Deep Learning-based Framework Research Articles

Predicting microbe–disease association based on graph autoencoder and inductive matrix completion with multi-similarities fusion

BackgroundClinical studies have demonstrated that microbes play a crucial role in human health and disease. The identification of microbe-disease interactions can provide insights into the pathogenesis and promote the diagnosis, treatment, and prevention of disease. Although a large number of computational methods are designed to screen novel microbe-disease associations, the accurate and efficient methods are still lacking due to data inconsistence, underutilization of prior information, and model performance.MethodsIn this study, we proposed an improved deep learning-based framework, named GIMMDA, to identify latent microbe-disease associations, which is based on graph autoencoder and inductive matrix completion. By co-training the information from microbe and disease space, the new representations of microbes and diseases are used to reconstruct microbe-disease association in the end-to-end framework. In particular, a similarity fusion strategy is conducted to improve prediction performance.ResultsThe experimental results show that the performance of GIMMDA is competitive with that of existing state-of-the-art methods on 3 datasets (i.e., HMDAD, Disbiome, and multiMDA). In particular, it performs best with the area under the receiver operating characteristic curve (AUC) of 0.9735, 0.9156, 0.9396 on abovementioned 3 datasets, respectively. And the result also confirms that different similarity fusions can improve the prediction performance. Furthermore, case studies on two diseases, i.e., asthma and obesity, validate the effectiveness and reliability of our proposed model.ConclusionThe proposed GIMMDA model show a strong capability in predicting microbe-disease associations. We expect that GPUDMDA will help identify potential microbe-related diseases in the future.

Deep Learning-Based Obesity Identification System for Young Adults Using Smartphone Inertial Measurements.

Obesity recognition in adolescents is a growing concern. This study presents a deep learning-based obesity identification framework that integrates smartphone inertial measurements with deep learning models to address this issue. Utilizing data from accelerometers, gyroscopes, and rotation vectors collected via a mobile health application, we analyzed gait patterns for obesity indicators. Our framework employs three deep learning models: convolutional neural networks (CNNs), long-short-term memory network (LSTM), and a hybrid CNN-LSTM model. Trained on data from 138 subjects, including both normal and obese individuals, and tested on an additional 35 subjects, the hybrid model achieved the highest accuracy of 97%, followed by the LSTM model at 96.31% and the CNN model at 95.81%. Despite the promising outcomes, the study has limitations, such as a small sample and the exclusion of individuals with distorted gait. In future work, we aim to develop more generalized models that accommodate a broader range of gait patterns, including those with medical conditions.

Online chatter detection via lightweight deep learning framework with efficient signal pre-processing

Deep learning methods have attracted considerable attention in the field of chatter detection due to their exceptional feature extraction and identification capabilities. However, it is challenging to adapt these methods to diverse machining conditions with limited sensor resources. In addition, the computational efficiency of these methods is compromised by their complexity, resulting in a reduced real-time performance for online chatter detection. To address them, this paper proposes a novel deep learning-based framework that combines an efficient signal pre-processing method with a novel lightweight parallel convolutional neural network (LPNN). A fast continuous wavelet transform algorithm is applied to the raw signal, effectively generating time–frequency maps that represent chatter characteristics. The LPNN incorporates the condensed block as a feature extraction unit, which is constructed using depthwise separable convolution. The introduction of the condensed block and the parallel structure increases the depth and width of the designed model while maintaining fewer parameters and higher computational efficiency. A transfer learning strategy is adopted to train the model, reducing the training time and data requirements for online chatter detection. Experimental results using large open-source turning datasets demonstrate the effectiveness of the proposed method in identifying chatter under various machining conditions. The proposed method outperforms other methods with a classification accuracy of 96.5% on the test set. Furthermore, the performance of the proposed method for online chatter detection is validated in our milling experiments, exhibiting enhanced real-time performance compared to other methods.

Prediction of mutation-induced protein stability changes based on the geometric representations learned by a self-supervised method

BackgroundThermostability is a fundamental property of proteins to maintain their biological functions. Predicting protein stability changes upon mutation is important for our understanding protein structure–function relationship, and is also of great interest in protein engineering and pharmaceutical design.ResultsHere we present mutDDG-SSM, a deep learning-based framework that uses the geometric representations encoded in protein structure to predict the mutation-induced protein stability changes. mutDDG-SSM consists of two parts: a graph attention network-based protein structural feature extractor that is trained with a self-supervised learning scheme using large-scale high-resolution protein structures, and an eXtreme Gradient Boosting model-based stability change predictor with an advantage of alleviating overfitting problem. The performance of mutDDG-SSM was tested on several widely-used independent datasets. Then, myoglobin and p53 were used as case studies to illustrate the effectiveness of the model in predicting protein stability changes upon mutations. Our results show that mutDDG-SSM achieved high performance in estimating the effects of mutations on protein stability. In addition, mutDDG-SSM exhibited good unbiasedness, where the prediction accuracy on the inverse mutations is as well as that on the direct mutations.ConclusionMeaningful features can be extracted from our pre-trained model to build downstream tasks and our model may serve as a valuable tool for protein engineering and drug design.

A Point Cloud Graph Neural Network for Protein-Ligand Binding Site Prediction.

Predicting protein-ligand binding sites is an integral part of structural biology and drug design. A comprehensive understanding of these binding sites is essential for advancing drug innovation, elucidating mechanisms of biological function, and exploring the nature of disease. However, accurately identifying protein-ligand binding sites remains a challenging task. To address this, we propose PGpocket, a geometric deep learning-based framework to improve protein-ligand binding site prediction. Initially, the protein surface is converted into a point cloud, and then the geometric and chemical properties of each point are calculated. Subsequently, the point cloud graph is constructed based on the inter-point distances, and the point cloud graph neural network (GNN) is applied to extract and analyze the protein surface information to predict potential binding sites. PGpocket is trained on the scPDB dataset, and its performance is verified on two independent test sets, Coach420 and HOLO4K. The results show that PGpocket achieves a 58% success rate on the Coach420 dataset and a 56% success rate on the HOLO4K dataset. These results surpass competing algorithms, demonstrating PGpocket's advancement and practicality for protein-ligand binding site prediction.

A Deep Learning-Based Framework for Highly Accelerated Prostate MR Dispersion Imaging.

Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) measures microvascular perfusion by capturing the temporal changes of an MRI contrast agent in a target tissue, and it provides valuable information for the diagnosis and prognosis of a wide range of tumors. Quantitative DCE-MRI analysis commonly relies on the nonlinear least square (NLLS) fitting of a pharmacokinetic (PK) model to concentration curves. However, the voxel-wise application of such nonlinear curve fitting is highly time-consuming. The arterial input function (AIF) needs to be utilized in quantitative DCE-MRI analysis. and in practice, a population-based arterial AIF is often used in PK modeling. The contribution of intravascular dispersion to the measured signal enhancement is assumed to be negligible. The MR dispersion imaging (MRDI) model was recently proposed to account for intravascular dispersion, enabling more accurate PK modeling. However, the complexity of the MRDI hinders its practical usability and makes quantitative PK modeling even more time-consuming. In this paper, we propose fast MR dispersion imaging (fMRDI) to effectively represent the intravascular dispersion and highly accelerated PK parameter estimation. We also propose a deep learning-based, two-stage framework to accelerate PK parameter estimation. We used a deep neural network (NN) to estimate PK parameters directly from enhancement curves. The estimation from NN was further refined using several steps of NLLS, which is significantly faster than performing NLLS from random initializations. A data synthesis module is proposed to generate synthetic training data for the NN. Two data-processing modules were introduced to improve the model's stability against noise and variations. Experiments on our in-house clinical prostate MRI dataset demonstrated that our method significantly reduces the processing time, produces a better distinction between normal and clinically significant prostate cancer (csPCa) lesions, and is more robust against noise than conventional DCE-MRI analysis methods.

Human sleeping pose estimation from IR images for in-bed patient monitoring using image processing and deep learning techniques

Human Pose Estimation (HPE) is a crucial step towards understanding people in images and videos. HPE provides geometric and motion information of the human body, which has been applied to a wide range of applications (e.g., human-computer interaction, motion analysis, augmented reality, virtual reality, healthcare, etc.). An extremely useful task of this kind is the 2D pose estimation of bedridden patients from infrared (IR) images. Here, the IR imaging modality is preferred due to privacy concerns and the need for monitoring both uncovered and covered patients at different levels of illumination. The major drawback of this research problem is the unavailability of covered examples, which are very costly to collect and time-consuming to label. In this work, a deep learning-based framework was developed for human sleeping pose estimation on covered images using only the uncovered training images. In the training scheme, two different image augmentation techniques, a statistical approach as well as a GAN-based approach, were explored for domain adaptation, where the statistical approach performed better. The accuracy of the model trained on the statistically augmented dataset was improved by 124 % as compared with the model trained on non-augmented images. To handle the scarcity of training infrared images, a transfer learning strategy was used by pre-training the model on an RGB pose estimation dataset, resulting in a further increment in accuracy of 4 %. Semi-supervised learning techniques, with a novel pose discriminator model in the loop, were adopted to utilize the unannotated training data, resulting in a further 3 % increase in accuracy. Thus, significant improvement has been shown in the case of 2D pose estimation from infrared images, with a comparatively small amount of annotated data and a large amount of unannotated data by using the proposed training pipeline powered by heavy augmentation.

A deep learning classification framework for research methods of marine protected area management

The latest emerging transdisciplinary marine protected area (MPA) research scheme requires efficient approaches for theoretically based and data-driven method integration. However, due to the rapid development and diversification of research methods, it is growingly difficult to locate new methods in methodological dimensions and integrate them to the utmost utility. This study proposes a deep learning-based classification framework for MPA management methods focused particularly on data and theory capabilities using natural language processing (NLP). It extracted keywords from academic sources and performed clustering based on semantic similarity, generating benchmark texts for abstract labeling. By training the deep learning NLP model and analyzing the abstracts of 9049 MPA management empirical research articles from 1986 to 2024, the data and theory scores were attributed to each article, and a total of 19 major method categories and 110 segment branches were identified in qualitative, quantitative, and mixed genres. Combination types of research methods were summarized, yielding the data-theory neutralization principle where the average data and theory scores tend to approximate 0.50. Applying the principle broadens traditional boundaries for method integration and extends method synthesis to higher numbers, generating a practical research 2paradigm for future MPA research. Implications include bridging social and ecological data, theorizing emergent challenges in complex systems and integrating theory construction and data science. The framework is applicable to quantification of other environmental management disciplines and can serve as guidance for multidisciplinary method integration.© 2017 Elsevier Inc. All rights reserved.

A deep learning based experimental framework for automatic staging of pressure ulcers from thermal images

ABSTRACT Pressure ulcers can be life-threatening, and accurate staging is crucial for proper treatment. Manual methods for ulcer detection and staging are subjective and prone to errors. Since the wound temperature holds valuable and informative data, we propose thermal imaging as a tool to detect the stages of pressure ulcers automatically. This paper implements a deep learning-based framework for the automatic detection and classification of pressure ulcer stages 2–4 using thermal images. 450 thermal images were used to train the classification model. The framework combines ‘R50-FPN’, retrained using transfer learning, for ulcer detection and cropping. An ensemble model is used for ulcer stage classification. The model was trained on a dataset consisting of 68 original images and their augmentations. The proposed method achieves a mAP for object detection of 0.70 and an ulcer stage classification accuracy of 0.9545 during evaluation. Class activation maps were also employed to study the model performance further, indicating that the trained model focuses on the wound periphery and the wound bed for classification. The results show a significant improvement in classification accuracy when compared with related works demonstrating the potential of deep learning-based thermal imaging as an objective method for the automated staging of pressure ulcers.

Accelerating the discovery of acceptor materials for organic solar cells by deep learning

It is a time-consuming and costly process to develop affordable and high-performance organic photovoltaic materials. Computational methods are essential for accelerating the material discovery process by predicting power conversion efficiencies (PCE). In this study, we propose a deep learning-based framework (DeepAcceptor) to design and discover highly efficient small molecule acceptor materials. Specifically, an experimental dataset is constructed by collecting acceptor data from publications. Then, a deep learning-based model is customized to predict PCEs by applying graph representation learning to Bidirectional Encoder Representations from Transformers (BERT), with the atom, bond, and connection information in acceptor molecular structures as the input (abcBERT). The computational dataset derived from density functional theory (DFT) calculations and the experimental dataset from literature are used to pre-train and fine-tune the model, respectively. The abcBERT model outperforms other state-of-the-art models for the PCE prediction with MAE = 1.78 and R2 = 0.67 on the test set. A molecular generation and screening process is built to find new high-performance acceptors for PM6. Three discovered candidates are further validated by experiment, and the best PCE reaches 14.61%. The released user-friendly interface of DeepAcceptor greatly boosts the accessibility and efficiency of designing and discovering high-performance acceptors. Altogether, the DeepAcceptor framework with abcBERT is promising to predict the PCE and accelerate the discovery of high-performance acceptor materials.

An efficient deep learning-based framework for image distortion correction

An efficient deep learning-based framework for image distortion correction

Bridge the gap between simulated and real-world data in optical fiber mode decomposition for accuracy improvement: A deep learning-based co-learning framework with visual similarity-based matching

In the area of optical fiber mode decomposition (MD), data-driven deep-learning (DL) algorithms have achieved remarkable strides with unlimited theoretically synthetic data, but the differences between theoretical and real-world data in the DL context are rarely investigated. The two differences, namely the information bottleneck and the domain discrepancy, between the simulation and realistic domain remain significant obstacles that limit the practical deployment of DL-based MD approaches in real-world optical systems. The paper aims to mitigate their influence on the accuracy limitation and the accuracy degradation of DL-based MD. First, we present a DL-based co-learning framework for MD for the first time. With the co-learning process, the knowledge of the teacher model learned from rich input information sources (near-field, far-field, and phase-distribution optical images) is transferred to the student model with the only near-field input source, which breaks the information bottleneck by allowing the student model to learn more knowledge in a simulation context. Furthermore, to settle the domain discrepancy between ideal data and practical observation, we propose a visual similarity-based matching to assign real values of modal coefficients to the practical optical images whose ground truth values cannot be easily acquired, which enables the transfer learning from the synthetic data to the real-world data for accuracy enhancement. Simulated and experimental results show that the proposed co-learning framework with visual similarity-based matching has realized high precision and robustness in MD simulated and practical tasks. By bridging the gap between theoretical and practical data, our study offers novel perspectives on DL-based MD practice.

A systematic literature review on machine learning and deep learning-based covid-19 detection frameworks using X-ray Images

Coronavirus is an endangered disease to kills more than millions of people, but it has also put tremendous pressure on the whole medical system. The initial stage of identification of COVID-19 is necessary to isolate the patients with positive cases in order to stop the disease from spreading. The amalgamation of imaging techniques and deep learning algorithms takes less time and leads to more accurate outcomes for COVID-19 detection. Deep learning techniques have been employed by scientists to identify coronavirus infection in lung images during the COVID-19 worldwide epidemic. In this review, a review of the Covid-19 detection framework based on machine learning and deep learning techniques using X-ray images is done. First, the review of existing Covid-19 detection models is done. For this purpose, a detailed literature survey is carried out on Covid-19 detection papers from 2019 to 2023. Following the literature survey, the pre-processing procedures, the segmentation process, and the classification techniques used for Covid-19 detection using deep learning, machine learning, and optimization algorithms are reviewed and categorized. After that, the dataset and the implementation tool which are utilized for Covid-19 detection works are analyzed and grouped. Finally, the performance metrics validation such as accuracy, recall, F1-score, NPV, precision, sensitivity, and specificity is carried out. The research gaps in the existing Covid-19 detection techniques are provided further as references to aid in future works.

A Deep Learning-Based Framework for Predicting Intracerebral Hemorrhage Hematoma Expansion Using Head Non-contrast CT Scan

Rationale and ObjectivesHematoma expansion (HE) in intracerebral hemorrhage (ICH) is a critical factor affecting patient outcomes, yet effective clinical tools for predicting HE are currently lacking. We aim to develop a fully automated framework based on deep learning for predicting HE using only clinical non-contrast CT (NCCT) scans. Materials and MethodsA large retrospective dataset (n=2,484) was collected from 84 centers, while a prospective dataset (n=500) was obtained from 26 additional centers. Baseline NCCT scans and follow-up NCCT scans were conducted within 6 hours and 48 hours from symptom onset, respectively. HE was defined as a volume increase of more than 6 mL on the follow-up NCCT. The retrospective dataset was divided into a training set (n=1,876) and a validation set (n=608) by patient inclusion time. A two-stage framework was trained to predict HE, and its performance was evaluated on both the validation and prospective sets. Receiver operating characteristics area under the curve (AUC), sensitivity, and specificity were leveraged. ResultsOur two-stage framework achieved an AUC of 0.760 (95% CI 0.724-0.799) on the retrospective validation set and 0.806 (95% CI 0.750-0.859) on the prospective set, outperforming the commonly used BAT score, which had AUCs of 0.582 and 0.699, respectively. ConclusionOur framework can automatically and robustly identify ICH patients at high risk of HE using admission head NCCT scans, providing more accurate predictions than the BAT score.

High-dimensional multi-objective optimization of coupled cross-laminated timber walls building using deep learning

This paper introduces a deep learning-based multi-objective optimization framework, applying for advancing the design of the Cross-Laminated Timber Coupled Wall system. While traditional optimization methods often struggle with the curse of dimensionality in high-dimensional problems, the approach proposed in this study employs an autoencoder to effectively reduce the dimensionality of the design space. Subsequently, a neural network establishes a mapping between input variables and latent spaces, with another neural network forming the crucial link between these latent variables and the output responses. The proposed framework is integrated into the design process of a 20-story Cross-Laminated Timber Coupled Wall system, where uncertainties inherent in connection elements are systematically addressed to optimize the structural parameters. The non-dominated sorting genetic algorithm-II is utilized to estimate optimal design variables by minimizing three conflicting objective functions, thereby generating a Pareto front. This optimized design is then bench-marked against three deterministic models with varying coupling beam shear strengths. A two-dimensional numerical model developed in OpenSees facilitates nonlinear time history analysis using 50 bi-directional ground motion records, representative of the seismicity of Vancouver, Canada. The results of this study not only highlight the efficacy of the deep learning-based framework in enhancing the structural integrity and resilience of high-rise timber structures in seismic regions but also significantly contribute to the evolution of computational approaches in structural engineering.

High energy absorption design of porous metals using deep learning

Due to its remarkable energy absorption properties, porous metals have widespread applications in engineering. However, the high randomness of pore morphology greatly hinders the effective design and analysis of high energy absorption structures. To address this challenge, this paper first introduces a deep learning-based framework for high energy absorption-oriented design of random porous metals structures. The framework comprises two steps: (i) a generator powered by Wasserstein deep convolutional generative adversarial network is developed to swiftly generate a vast design space (∼one million samples) of porous metals with real random pore morphology. (ii) an inverse search strategy based on convolutional neural network is applied to quickly pick out the optimal structure with the best energy absorption from the design space. Results show that the optimal energy absorption is about 17.71 % higher than the maximum value of initial structures from CT scan. Additionally, a 575-fold increase in computational efficiency is achieved compared to the traversal search using finite element method. Subsequently, the deformation process of the optimal structure is analyzed focusing on the pore morphology and compression performance, showing that random porous metals with uniformly sized pores are capable of withstanding higher stress under the same strain and exhibit no yield band during compression. Inspired by this, a structural homogenization method is introduced and validated to create porous metal structure with stable microstructure evolution, extended plateau stress and high energy absorption.

Development and validation of a deep learning-based framework for automated lung CT segmentation and acute respiratory distress syndrome prediction: a multicenter cohort study

Acute respiratory distress syndrome (ARDS) is a life-threatening condition with a high incidence and mortality rate in intensive care unit (ICU) admissions. Early identification of patients at high risk for developing ARDS is crucial for timely intervention and improved clinical outcomes. However, the complex pathophysiology of ARDS makes early prediction challenging. This study aimed to develop an artificial intelligence (AI) model for automated lung lesion segmentation and early prediction of ARDS to facilitate timely intervention in the intensive care unit. A total of 928 ICU patients with chest computed tomography (CT) scans were included from November 2018 to November 2021at three centers in China. Patients were divided into a retrospective cohort for model development and internal validation, and three independent cohorts for external validation. A deep learning-based framework using the UNet Transformer (UNETR) model was developed to perform the segmentation of lung lesions and early prediction of ARDS. We employed various data augmentation techniques using the Medical Open Network for AI (MONAI) framework, enhancing the training sample diversity and improving the model's generalization capabilities. The performance of the deep learning-based framework was compared with a Densenet-based image classification network and evaluated in external and prospective validation cohorts. The segmentation performance was assessed using the Dice coefficient (DC), and the prediction performance was assessed using area under the receiver operating characteristic curve (AUC), sensitivity, and specificity. The contributions of different features to ARDS prediction were visualized using Shapley Explanation Plots. This study was registered with the China Clinical Trial Registration Centre (ChiCTR2200058700). The segmentation task using the deep learning framework achieved a DC of 0.734±0.137 in the validation set. For the prediction task, the deep learning-based framework achieved AUCs of 0.916 [0.858-0.961], 0.865 [0.774-0.945], 0.901 [0.835-0.955], and 0.876 [0.804-0.936] in the internal validation cohort, external validation cohort I, external validation cohort II, and prospective validation cohort, respectively. It outperformed the Densenet-based image classification network in terms of prediction accuracy. Moreover, the ARDS prediction model identified lung lesion features and clinical parameters such as C-reactive protein, albumin, bilirubin, platelet count, and age as significant contributors to ARDS prediction. The deep learning-based framework using the UNETR model demonstrated high accuracy and robustness in lung lesion segmentation and early ARDS prediction, and had good generalization ability and clinical applicability. This study was supported by grants from the Shanghai Renji Hospital Clinical Research Innovation and Cultivation Fund (RJPY-DZX-008) and Shanghai Science and Technology Development Funds (22YF1423300).

HTINet2: herb-target prediction via knowledge graph embedding and residual-like graph neural network.

Target identification is one of the crucial tasks in drug research and development, as it aids in uncovering the action mechanism of herbs/drugs and discovering new therapeutic targets. Although multiple algorithms of herb target prediction have been proposed, due to the incompleteness of clinical knowledge and the limitation of unsupervised models, accurate identification for herb targets still faces huge challenges of data and models. To address this, we proposed a deep learning-based target prediction framework termed HTINet2, which designed three key modules, namely, traditional Chinese medicine (TCM) and clinical knowledge graph embedding, residual graph representation learning, and supervised target prediction. In the first module, we constructed a large-scale knowledge graph that covers the TCM properties and clinical treatment knowledge of herbs, and designed a component of deep knowledge embedding to learn the deep knowledge embedding of herbs and targets. In the remaining two modules, we designed a residual-like graph convolution network to capture the deep interactions among herbs and targets, and a Bayesian personalized ranking loss to conduct supervised training and target prediction. Finally, we designed comprehensive experiments, of which comparison with baselines indicated the excellent performance of HTINet2 (HR@10 increased by 122.7% and NDCG@10 by 35.7%), ablation experiments illustrated the positive effect of our designed modules of HTINet2, and case study demonstrated the reliability of the predicted targets of Artemisia annua and Coptis chinensis based on the knowledge base, literature, and molecular docking.

A deep reinforcement learning-based scheduling framework for real-time workflows in the cloud environment

Benefiting from the flexibility, scalability and many other advantages of cloud computing, an increasing number of research institutions and enterprises are using cloud platforms to deploy their scientific or commercial workflow applications. From the perspective of cloud service providers, how to schedule workflows reasonably to improve user satisfaction and resource utilization is a big challenge. This challenge becomes more prominent in a dynamic cloud environment, where the workload of user-submitted workflows is constantly changing and unpredictable. Traditional scheduling methods designed for fixed workflows or real-time independent tasks are not applicable in such scenarios. In this paper, we present an online scheduling framework for multiple real-time workflows in a dynamic environment. In this framework, the scheduler employs a deep reinforcement learning based algorithm (R-DQN) to assign each task in workflows to a suitable virtual machine instance according to the current situation. Its goal is to achieve optimal task scheduling and resource allocation, so as to minimize workflow makespan and maximize resource utilization. We provide a comprehensive overview of the design and implementation of our approach, and we conduct experiments using various workflow benchmarks with different structures and scales on the WorkflowSim platform. The experimental results demonstrate that our algorithm significantly outperforms the compared algorithms in term of makespan and other metrics. Moreover, our algorithm demonstrates strong performance in generality, robustness and scalability.

Automatic Annotation Diagnostic Framework for Nasopharyngeal Carcinoma via Pathology-Fidelity GAN and Prior-Driven Classification.

Non-keratinizing carcinoma is the most common subtype of nasopharyngeal carcinoma (NPC). Its poorly differentiated tumor cells and complex microenvironment present challenges to pathological diagnosis. AI-based pathological models have demonstrated potential in diagnosing NPC, but the reliance on costly manual annotation hinders development. To address the challenges, this paper proposes a deep learning-based framework for diagnosing NPC without manual annotation. The framework includes a novel unpaired generative network and a prior-driven image classification system. With pathology-fidelity constraints, the generative network achieves accurate digital staining from H&E to EBER images. The classification system leverages staining specificity and pathological prior knowledge to annotate training data automatically and to classify images for NPC diagnosis. This work used 232 cases for study. The experimental results show that the classification system reached a 99.59% accuracy in classifying EBER images, which closely matched the diagnostic results of pathologists. Utilizing PF-GAN as the backbone of the framework, the system attained a specificity of 0.8826 in generating EBER images, markedly outperforming that of other GANs (0.6137, 0.5815). Furthermore, the F1-Score of the framework for patch level diagnosis was 0.9143, exceeding those of fully supervised models (0.9103, 0.8777). To further validate its clinical efficacy, the framework was compared with experienced pathologists at the WSI level, showing comparable NPC diagnosis performance. This low-cost and precise diagnostic framework optimizes the early pathological diagnosis method for NPC and provides an innovative strategic direction for AI-based cancer diagnosis.