Image Database Research Articles

Scan-Specific Self-Supervised Bayesian Deep Non-Linear Inversion for Undersampled MRI Reconstruction.

Magnetic resonance imaging is subject to slow acquisition times due to the inherent limitations in data sampling. Recently, supervised deep learning has emerged as a promising technique for reconstructing sub-sampled MRI. However, supervised deep learning requires a large dataset of fully-sampled data. Although unsupervised or self-supervised deep learning methods have emerged to address the limitations of supervised deep learning approaches, they still require a database of images. In contrast, scan-specific deep learning methods learn and reconstruct using only the sub-sampled data from a single scan. Here, we introduce Scan-Specific Self-Supervised Bayesian Deep Non-Linear Inversion (DNLINV) that does not require an auto calibration scan region. DNLINV utilizes a Deep Image Prior-type generative modeling approach and relies on approximate Bayesian inference to regularize the deep convolutional neural network. We demonstrate our approach on several anatomies, contrasts, and sampling patterns and show improved performance over existing approaches in scan-specific calibrationless parallel imaging and compressed sensing.

Optimization of CPBIS methods applied on enhanced fibrin microbeads approach for image segmentation in dynamic databases

In the empire of image processing and computer vision, the demand for advanced segmentation techniques has intensified with the growing complexity of visual data. This study focuses on the innovative paradigm of fuzzy mountain-based image segmentation, a method that harnesses the power of fuzzy logic and topographical inspiration to achieve nuanced and adaptable delineation of image regions. This research primarily concentrates on determining the age of tigers, a critical and challenging task in the current scenario. The primary objectives include the development of a comprehensive framework for FMBIS and an in-depth investigation into its adaptability to different image characteristics. This research work incorporates those domains of image processing and data mining to predict the age of the tiger using different kinds of color images. Fuzzy mountain-based pixel segmentation arises from the need to capture the subtle gradients and uncertainties present in images, offering a novel approach to achieving high-fidelity segmentations in diverse and complex scenarios. The proposed methods enable image enhancement and filtering and are then assessed during process time, retrieval time, to give a more accurate and reduced error rate for producing higher results for real-time tiger image database.

Study on Cavitation Determination of Pump Turbine Model Test

Abstract At present, it is difficult to determine whether cavitation has occurred in the pump turbine model test, with a large human error and no more accurate judgment guidelines, this paper proposes an intelligent algorithm based on image recognition to determine when cavitation has occurred. The method collects images of water bodies in the recognition area, extracts and processes the features of fixed parts, and compares them with database image data to finally realize the recognition of cavitation phenomena. The results show that the method captures the cavitation phenomenon more accurately and efficiently, and achieves the intelligent judgment of the cavitation phenomenon.

Validation study on the practical accuracy of wood species identification via deep learning from visible microscopic images

This study aimed to validate the accuracy of identifying Japanese hardwood species from microscopic cross-sectional images using convolutional neural networks (CNN). The overarching goal is to create a versatile model that can handle microscopic cross-sectional images of wood. To gauge the practical accuracy, a comprehensive database of microscopic images of Japanese hardwood species was provided by the Forest Research and Management Organization. These images, captured from various positions on wood blocks, different trees, and diverse production areas, resulted in substantial intra-species image variation. To assess the effect of data distribution on accuracy, two datasets, D1 and D2, representing a segregated and a non-segregated dataset, respectively—from 1,000 images (20 images from each of the 50 species) were compiled. For D1, distinct images were allocated to the training, validation, and testing sets. However, in D2, the same images were used for both training and testing. Furthermore, the influence of the evaluation methodology on the identification accuracy was investigated by comparing two approaches: patch evaluation and E2 image evaluation. The accuracy of the model for uniformly sized images was approximately 90%, whereas that for variably sized images it was approximately 70%.

A Real Time Facial Recognition and Tracking System for Personnel Presence

Abstract: Our project aims to revolutionize the conventional attendance system by implementing a facial recognition-based solution. The existing manual methods are prone to human error and require significant maintenance. By leveraging facial recognition technology, our system will offer improved precision and efficiency, reducing the need for manual work. The system will maintain a database of students' images, matching them during class to mark attendance accurately. Utilizing machine learning techniques, specially the Haar-Cascade classifier and local binary pattern histogram method for face detection and recognition, respectively, our system will ensure reliable attendance tracking. The attendance data will be stored in a MySQL database and Microsoft Excel file, providing a streamlined and secure alternative to the traditional attendance process.

Image driven deep learning method with FFT solver for predicting the microscale full field stress of stochastic boundary composites

AbstractA novel image‐driven deep learning approach embedded with model‐data‐knowledge information can directly predict the full field stress distribution and mechanical properties of composites solely based on initial scanning geometry images. The fiber spatial distribution and morphological features are identified by Scanning Electron Microscopy (SEM) initially. An effective random fiber generation algorithm is further utilized to generate images database comprising equivalent geometric models of various microstructures. Subsequently, the geometry model images database is analyzed by fast Fourier transform (FFT) to produce the database including the elastic modulus parameters and full field stress distributions of composites with distinct microstructures. Finally, an innovative image‐learning comprehensive paradigm uniting convolutional neural networks and convolutional autoencoders is elaborated systematically to learn the inherent laws of geometry images and field images. The results show that the proposed method have capacity to effectively and accurately predict the elastic modulus and full field stress distributions combined with error analysis even for stochastic boundary composites. The proposed methodology is a symbiosis of cutting‐edge image learning and non‐destructive testing techniques for composites, which can directly use the local non‐destructive or scanning images of composite structural components to quickly obtain field information and further predict damage evolution online.Highlights Proposing a novel deep learning approach combined with FFT directly to predict stress distribution of stochastic boundary composites solely based on micro‐images. Adopting equivalent geometric models dispersed by higher pixels mesh density considering extreme thin interface. Incorporate adaptive algorithms for streamlined training optimization. An innovative integration of deep learning and non‐destructive testing technology for the stress distribution and properties prediction online.

Research on heritage characteristics based on railway architectural heritage database in Jinzhou section of the Peking–Mukden Railway

This research interprets the background of Jinzhou section of the Peking–Mukden Railway, and puts forward 65 heritages as cases based on the scope definition and investigation. After the data collection, processing, and visualization, the database composed of 9 sub-databases, with B/S architecture mode, is constructed based on SQL server platform. The ArcGIS tool is used to analyze the distribution of the heritages, including spatial distribution characteristics, spatial agglomeration, and spatial equilibrium. “Image and model information database” and “text attribute information database” is used to analyze the architectural ontology features. The conclusions are drawn as follows: 1) The integral distribution has the characteristics of “cohesion”, while the 5 medium types of heritages show obvious and different directions. 2) The overall pattern of spatial agglomeration is characterized by high cohesion with a single high agglomeration point as the core. The low agglomeration area shows a point-line-point pattern. 3) The integral heritages and three main types of buildings differ in distribution, and the equilibrium is low. The architectural ontology analysis shows that the image information can be used as the basis for ontology characteristics analysis, architectural form and style judgment, and architectural functional space analysis.

Quantifying lung fissure integrity using a three-dimensional patch-based convolutional neural network on CT images for emphysema treatment planning.

Evaluation of lung fissure integrity is required to determine whether emphysema patients have complete fissures and are candidates for endobronchial valve (EBV) therapy. We propose a deep learning (DL) approach to segment fissures using a three-dimensional patch-based convolutional neural network (CNN) and quantitatively assess fissure integrity on CT to evaluate it in subjects with severe emphysema. From an anonymized image database of patients with severe emphysema, 129 CT scans were used. Lung lobe segmentations were performed to identify lobar regions, and the boundaries among these regions were used to construct approximate interlobar regions of interest (ROIs). The interlobar ROIs were annotated by expert image analysts to identify voxels where the fissure was present and create a reference ROI that excluded non-fissure voxels (where the fissure is incomplete). A CNN configured by nnU-Net was trained using 86 CT scans and their corresponding reference ROIs to segment the ROIs of left oblique fissure (LOF), right oblique fissure (ROF), and right horizontal fissure (RHF). For an independent test set of 43 cases, fissure integrity was quantified by mapping the segmented fissure ROI along the interlobar ROI. A fissure integrity score (FIS) was then calculated as the percentage of labeled fissure voxels divided by total voxels in the interlobar ROI. Predicted FIS (p-FIS) was quantified from the CNN output, and statistical analyses were performed comparing p-FIS and reference FIS (r-FIS). The absolute percent error mean (±SD) between r-FIS and p-FIS for the test set was 4.0% (), 6.0% (), and 12.2% () for the LOF, ROF, and RHF, respectively. A DL approach was developed to segment lung fissures on CT images and accurately quantify FIS. It has potential to assist in the identification of emphysema patients who would benefit from EBV treatment.

Vertebral Bone Density Variations in Scoliotic vs. Non-Scoliotic Juveniles, and Its Implications for Schmorl’s Node Development: A CT-Based Analysis Using the New Mexico Decedent Image Database

Scoliosis, characterized by an abnormal lateral curvature of the spine, is predominantly idiopathic, underscoring the need to delve into its underlying causes for effective treatment and preventive strategies. This study investigates a potential correlation between scoliosis and Schmorl’s nodes (intervertebral disc herniations) influenced by Wolff’s law, which posits that bones adapt to external pressures. We analyzed CT scans from 108 juvenile decedents, including 56 with scoliosis and 52 without. After running multiple statistical tests, there was no significance between the mean bone density when compared to having scoliosis. An independent t-test provided a t-value of 0.041, which, when compared to the original significance level of 0.05, is statistically significant, although weak. When compared to the Bonferroni correction level of 0.008, it throws out the significance to give a result of not being statistically significant. It was the same in the cases of L3 (t = 0.103), L2 (t = 0.084), and L1 (t = 0.053). If compared to the regular significance level of 0.05, T12 (t = 0.012) and T11 (t = 0.042) had weak significance, but that was then excluded when the Bonferroni correction was applied. When looking at any significance of densities in different vertebral regions, the results from a one-way ANOVA (p-value = 0.213) suggest that it is likely that the results are due to random variability or chance, and that there is no statistical significance. With a value of 0.273 from a Chi-squared (χ2)/Fisher’s exact test, it suggests that there is no statistically significant correlation or difference between the variables of scoliosis and Schmorl’s nodes. The general pattern seems to follow that as the spine ascends, the density increases, and this is true in both scoliotic and non-scoliotic individuals. As a whole, it is evident that those with scoliosis have a lower vertebral density than those without, in all of the vertebral regions. There is, however, a weak negative linear relationship between bone density and age in both scoliotic and non-scoliotic individuals. A p-value of −0.229 obtained from a Pearson correlation coefficient analysis in non-scoliotic individuals, as well as a p-value of −0.069 in scoliotic individuals, was obtained. Overall, the findings of this study are comparable to some existing studies on similar topics, but there are few results that hold statistical significance and so this would be interesting to research further, potentially using a different dataset or a larger sample size that is more representative.

Prevalence of Chronic Rhino Sinusitis in Patients with Concha Bullosa.

This study aims to investigate the prevalence of chronic rhino sinusitis in patients with concha bullosa, a common anatomical variation characterised by pneumatization of the middle turbinates. Chronic rhino sinusitis is an inflammatory disorder of the paranasal sinuses and the linings of the nasal passages that lasts for 12weeks or longer. The incidence of chronic rhino sinusitis was assessed among patients with diagnosed concha bullosa, focusing on the presence of symptoms, and severity. There seems to be a strong relationship between concha bullosa and the development of chronic rhino sinusitis To screen the patients of concha bullosa for chronic rhino sinusitis. To find out of the percentage of chronic rhino sinusitis in patients with concha bullosa. To determine the sinuses affected and their respective sides. This is a retrospective study to investigate the incidence of chronic rhino sinusitis in patients diagnosed with concha bullosa. The study was conducted at Vydehi institute of medical science and research centre, Whitefield, Bangalore and the data from 01/02/23 to 01/09/23 were analyzed. Medical records of patients with diagnosed concha bullosa were identified through electronic medical records and imaging databases (CT scan paranasal sinuses). Inclusion criteria consisted of patients with a confirmed diagnosis of concha bullosa based on radiological imaging (CT scans) and clinical assessment.Exclusion criteria: patients with previous sinus surgeries and malignancy. The study's methods focused on retrospectively assessing the incidence of chronic rhino sinusitis in patients diagnosed with concha bullosa. In the above study it is concluded that concha bullosa plays major role in development of chronic rhino sinusitis. Thus after confirming the presence of concha bullosa practitioner should keep in mind the possibility of development of chronic rhino sinusitis.

COVID-19 Detection from Computed Tomography Images Using Slice Processing Techniques and a Modified Xception Classifier.

This paper extends our previous method for COVID-19 diagnosis, proposing an enhanced solution for detecting COVID-19 from computed tomography (CT) images using a lean transfer learning-based model. To decrease model misclassifications, two key steps of image processing were employed. Firstly, the uppermost and lowermost slices were removed, preserving sixty percent of each patient's slices. Secondly, all slices underwent manual cropping to emphasize the lung areas. Subsequently, resized CT scans (224 × 224) were input into an Xception transfer learning model with a modified output. Both Xception's architecture and pretrained weights were leveraged in the method. A big and rigorously annotated database of CT images was used to verify the method. The number of patients/subjects in the dataset is more than 5000, and the number and shape of the slices in each CT scan varies greatly. Verification was made both on the validation partition and on the test partition of unseen images. Results on the COV19-CT database showcased not only improvement from our previous solution and the baseline but also comparable performance to the highest-achieving methods on the same dataset. Further validation studies could explore the scalability and adaptability of the developed methodologies across diverse healthcare settings and patient populations. Additionally, investigating the integration of advanced image processing techniques, such as automated region of interest detection and segmentation algorithms, could enhance the efficiency and accuracy of COVID-19 diagnosis.

The Diverse Mountainous Landslide Dataset (DMLD): A High-Resolution Remote Sensing Landslide Dataset in Diverse Mountainous Regions

The frequent occurrence of landslides poses a serious threat to people’s lives and property. In order to evaluate disaster hazards based on remote sensing images via machine learning means, it is essential to establish an image database with manually labeled boundaries of landslides. However, the existing datasets do not cover diverse types of mountainous landslides. To address this issue, we propose a high-resolution (1 m) diverse mountainous landslide remote sensing dataset (DMLD), including 990 landslide instances across different terrain in southwestern China. To evaluate the performance of the DMLD, seven state-of-the-art deep learning models with different loss functions were implemented on it. The experiment results demonstrate not only that all of these deep learning methods with different characteristics can adapt well to the DMLD, but also that the DMLD has potential adaptability to other geographical regions.

A Deep Learning Model for Automatically Quantifying the Anterior Segment in Ultrasound Biomicroscopy Images of Implantable Collamer Lens Candidates

ObjectiveThis study aimed to develop and evaluate a deep learning-based model that could automatically measure anterior segment (AS) parameters on preoperative ultrasound biomicroscopy (UBM) images of implantable Collamer lens (ICL) surgery candidates. MethodsA total of 1164 panoramic UBM images were preoperatively obtained from 321 patients who received ICL surgery in the Eye Center of Renmin Hospital of Wuhan University (Wuhan, China) to develop an imaging database. First, the UNet++ network was utilized to segment AS tissues automatically, such as corneal lens and iris. In addition, image processing techniques and geometric localization algorithms were developed to automatically identify the anatomical landmarks (ALs) of pupil diameter (PD), anterior chamber depth (ACD), angle-to-angle distance (ATA), and sulcus-to-sulcus distance (STS). Based on the results of the latter two processes, PD, ACD, ATA, and STS can be measured. Meanwhile, an external dataset of 294 images from Huangshi Aier Eye Hospital was employed to further assess the model's performance in other center. Lastly, a subset of 100 random images from the external test set was chosen to compare the performance of the model with senior experts. ResultsWhether in the internal test dataset or external test dataset, using manual labeling as the reference standard, the models achieved a mean Dice coefficient exceeding 0.880. Additionally, the intra-class correlation coefficients (ICCs) of ALs’ coordinates were all greater than 0.947, and the percentage of Euclidean distance distribution of ALs within 250 μm was over 95.24%.While the ICCs for PD, ACD, ATA, and STS were greater than 0.957, furthermore, the average relative error (ARE) of PD, ACD, ATA, and STS were below 2.41%. In terms of human versus machine performance, the ICCs between the measurements performed by the model and those by senior experts were all greater than 0.931. ConclusionA deep learning-based model could measure AS parameters using UBM images of ICL candidates, and exhibited a performance similar to that of a senior ophthalmologist.

Investigating visual navigation using spiking neural network models of the insect mushroom bodies.

Ants are capable of learning long visually guided foraging routes with limited neural resources. The visual scene memory needed for this behaviour is mediated by the mushroom bodies; an insect brain region important for learning and memory. In a visual navigation context, the mushroom bodies are theorised to act as familiarity detectors, guiding ants to views that are similar to those previously learned when first travelling along a foraging route. Evidence from behavioural experiments, computational studies and brain lesions all support this idea. Here we further investigate the role of mushroom bodies in visual navigation with a spiking neural network model learning complex natural scenes. By implementing these networks in GeNN-a library for building GPU accelerated spiking neural networks-we were able to test these models offline on an image database representing navigation through a complex outdoor natural environment, and also online embodied on a robot. The mushroom body model successfully learnt a large series of visual scenes (400 scenes corresponding to a 27 m route) and used these memories to choose accurate heading directions during route recapitulation in both complex environments. Through analysing our model's Kenyon cell (KC) activity, we were able to demonstrate that KC activity is directly related to the respective novelty of input images. Through conducting a parameter search we found that there is a non-linear dependence between optimal KC to visual projection neuron (VPN) connection sparsity and the length of time the model is presented with an image stimulus. The parameter search also showed training the model on lower proportions of a route generally produced better accuracy when testing on the entire route. We embodied the mushroom body model and comparator visual navigation algorithms on a Quanser Q-car robot with all processing running on an Nvidia Jetson TX2. On a 6.5 m route, the mushroom body model had a mean distance to training route (error) of 0.144 ± 0.088 m over 5 trials, which was performance comparable to standard visual-only navigation algorithms. Thus, we have demonstrated that a biologically plausible model of the ant mushroom body can navigate complex environments both in simulation and the real world. Understanding the neural basis of this behaviour will provide insight into how neural circuits are tuned to rapidly learn behaviourally relevant information from complex environments and provide inspiration for creating bio-mimetic computer/robotic systems that can learn rapidly with low energy requirements.

Multi-centre benchmarking of deep learning models for COVID-19 detection in chest x-rays.

This study is a retrospective evaluation of the performance of deep learning models that were developed for the detection of COVID-19 from chest x-rays, undertaken with the goal of assessing the suitability of such systems as clinical decision support tools. Models were trained on the National COVID-19 Chest Imaging Database (NCCID), a UK-wide multi-centre dataset from 26 different NHS hospitals and evaluated on independent multi-national clinical datasets. The evaluation considers clinical and technical contributors to model error and potential model bias. Model predictions are examined for spurious feature correlations using techniques for explainable prediction. Models performed adequately on NHS populations, with performance comparable to radiologists, but generalised poorly to international populations. Models performed better in males than females, and performance varied across age groups. Alarmingly, models routinely failed when applied to complex clinical cases with confounding pathologies and when applied to radiologist defined "mild" cases. This comprehensive benchmarking study examines the pitfalls in current practices that have led to impractical model development. Key findings highlight the need for clinician involvement at all stages of model development, from data curation and label definition, to model evaluation, to ensure that all clinical factors and disease features are appropriately considered during model design. This is imperative to ensure automated approaches developed for disease detection are fit-for-purpose in a clinical setting.

Scalable approach to create annotated disaster image database supporting AI-driven damage assessment

As coastal populations surge, the devastation caused by hurricanes becomes more catastrophic. Understanding the extent of the damage is essential as this knowledge helps shape our plans and decisions to reduce the effects of hurricanes. While community and property-level damage post-hurricane damage assessments are common, evaluations at the building component level, such as roofs, windows, and walls, are rarely conducted. This scarcity is attributed to the challenges inherent in automating precise object detections. Moreover, a significant disconnection exists between manual damage assessments, typically logged-in spreadsheets, and images of the damaged buildings. Extracting historical damage insights from these datasets becomes arduous without a digital linkage. This study introduces an innovative workflow anchored in state-of-the-art deep learning models to address these gaps. The methodology offers enhanced image annotation capabilities by leveraging large-scale pre-trained instance segmentation models and accurate damaged building component segmentation from transformer-based fine-tuning detection models. Coupled with a novel data repository structure, this study merges the segmentation mask of hurricane-affected components with manual damage assessment data, heralding a transformative approach to hurricane-induced building damage assessments and visualization.

Human–machine partnerships at the exascale: exploring simulation ensembles through image databases

Human–machine partnerships at the exascale: exploring simulation ensembles through image databases

Implications of trial eligibility in patients with heart failure with mildly reduced or preserved ejection fraction.

Clinical trials in heart failure with mildly reduced or preserved ejection fraction (HFmrEF/HFpEF) commonly have detailed eligibility criteria. This may contribute to challenges with efficient enrolment and questions regarding the generalizability of trial findings. Patients with HFmrEF/HFpEF from a large US healthcare system were identified through a computable phenotype applied in linked imaging and electronic health record databases. We evaluated shared eligibility criteria from five recent/ongoing HFmrEF/HFpEF trials (PARAGON-HF, EMPEROR-Preserved, DELIVER, FINE-ARTS, and SPIRRIT-HFpEF) and compared clinical and echocardiographic features as well as outcomes between trial-eligible and trial-ineligible patients. Among 5552 patients with HFpEF/HFmrEF, 792 (14%) were eligible for trial consideration, having met all criteria assessed. Causes of ineligibility included lack of recent loop diuretics (37%), significant pulmonary disease (24%), reduced estimated glomerular filtration rate (17%), recent stroke/transient ischaemic attack (13%), or low natriuretic peptides (12%); 53% of ineligible patients had >1 reason for exclusion. Compared with eligible patients, ineligible patients were younger (age 71 vs. 75years, P<0.001) with higher rates of coronary artery disease (66% vs. 59%, P<0.001) and peripheral vascular disease (40% vs. 33%, P<0.001), but less mitral regurgitation, lower E/e' ratio, and smaller left atrial sizes. Both eligible and ineligible patients demonstrated high rates of structural heart disease consistent with HFpEF [elevated left atrial size or left ventricular (LV) hypertrophy/increased LV mass], although this was slightly higher among eligible patients (95% vs. 92%, P=0.001). The two cohorts demonstrated similar LV global longitudinal strain along with a similar prevalence of atrial fibrillation/flutter, hypertension, and obesity. Ineligible patients had similar all-cause mortality (33% vs. 33% at 3years) to those eligible but lower rates of heart failure hospitalization (20% vs. 28% at 3years, P<0.001). Among patients with HFmrEF/HFpEF from a large health system, approximately one in seven were eligible for major trials based on key criteria applied through a clinical computable phenotype. These findings highlight the large proportion of patients with HFmrEF/HFpEF ineligible for contemporary trials for whom the generalizability of trial findings may be questioned and further investigation would be beneficial.

Automated Reservoir Characterization of Carbonate Rocks using Deep Learning Image Segmentation Approach

Summary The objective of this study is to develop a systematic and novel workflow for the automated and objective characterization of carbonate reservoirs with the help of deep learning architectures. An image database of more than 6,000 carbonate thin-section images was generated using the optical microscope and image augmentation techniques. Five features, namely clay/silt/mineral, calcite, pores, fossils, and opaque minerals, were identified with the help of manual petrography of the thin sections under the microscope. A total of four deep learning models were developed, which included U-Net, U-Net with ResNet34 backbone, U-Net with Mobilenetv2 backbone, and LinkNet with ResNet34 backbone. The Ensemble model of U-Net + ResNet34 and U-Net + MobileNetv2 yielded the highest intersection over union (IoU) score of 75%, followed by the U-Net + ResNet34 model with an IoU score of 61%. The models struggled with class imbalance, which was very prominent in the image database, with classes such as fossils and opaques considered to be rare. The statistical analysis of the relative errors revealed that the major classes play a more important role in increasing the final IoU score as opposed to the common understanding that the rare classes affect the model performance. The novel workflow developed in this paper can be extended to real carbonate reservoirs for time efficient, objective, and accurate characterization.

G2NPAN: GAN-guided nuance perceptual attention network for multimodal medical fusion image quality assessment.

Multimodal medical fusion images (MMFI) are formed by fusing medical images of two or more modalities with the aim of displaying as much valuable information as possible in a single image. However, due to the different strategies of various fusion algorithms, the quality of the generated fused images is uneven. Thus, an effective blind image quality assessment (BIQA) method is urgently required. The challenge of MMFI quality assessment is to enable the network to perceive the nuances between fused images of different qualities, and the key point for the success of BIQA is the availability of valid reference information. To this end, this work proposes a generative adversarial network (GAN) -guided nuance perceptual attention network (G2NPAN) to implement BIQA for MMFI. Specifically, we achieve the blind evaluation style via the design of a GAN and develop a Unique Feature Warehouse module to learn the effective features of fused images from the pixel level. The redesigned loss function guides the network to perceive the image quality. In the end, the class activation mapping supervised quality assessment network is employed to obtain the MMFI quality score. Extensive experiments and validation have been conducted in a database of medical fusion images, and the proposed method is superior to the state-of-the-art BIQA method.