually measured results were 0.96**,0.97**,and 0.93**,respectively,with root mean square errors(RMSE)of 8.13 cm,1.68 cm^(2),and 0.05 cm^(3),processed at a rate of 9.74 s per image,indicating satisfying accuracy.faCRSA has also demonstrated satisfactory performance in dynamically monitoring root system changes under various stress conditions,such as drought or waterlogging.The detailed code and deployable package of faCRSA are provided for researchers with the potential to replace manual and semi-automated methods.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
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
Cerebral small vessel disease encompasses a group of neurological disorders characterized by injury to small blood vessels,often leading to stroke and dementia.Due to its diverse etiologies and complex pathological me...Cerebral small vessel disease encompasses a group of neurological disorders characterized by injury to small blood vessels,often leading to stroke and dementia.Due to its diverse etiologies and complex pathological mechanisms,preventing and treating cerebral small vessel vasculopathy is challenging.Recent studies have shown that the glymphatic system plays a crucial role in interstitial solute clearance and the maintenance of brain homeostasis.Increasing evidence also suggests that dysfunction in glymphatic clearance is a key factor in the progression of cerebral small vessel disease.This review begins with a comprehensive introduction to the structure,function,and driving factors of the glymphatic system,highlighting its essential role in brain waste clearance.Afterwards,cerebral small vessel disease was reviewed from the perspective of the glymphatic system,after which the mechanisms underlying their correlation were summarized.Glymphatic dysfunction may lead to the accumulation of metabolic waste in the brain,thereby exacerbating the pathological processes associated with cerebral small vessel disease.The review also discussed the direct evidence of glymphatic dysfunction in patients and animal models exhibiting two subtypes of cerebral small vessel disease:arteriolosclerosis-related cerebral small vessel disease and amyloid-related cerebral small vessel disease.Diffusion tensor image analysis along the perivascular space is an important non-invasive tool for assessing the clearance function of the glymphatic system.However,the effectiveness of its parameters needs to be enhanced.Among various nervous system diseases,including cerebral small vessel disease,glymphatic failure may be a common final pathway toward dementia.Overall,this review summarizes prevention and treatment strategies that target glymphatic drainage and will offer valuable insight for developing novel treatments for cerebral small vessel disease.展开更多
Post-kidney transplant rejection is a critical factor influencing transplant success rates and the survival of transplanted organs.With the rapid advancement of artificial intelligence technologies,machine learning(ML...Post-kidney transplant rejection is a critical factor influencing transplant success rates and the survival of transplanted organs.With the rapid advancement of artificial intelligence technologies,machine learning(ML)has emerged as a powerful data analysis tool,widely applied in the prediction,diagnosis,and mechanistic study of kidney transplant rejection.This mini-review systematically summarizes the recent applications of ML techniques in post-kidney transplant rejection,covering areas such as the construction of predictive models,identification of biomarkers,analysis of pathological images,assessment of immune cell infiltration,and formulation of personalized treatment strategies.By integrating multi-omics data and clinical information,ML has significantly enhanced the accuracy of early rejection diagnosis and the capability for prognostic evaluation,driving the development of precision medicine in the field of kidney transplantation.Furthermore,this article discusses the challenges faced in existing research and potential future directions,providing a theoretical basis and technical references for related studies.展开更多
By using computer imaging analysis system combined with light microscopy, the glomeruli lesions on biopsy specimens sections were quantitatively analysed. The patholoical changes of mesangial proliferative glomerulone...By using computer imaging analysis system combined with light microscopy, the glomeruli lesions on biopsy specimens sections were quantitatively analysed. The patholoical changes of mesangial proliferative glomerulonephritis (MsPGN) in children were quantitatively evaluated and the correlation between the mesangial matrix area (MA) and some clinical data were also analysed. The results indicated that the levels of various glomerular parameters in MsPGN group were higher than those of normal controls. No correlation was found between MA and 24 h urinary protein excretion, but a negative correlation was revealed between MA and estimated GFR. MA was also correlated with the duration of MsPGN recovery. It was suggested that the quantitative analysis of glomerular parameters by computer is a reproducible methods. The parameter of MA may be used for evaluation of the renal function, determination of the duration of therapy and evaluation of prognosis of MsPGN in children.展开更多
The fractal dimensions in different topological spaces of polyferric chloride-humic acid （PFC-HA） flocs, formed in flocculating different kinds of humic acids （HA） water at different initial pH （9.0, 7.0, 5.0） a...The fractal dimensions in different topological spaces of polyferric chloride-humic acid （PFC-HA） flocs, formed in flocculating different kinds of humic acids （HA） water at different initial pH （9.0, 7.0, 5.0） and PFC dosages, were calculated by effective densitymaximum diameter, image analysis, and N2 absorption-desorption methods, respectively. The mass fractal dimensions （De） of PFC-HA floes were calculated by bi-logarithm relation of effective density with maximum diameter and Logan empirical equation. The Df value was more than 2.0 at initial pH of 7,0, which was 11% and 13% higher than those at pH 9.0 and 5.0, respecively, indicating the most compact flocs formed in flocculated HA water at initial pH of 7.0. The image analysis for those flocs indicates that after flocculating the HA water at initial pH greater than 7.0 with PFC flocculant, the fractal dimensions of D2 （logA vs. logdL） and D3 （logVsphere vs. logdL） of PFC-HA floes decreased with the increase of PFC dosages, and PFC-HA floes showed a gradually looser structure. At the optimum dosage of PFC, the D2 （logA vs. logdL） values of the flocs show 14%-43% difference with their corresponding Dr, and they even had different tendency with the change of initial pH values. However, the D2 values of the floes formed at three different initial pH in HA solution had a same tendency with the corresponding Df. Based on fractal Frenkel-Halsey-HiU （FHH） adsorption and desorption equations, the pore surface fractal dimensions （Ds） for dried powders of PFC-HA flocs formed in HA water with initial pH 9.0 and 7.0 were all close to 2.9421, and the Ds values of flocs formed at initial pH 5.0 were less than 2.3746. It indicated that the pore surface fractal dimensions of PFC-HA floes dried powder mainly show the irregularity from the mesopore-size distribution and marcopore-size distribution.展开更多
Mineral dissemination and pore space distribution in ore particles are important features that influence heap leaching performance.To quantify the mineral dissemination and pore space distribution of an ore particle,a...Mineral dissemination and pore space distribution in ore particles are important features that influence heap leaching performance.To quantify the mineral dissemination and pore space distribution of an ore particle,a cylindrical copper oxide ore sample(I center dot 4.6 mm x 5.6 mm)was scanned using high-resolution X-ray computed tomography(HRXCT),a nondestructive imaging technology,at a spatial resolution of 4.85 mu m.Combined with three-dimensional(3D)image analysis techniques,the main mineral phases and pore space were segmented and the volume fraction of each phase was calculated.In addition,the mass fraction of each mineral phase was estimated and the result was validated with that obtained using traditional techniques.Furthermore,the pore phase features,including the pore size distribution,pore surface area,pore fractal dimension,pore centerline,and the pore connectivity,were investigated quantitatively.The pore space analysis results indicate that the pore size distribution closely fits a log-normal distribution and that the pore space morphology is complicated,with a large surface area and low connectivity.This study demonstrates that the combination of HRXCT and 3D image analysis is an effective tool for acquiring 3D mineralogical and pore structural data.展开更多
With the wide use of high-resolution remotely sensed imagery, the object-oriented remotely sensed informa- tion classification pattern has been intensively studied. Starting with the definition of object-oriented remo...With the wide use of high-resolution remotely sensed imagery, the object-oriented remotely sensed informa- tion classification pattern has been intensively studied. Starting with the definition of object-oriented remotely sensed information classification pattern and a literature review of related research progress, this paper sums up 4 developing phases of object-oriented classification pattern during the past 20 years. Then, we discuss the three aspects of method- ology in detail, namely remotely sensed imagery segmentation, feature analysis and feature selection, and classification rule generation, through comparing them with remotely sensed information classification method based on per-pixel. At last, this paper presents several points that need to be paid attention to in the future studies on object-oriented RS in- formation classification pattern: 1) developing robust and highly effective image segmentation algorithm for multi-spectral RS imagery; 2) improving the feature-set including edge, spatial-adjacent and temporal characteristics; 3) discussing the classification rule generation classifier based on the decision tree; 4) presenting evaluation methods for classification result by object-oriented classification pattern.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 projects of the National Key Research and Development Program of China(2024YFD2301305)Jiangsu Innovation Support Program for International Science and Technology Cooperation Project(BZ2023049)+2 种基金the projects of the National Natural Science Foundation of China(32272213)the China Agriculture Research System(CARS-03)Jiangsu Collaborative Innovation Center for Modern Crop Production(JCIC-MCP).
文摘Optimizing root system architecture(RSA)is essential for plants because of its critical role in acquiring water and nutrients from the soil.However,the subterranean nature of roots complicates the measurement of RSA traits.Recently developed rhizobox methods allow for the rapid acquisition of root images.Nevertheless,effective and precise approaches for extracting RSA features from these images remain underdeveloped.Deep learning(DL)technology can enhance image segmentation and facilitate RSA trait extraction.However,comprehensive pipelines that integrate DL technologies into image-based root phenotyping techniques are still scarce,hampering their implementation.To address this challenge,we present a reproducible pipeline(faCRSA)for automated RSA traits analysis,consisting of three modules:(1)the RSA traits extraction module functions to segment soil-root images and calculate RSA traits.A lightweight convolutional neural network(CNN)named RootSeg was proposed for efficient and accurate segmentation;(2)the data storage module,which stores image and text data from other modules;and(3)the web application module,which allows researchers to analyze data online in a user-friendly manner.The correlation coefficients(R^(2))of total root length,root surface area,and root volume calculated from faCRSA and manually measured results were 0.96**,0.97**,and 0.93**,respectively,with root mean square errors(RMSE)of 8.13 cm,1.68 cm^(2),and 0.05 cm^(3),processed at a rate of 9.74 s per image,indicating satisfying accuracy.faCRSA has also demonstrated satisfactory performance in dynamically monitoring root system changes under various stress conditions,such as drought or waterlogging.The detailed code and deployable package of faCRSA are provided for researchers with the potential to replace manual and semi-automated methods.
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.
基金supported by the National Natural Science Foundation of China,No.82274304(to YH)the Major Clinical Study Projects of Shanghai Shenkang Hospital Development Center,No.SHDC2020CR2046B(to YH)Shanghai Municipal Health Commission Talent Plan,No.2022LJ010(to YH).
文摘Cerebral small vessel disease encompasses a group of neurological disorders characterized by injury to small blood vessels,often leading to stroke and dementia.Due to its diverse etiologies and complex pathological mechanisms,preventing and treating cerebral small vessel vasculopathy is challenging.Recent studies have shown that the glymphatic system plays a crucial role in interstitial solute clearance and the maintenance of brain homeostasis.Increasing evidence also suggests that dysfunction in glymphatic clearance is a key factor in the progression of cerebral small vessel disease.This review begins with a comprehensive introduction to the structure,function,and driving factors of the glymphatic system,highlighting its essential role in brain waste clearance.Afterwards,cerebral small vessel disease was reviewed from the perspective of the glymphatic system,after which the mechanisms underlying their correlation were summarized.Glymphatic dysfunction may lead to the accumulation of metabolic waste in the brain,thereby exacerbating the pathological processes associated with cerebral small vessel disease.The review also discussed the direct evidence of glymphatic dysfunction in patients and animal models exhibiting two subtypes of cerebral small vessel disease:arteriolosclerosis-related cerebral small vessel disease and amyloid-related cerebral small vessel disease.Diffusion tensor image analysis along the perivascular space is an important non-invasive tool for assessing the clearance function of the glymphatic system.However,the effectiveness of its parameters needs to be enhanced.Among various nervous system diseases,including cerebral small vessel disease,glymphatic failure may be a common final pathway toward dementia.Overall,this review summarizes prevention and treatment strategies that target glymphatic drainage and will offer valuable insight for developing novel treatments for cerebral small vessel disease.
文摘Post-kidney transplant rejection is a critical factor influencing transplant success rates and the survival of transplanted organs.With the rapid advancement of artificial intelligence technologies,machine learning(ML)has emerged as a powerful data analysis tool,widely applied in the prediction,diagnosis,and mechanistic study of kidney transplant rejection.This mini-review systematically summarizes the recent applications of ML techniques in post-kidney transplant rejection,covering areas such as the construction of predictive models,identification of biomarkers,analysis of pathological images,assessment of immune cell infiltration,and formulation of personalized treatment strategies.By integrating multi-omics data and clinical information,ML has significantly enhanced the accuracy of early rejection diagnosis and the capability for prognostic evaluation,driving the development of precision medicine in the field of kidney transplantation.Furthermore,this article discusses the challenges faced in existing research and potential future directions,providing a theoretical basis and technical references for related studies.
文摘By using computer imaging analysis system combined with light microscopy, the glomeruli lesions on biopsy specimens sections were quantitatively analysed. The patholoical changes of mesangial proliferative glomerulonephritis (MsPGN) in children were quantitatively evaluated and the correlation between the mesangial matrix area (MA) and some clinical data were also analysed. The results indicated that the levels of various glomerular parameters in MsPGN group were higher than those of normal controls. No correlation was found between MA and 24 h urinary protein excretion, but a negative correlation was revealed between MA and estimated GFR. MA was also correlated with the duration of MsPGN recovery. It was suggested that the quantitative analysis of glomerular parameters by computer is a reproducible methods. The parameter of MA may be used for evaluation of the renal function, determination of the duration of therapy and evaluation of prognosis of MsPGN in children.
基金supported by the National Natural Science Foundation of China (No. 20407004, 50578012, 50178009)the High-Tech Research and Development Program (863) of China (No. 2007AA06Z301)+2 种基金the Fok Ying Tung Education Foundation of National Education Ministry of China (No. 91078)the Beijing Municipal Commission of Education Project, Program for New Cen- tury Excellent Talents in University (No. NCET-06-0120)the Beijing Nova of Science and Technology, Beijing Key Subject (No. XK100220555).
文摘The fractal dimensions in different topological spaces of polyferric chloride-humic acid （PFC-HA） flocs, formed in flocculating different kinds of humic acids （HA） water at different initial pH （9.0, 7.0, 5.0） and PFC dosages, were calculated by effective densitymaximum diameter, image analysis, and N2 absorption-desorption methods, respectively. The mass fractal dimensions （De） of PFC-HA floes were calculated by bi-logarithm relation of effective density with maximum diameter and Logan empirical equation. The Df value was more than 2.0 at initial pH of 7,0, which was 11% and 13% higher than those at pH 9.0 and 5.0, respecively, indicating the most compact flocs formed in flocculated HA water at initial pH of 7.0. The image analysis for those flocs indicates that after flocculating the HA water at initial pH greater than 7.0 with PFC flocculant, the fractal dimensions of D2 （logA vs. logdL） and D3 （logVsphere vs. logdL） of PFC-HA floes decreased with the increase of PFC dosages, and PFC-HA floes showed a gradually looser structure. At the optimum dosage of PFC, the D2 （logA vs. logdL） values of the flocs show 14%-43% difference with their corresponding Dr, and they even had different tendency with the change of initial pH values. However, the D2 values of the floes formed at three different initial pH in HA solution had a same tendency with the corresponding Df. Based on fractal Frenkel-Halsey-HiU （FHH） adsorption and desorption equations, the pore surface fractal dimensions （Ds） for dried powders of PFC-HA flocs formed in HA water with initial pH 9.0 and 7.0 were all close to 2.9421, and the Ds values of flocs formed at initial pH 5.0 were less than 2.3746. It indicated that the pore surface fractal dimensions of PFC-HA floes dried powder mainly show the irregularity from the mesopore-size distribution and marcopore-size distribution.
基金financially supported by the National Natural Science Foundation of China(No.51304076)the Natural Science Foundation of Hunan Province,China(No.14JJ4064)
文摘Mineral dissemination and pore space distribution in ore particles are important features that influence heap leaching performance.To quantify the mineral dissemination and pore space distribution of an ore particle,a cylindrical copper oxide ore sample(I center dot 4.6 mm x 5.6 mm)was scanned using high-resolution X-ray computed tomography(HRXCT),a nondestructive imaging technology,at a spatial resolution of 4.85 mu m.Combined with three-dimensional(3D)image analysis techniques,the main mineral phases and pore space were segmented and the volume fraction of each phase was calculated.In addition,the mass fraction of each mineral phase was estimated and the result was validated with that obtained using traditional techniques.Furthermore,the pore phase features,including the pore size distribution,pore surface area,pore fractal dimension,pore centerline,and the pore connectivity,were investigated quantitatively.The pore space analysis results indicate that the pore size distribution closely fits a log-normal distribution and that the pore space morphology is complicated,with a large surface area and low connectivity.This study demonstrates that the combination of HRXCT and 3D image analysis is an effective tool for acquiring 3D mineralogical and pore structural data.
基金Under the auspices of the National Natural Science Foundation of China (No. 40301038), Talents Recruitment Foun-dation of Nanjing University
文摘With the wide use of high-resolution remotely sensed imagery, the object-oriented remotely sensed informa- tion classification pattern has been intensively studied. Starting with the definition of object-oriented remotely sensed information classification pattern and a literature review of related research progress, this paper sums up 4 developing phases of object-oriented classification pattern during the past 20 years. Then, we discuss the three aspects of method- ology in detail, namely remotely sensed imagery segmentation, feature analysis and feature selection, and classification rule generation, through comparing them with remotely sensed information classification method based on per-pixel. At last, this paper presents several points that need to be paid attention to in the future studies on object-oriented RS in- formation classification pattern: 1) developing robust and highly effective image segmentation algorithm for multi-spectral RS imagery; 2) improving the feature-set including edge, spatial-adjacent and temporal characteristics; 3) discussing the classification rule generation classifier based on the decision tree; 4) presenting evaluation methods for classification result by object-oriented classification pattern.
