Region Of Interest Extraction Research Articles

Bilinear Perceptual Fusion Algorithm Based on Brain Functional and Structural Data for ASD Diagnosis and Regions of Interest Identification.

Autism spectrum disorder (ASD) is a serious mental disorder with a complex pathogenesis mechanism and variable presentation among individuals. Although many deep learning algorithms have been used to diagnose ASD, most of them focus on a single modality of data, resulting in limited information extraction and poor stability. In this paper, we propose a bilinear perceptual fusion (BPF) algorithm that leverages data from multiple modalities. In our algorithm, different schemes are used to extract features according to the characteristics of functional and structural data. Through bilinear operations, the associations between the functional and structural features of each region of interest (ROI) are captured. Then the associations are used to integrate the feature representation. Graph convolutional neural networks (GCNs) can effectively utilize topology and node features in brain network analysis. Therefore, we design a deep learning framework called BPF-GCN and conduct experiments on publicly available ASD dataset. The results show that the classification accuracy of BPF-GCN reached 82.35%, surpassing existing methods. This demonstrates the superiority of its classification performance, and the framework can extract ROIs related to ASD. Our work provides a valuable reference for the timely diagnosis and treatment of ASD.

A ROI Extraction Method for Wrist Imaging Applied in Smart Bone-Age Assessment System.

Bone Age (BA) is reckoned to be closely associated with the growth and development of teenagers, whose assessment highly depends on the accurate extraction of the reference bone from the carpal bone. Being uncertain in its proportion and irregular in its shape, wrong judgment and poor average extraction accuracy of the reference bone will no doubt lower the accuracy of Bone Age Assessment (BAA). In recent years, machine learning and data mining are widely embraced in smart healthcare systems. Using these two instruments, this article aims to tackle the aforementioned problems by proposing a Region of Interest (ROI) extraction method for wrist X-ray images based on optimized YOLO model. The method combines Deformable convolution-focus (Dc-focus), Coordinate attention (Ca) module, Feature level expansion, and Efficient Intersection over Union (EIoU) loss all together as YOLO-DCFE. With the improvement, the model can better extract the features of irregular reference bone and reduce the potential misdiscrimination between the reference bone and other similarly shaped reference bones, improving the detection accuracy. We select 10041 images taken by professional medical cameras as the dataset to test the performance of YOLO-DCFE. Statistics show the advantages of YOLO-DCFE in detection speed and high accuracy. The detection accuracy of all ROIs is 99.8%, which is higher than other models. Meanwhile, YOLO-DCFE is the fastest of all comparison models, with the Frames Per Second (FPS) reaching 16.

Efficient Extraction of Coronary Artery Vessels from Computed Tomography Angiography Images Using ResUnet and Vesselness.

Accurate and efficient segmentation of coronary arteries from CTA images is crucial for diagnosing and treating cardiovascular diseases. This study proposes a structured approach that combines vesselness enhancement, heart region of interest (ROI) extraction, and the ResUNet deep learning method to accurately and efficiently extract coronary artery vessels. Vesselness enhancement and heart ROI extraction significantly improve the accuracy and efficiency of the segmentation process, while ResUNet enables the model to capture both local and global features. The proposed method outperformed other state-of-the-art methods, achieving a Dice similarity coefficient (DSC) of 0.867, a Recall of 0.881, and a Precision of 0.892. The exceptional results for segmenting coronary arteries from CTA images demonstrate the potential of this method to significantly contribute to accurate diagnosis and effective treatment of cardiovascular diseases.

Kidney, ureter, and urinary bladder segmentation based on non-contrast enhanced computed tomography images using modified U-Net

This study was performed to segment the urinary system as the basis for diagnosing urinary system diseases on non-contrast computed tomography (CT). This study was conducted with images obtained between January 2016 and December 2020. During the study period, non-contrast abdominopelvic CT scans of patients and diagnosed and treated with urinary stones at the emergency departments of two institutions were collected. Region of interest extraction was first performed, and urinary system segmentation was performed using a modified U-Net. Thereafter, fivefold cross-validation was performed to evaluate the robustness of the model performance. In fivefold cross-validation results of the segmentation of the urinary system, the average dice coefficient was 0.8673, and the dice coefficients for each class (kidney, ureter, and urinary bladder) were 0.9651, 0.7172, and 0.9196, respectively. In the test dataset, the average dice coefficient of best performing model in fivefold cross validation for whole urinary system was 0.8623, and the dice coefficients for each class (kidney, ureter, and urinary bladder) were 0.9613, 0.7225, and 0.9032, respectively. The segmentation of the urinary system using the modified U-Net proposed in this study could be the basis for the detection of kidney, ureter, and urinary bladder lesions, such as stones and tumours, through machine learning.

Development of A Palm-Vein Recognition System for Identification and Verification Systems using Enhanced Convolutional Neural Network

Biometric authentication systems have gained significant attention in access control applications due to their ability to provide enhanced security and convenience. Among various biometric modalities, palm-vein recognition has emerged as a promising approach, offering high accuracy, reliability, and resistance to forgery. However, existing palm-vein recognition systems often face challenges in implementation costs, computational efficiency, and performance limitations. This research aimed to develop an enhanced palm-vein recognition system for access control applications by optimizing a Convolutional Neural Network (CNN) architecture. A palm-vein dataset comprising 1000 images from 200 LAUTECH students was acquired, with 5 images per individual. The dataset was split into 700 training images and 300 testing images. The acquired images were pre-processed for quality enhancement and region of interest extraction. A Gravitational Search Algorithm (GSA) optimized CNN (GSA-CNN) was then employed to extract deep features from the pre-processed images, which were classified using a SoftMax layer. Experimental results revealed that the CNN technique achieved a specificity, sensitivity, False Positive Rate (FPR), accuracy of 74.60%, 79.89%, 25.40%, 77.67% at 117.52 seconds, respectively. In contrast, the proposed GSA-CNN technique demonstrated superior performance, achieving a specificity, sensitivity, FPR, accuracy of 92.06%, 92.53%, 7.94%, 92.33% at 97.14 seconds, respectively. The GSA-CNN system outperformed the conventional CNN approach in terms of accuracy, specificity, sensitivity, FPR, and processing time, demonstrating its potential for reliable and efficient palm-vein recognition in access control applications. The findings have significant implications for developing robust and secure access control systems, contributing to enhanced privacy and security across various domains.

SASAN: ground truth for the effective segmentation and classification of skin cancer using biopsy images.

Early skin cancer diagnosis can save lives; however, traditional methods rely on expert knowledge and can be time-consuming. This calls for automated systems using machine learning and deep learning. However, existing datasets often focus on flat skin surfaces, neglecting more complex cases on organs or with nearby lesions. This work addresses this gap by proposing a skin cancer diagnosis methodology using a dataset named ASAN that covers diverse skin cancer cases but suffers from noisy features. To overcome the noisy feature problem, a segmentation dataset named SASAN is introduced, focusing on Region of Interest (ROI) extraction-based classification. This allows models to concentrate on critical areas within the images while ignoring learning the noisy features. Various deep learning segmentation models such as UNet, LinkNet, PSPNet, and FPN were trained on the SASAN dataset to perform segmentation-based ROI extraction. Classification was then performed using the dataset with and without ROI extraction. The results demonstrate that ROI extraction significantly improves the performance of these models in classification. This implies that SASAN is effective in evaluating performance metrics for complex skin cancer cases. This study highlights the importance of expanding datasets to include challenging scenarios and developing better segmentation methods to enhance automated skin cancer diagnosis. The SASAN dataset serves as a valuable tool for researchers aiming to improve such systems and ultimately contribute to better diagnostic outcomes.

Hybrid federated learning with brain-region attention network for multi-center Alzheimer's disease detection

Identifying reproducible and interpretable biomarkers for Alzheimer's disease (AD) detection remains a challenge. AD detection using multi-center datasets can expand the sample size to improve robustness but might lead to a data privacy problem. Moreover, due to the high cost of labeling data, a lot of unlabeled data in each center is not fully utilized. To address this, a hybrid FL (HFL) framework is proposed that not only uses unlabeled data to train deep learning networks, but also achieves data privacy protection. We propose a novel Brain-region Attention Network (BANet), which highlights important regions via attention to represent the region of interest (ROIs).Specifically, we use a brain template to extract ROI signals from the preprocessed structure magnetic resonance imaging (sMRI) data. In addition, we add a self-supervised loss to the current loss to guide the attention map generation to learn the representations from unlabeled data. Finally, we evaluate our method on a multi-center database which is constructed using five AD datasets. The experimental results show that the proposed method performs better than state-of-the-art methods, achieving mean accuracy rates of 85.69 %, 63.34 %, and 69.89 % on the AD vs. NC, MCI vs. NC, and AD vs. MCI respectively. The source code is available for reproducibility at: https://github.com/yuliangCarmelo/HFL.

An unconstrained palmprint region of interest extraction method based on lightweight networks.

Accurately extracting the Region of Interest (ROI) of a palm print was crucial for subsequent palm print recognition. However, under unconstrained environmental conditions, the user's palm posture and angle, as well as the background and lighting of the environment, were not controlled, making the extraction of the ROI of palm print a major challenge. In existing research methods, traditional ROI extraction methods relied on image segmentation and were difficult to apply to multiple datasets simultaneously under the aforementioned interference. However, deep learning-based methods typically did not consider the computational cost of the model and were difficult to apply to embedded devices. This article proposed a palm print ROI extraction method based on lightweight networks. Firstly, the YOLOv5-lite network was used to detect and preliminarily locate the palm, in order to eliminate most of the interference from complex backgrounds. Then, an improved UNet was used for keypoints detection. This network model reduced the number of parameters compared to the original UNet model, improved network performance, and accelerated network convergence. The output of this model combined Gaussian heatmap regression and direct regression and proposed a joint loss function based on JS loss and L2 loss for supervision. During the experiment, a mixed database consisting of 5 databases was used to meet the needs of practical applications. The results showed that the proposed method achieved an accuracy of 98.3% on the database, with an average detection time of only 28ms on the GPU, which was superior to other mainstream lightweight networks, and the model size was only 831k. In the open-set test, with a success rate of 93.4%, an average detection time of 5.95ms on the GPU, it was far ahead of the latest palm print ROI extraction algorithm and could be applied in practice.

CAM-QUS guided self-tuning modular CNNs with multi-loss functions for fully automated breast lesion classification in ultrasound images

Objective. Breast cancer is the major cause of cancer death among women worldwide. Deep learning-based computer-aided diagnosis (CAD) systems for classifying lesions in breast ultrasound images can help materialise the early detection of breast cancer and enhance survival chances. Approach. This paper presents a completely automated BUS diagnosis system with modular convolutional neural networks tuned with novel loss functions. The proposed network comprises a dynamic channel input enhancement network, an attention-guided InceptionV3-based feature extraction network, a classification network, and a parallel feature transformation network to map deep features into quantitative ultrasound (QUS) feature space. These networks function together to improve classification accuracy by increasing the separation of benign and malignant class-specific features and enriching them simultaneously. Unlike the categorical crossentropy (CCE) loss-based traditional approaches, our method uses two additional novel losses: class activation mapping (CAM)-based and QUS feature-based losses, to capacitate the overall network learn the extraction of clinically valued lesion shape and texture-related properties focusing primarily the lesion area for explainable AI (XAI). Main results. Experiments on four public, one private, and a combined breast ultrasound dataset are used to validate our strategy. The suggested technique obtains an accuracy of 97.28%, sensitivity of 93.87%, F1-score of 95.42% on dataset 1 (BUSI), and an accuracy of 91.50%, sensitivity of 89.38%, and F1-score of 89.31% on the combined dataset, consisting of 1494 images collected from hospitals in five demographic locations using four ultrasound systems of different manufacturers. These results outperform techniques reported in the literature by a considerable margin. Significance. The proposed CAD system provides diagnosis from the auto-focused lesion area of B-mode BUS images, avoiding the explicit requirement of any segmentation or region of interest extraction, and thus can be a handy tool for making accurate and reliable diagnoses even in unspecialized healthcare centers.

Region of interest (ROI) extraction and crack detection for UAV-based bridge inspection using point cloud segmentation and 3D-to-2D projection

For digital-image-based bridge inspection tasks, images captured by camera-carrying unmanned aircraft vehicles (UAVs) usually contain both the region of interest (ROI) and the background. However, accurately detecting cracks in concrete surface images containing background information is challenging. To improve UAV-based bridge inspection, an image extraction and crack detection methodology is presented in this paper. First, a deep-learning-based semantic segmentation network RandLA-BridgeNet for large-scale bridge point clouds, which can facilitate 3D ROI extraction, is trained and tested. Second, an image ROI extraction method based on 3D-to-2D projection is presented to generate images containing only the ROI. Finally, a data-driven deep learning convolutional neural network (CNN) called the grid-based classification and box-based detection fusion model (GCBD) is utilized to identify cracks in the processed images. An experiment is conducted on highway bridge images to validate the presented methodology. The overall semantic segmentation and image ROI extraction accuracies are 97.0% and 98.9%, respectively. After ROI extraction, 47.9% of the grid cells, which represent background misrecognition, are filtered, greatly improving the crack identification accuracy.

The Potential of a CT-Based Machine Learning Radiomics Analysis to Differentiate Brucella and Pyogenic Spondylitis.

Pyogenic spondylitis (PS) and Brucella spondylitis (BS) are common spinal infections with similar manifestations, making their differentiation challenging. This study aimed to explore the potential of CT-based radiomics features combined with machine learning algorithms to differentiate PS from BS. This retrospective study involved the collection of clinical and radiological information from 138 patients diagnosed with either PS or BS in our hospital between January 2017 and December 2022, based on histopathology examination and/or germ isolations. The region of interest (ROI) was defined by two radiologists using a 3D Slicer open-source platform, utilizing blind analysis of sagittal CT images against histopathological examination results. PyRadiomics, a Python package, was utilized to extract ROI features. Several methods were performed to reduce the dimensionality of the extracted features. Machine learning algorithms were trained and evaluated using techniques like the area under the receiver operating characteristic curve (AUC; confusion matrix-related metrics, calibration plot, and decision curve analysis to assess their ability to differentiate PS from BS. Additionally, permutation feature importance (PFI; local interpretable model-agnostic explanations (LIME; and Shapley additive explanation (SHAP) techniques were utilized to gain insights into the interpretabilities of the models that are otherwise considered opaque black-boxes. A total of 15 radiomics features were screened during the analysis. The AUC value and Brier score of best the model were 0.88 and 0.13, respectively. The calibration plot and decision curve analysis displayed higher clinical efficiency in the differential diagnosis. According to the interpretation results, the most impactful features on the model output were wavelet LHL small dependence low gray-level emphasis (GLDN). The CT-based radiomics models that we developed have proven to be useful in reliably differentiating between PS and BS at an early stage and can provide a reliable explanation for the classification results.

Generalized weld bead region of interest localization and improved faster R-CNN for weld defect recognition

Pipeline is the main transportation. In order to ensure the safety of pipeline transportation, the research on automatic recognition of defect features in X-ray film digital images has become an important research interest. Deep learning is highly favored via its powerful feature of self-extraction ability, which avoids the complexity and uncertainty associated with manual extraction in traditional algorithms. Nevertheless, in weld seam X-ray images, the weld bead only occupies a small portion of the entire image, and the welding defects exclusively exist within the weld bead. If the weld bead area can be determined first, followed by defect extraction within the weld bead, which will undoubtedly improve the accuracy of defect detection. Therefore, this paper first proposes a simple and general two-stage weld bead ROI (region of interest) extraction algorithm, which roughly determines weld bead ROI based on the distribution of the weld bead center and accurately locates the boundaries of the weld bead ROI based on the vertical distance curve. Then, to address the issues of Faster R-CNN (region-based convolutional neural networks) only utilizing the highest-level features and the poor adaptability of prior boxes setting, FPN (feature pyramid network) and CBAM (convolutional attention mechanism module) are incorporated to the feature extraction module and K-means clustering is employed to modify the setting of prior boxes. Experimental results conducted on two weld seam datasets demonstrate the effectiveness of the two-stage weld bead ROI extraction algorithm compared with the Otsu algorithm. The mAP (mean average precision) of defect recognition based on Faster R-CNN is 80.1%, which is greatly higher than the mAP achieved without extracting weld ROI (48.7%). In addition, experimental verification is performed on the Improved Faster R-CNN algorithm for identifying weld defects, and the cumulative improvement in mAP reached 6.8%, demonstrating the effectiveness and applicability of the proposed method.

An Inspection Technique Using Fit Clearance Based on Microscopic Vision in Precision Assembly.

Inspection is a crucial process to ensure product quality. In the precision assembly of an optic-mechanical device, a part with micro multi-section arcs needs to be inspected and assembled into another part. Actually, because of machining errors, including dimensional and geometric shapes, can lead to complex deformation modes for parts with micro multi-section arcs, posing challenges to their inspection. Furthermore, inconsistencies in feature images in microscopic vision may complicate the extraction of the Region of Interest (ROI). To address these issues, this paper proposes an ROI extraction method based on the CAD model for rough positioning of feature points and connected region detection for refinement. Subsequently, based on feature points, the CAD model is used again to obtain the ROI. For inspection purposes, this paper proposes a method suitable for micro multi-section arcs based on assembly fit requirements. Experimental testing was performed on parts with eight-section arcs and mirrors to verify the effectiveness of the proposed method. This method provides a suitable solution for the inspection of micro multi-section arcs in precision assembly with the potential to improve the accuracy of the inspection results.

TGMIL: A hybrid multi-instance learning model based on the Transformer and the Graph Attention Network for whole-slide images classification of renal cell carcinoma

The pathological diagnosis of renal cell carcinoma is crucial for treatment. Currently, the multi-instance learning method is commonly used for whole-slide image classification of renal cell carcinoma, which is mainly based on the assumption of independent identical distribution. But this is inconsistent with the need to consider the correlation between different instances in the diagnosis process. Furthermore, the problem of high resource consumption of pathology images is still urgent to be solved. Therefore, we propose a new multi-instance learning method to solve this problem. In this study, we proposed a hybrid multi-instance learning model based on the Transformer and the Graph Attention Network, called TGMIL, to achieve whole-slide image of renal cell carcinoma classification without pixel-level annotation or region of interest extraction. Our approach is divided into three steps. First, we designed a feature pyramid with the multiple low magnifications of whole-slide image named MMFP. It makes the model incorporates richer information, and reduces memory consumption as well as training time compared to the highest magnification. Second, TGMIL amalgamates the Transformer and the Graph Attention's capabilities, adeptly addressing the loss of instance contextual and spatial. Within the Graph Attention network stream, an easy and efficient approach employing max pooling and mean pooling yields the graph adjacency matrix, devoid of extra memory consumption. Finally, the outputs of two streams of TGMIL are aggregated to achieve the classification of renal cell carcinoma. On the TCGA-RCC validation set, a public dataset for renal cell carcinoma, the area under a receiver operating characteristic (ROC) curve (AUC) and accuracy of TGMIL were 0.98±0.0015,0.9191±0.0062, respectively. It showcased remarkable proficiency on the private validation set of renal cell carcinoma pathology images, attaining AUC of 0.9386±0.0162 and ACC of 0.9197±0.0124. Furthermore, on the public breast cancer whole-slide image test dataset, CAMELYON 16, our model showed good classification performance with an accuracy of 0.8792. TGMIL models the diagnostic process of pathologists and shows good classification performance on multiple datasets. Concurrently, the MMFP module efficiently diminishes resource requirements, offering a novel angle for exploring computational pathology images.

PTC-MAS: A Deep Learning-Based Preoperative Automatic Assessment of Lymph Node Metastasis in Primary Thyroid Cancer.

Identifying cervical lymph node metastasis (LNM) in primary thyroid cancer preoperatively using ultrasound is challenging. Therefore, a non-invasive method is needed to assess LNM accurately. To address this need, we developed the Primary Thyroid Cancer Lymph Node Metastasis Assessment System (PTC-MAS), a transfer learning-based and B-mode ultrasound images-based automatic assessment system for assessing LNM in primary thyroid cancer. The system has two parts: YOLO Thyroid Nodule Recognition System (YOLOS) for obtaining regions of interest (ROIs) of nodules, and LMM assessment system for building the LNM assessment system using transfer learning and majority voting with extracted ROIs as input. We retained the relative size features of nodules to improve the system's performance. We evaluated three transfer learning-based neural networks (DenseNet, ResNet, and GoogLeNet) and majority voting, which had the area under the curves (AUCs) of 0.802, 0.837, 0.823, and 0.858, respectively. Method III preserved relative size features and achieved higher AUCs than Method II, which fixed nodule size. YOLOS achieved high precision and sensitivity on a test set, indicating its potential for ROIs extraction. Our proposed PTC-MAS system effectively assesses primary thyroid cancer LNM based on preserving nodule relative size features. It has potential for guiding treatment modalities and avoiding inaccurate ultrasound results due to tracheal interference.

Traffic Analysis Through Deep-Learning-Based Image Segmentation From UAV Streaming

Alongside many traditional as well as novel applications, in the latest years, drones have been widely adopted as remote sensing platforms for road traffic monitoring in urban areas and on highways. The problem of traffic monitoring on Region of Interest (RoI) based on drone imagery is a challenging task, especially when the surveillance drone is constantly moving. In this work, two specific subtasks have been addressed. The goal of the first stage is to predict the RoI in drone imagery of traffic scenes using deep-learning (DL)-based approaches instead of traditional image processing; in this connection, the goal of the second task is to perform vehicle detection on the selected RoI. To ensure diversity and robustness, drone images with different altitudes, positions, and view points have been considered. To achieve these goals, two custom aerial data sets for RoI extraction and detection were built by collecting aerial sequences from flying unmanned aerial vehicles (UAVs) and by transmitting them to the base station leveraging 5G technology. Two different ad hoc DL-based architectures have been designed for the RoI extraction task to maximize the accuracy and inference speed, respectively, and have been evaluated on two different data sets: 1) a custom-built data set and 2) a Massachusetts roads data set. Finally, the models providing the best performance have been combined to further improve the overall results. Experimental tests show that the proposed framework represents a promising solution for drone-based road traffic monitoring in critical areas, exploiting imagery from a variety of viewing angles and altitudes.

FusionRCNN: LiDAR-Camera Fusion for Two-Stage 3D Object Detection

Accurate and reliable perception systems are essential for autonomous driving and robotics. To achieve this, 3D object detection with multi-sensors is necessary. Existing 3D detectors have significantly improved accuracy by adopting a two-stage paradigm that relies solely on LiDAR point clouds for 3D proposal refinement. However, the sparsity of point clouds, particularly for faraway points, makes it difficult for the LiDAR-only refinement module to recognize and locate objects accurately. To address this issue, we propose a novel multi-modality two-stage approach called FusionRCNN. This approach effectively and efficiently fuses point clouds and camera images in the Regions of Interest (RoI). The FusionRCNN adaptively integrates both sparse geometry information from LiDAR and dense texture information from the camera in a unified attention mechanism. Specifically, FusionRCNN first utilizes RoIPooling to obtain an image set with a unified size and gets the point set by sampling raw points within proposals in the RoI extraction step. Then, it leverages an intra-modality self-attention to enhance the domain-specific features, followed by a well-designed cross-attention to fuse the information from two modalities. FusionRCNN is fundamentally plug-and-play and supports different one-stage methods with almost no architectural changes. Extensive experiments on KITTI and Waymo benchmarks demonstrate that our method significantly boosts the performances of popular detectors. Remarkably, FusionRCNN improves the strong SECOND baseline by 6.14% mAP on Waymo and outperforms competing two-stage approaches.

Zebra-crossing detection based on cascaded Hough transform principle and vanishing point characteristics

Abstract In this study, a zebra-crossing detection method based on cascaded Hough transform (CHT) and vanishing point (VP) characteristics is proposed. In this method, the principle of detecting straight lines in the parallel coordinate system is applied to zebra-crossing detection. Each edge point of the image obtained by edge detection is represented in the parallel coordinate system to find the VP. Using the VP coordinate as the judgment condition, those straight lines that do not pass through the VP but meet the straight-line condition are excluded to obtain the straight lines passing through both sides of the zebra crossing, and finally fit the edge points on the straight line, and get the zebra-crossing fitting line segment. Experiments show that CHT has obvious advantages in detection accuracy and speed compared with the Hough transform. At the same time, VPs can be used to eliminate interference segments, which provide support for the accuracy of zebra-crossing detection. This method can get zebra-crossing location information without using region of interest extraction, which also provides a reference method for road detection in some specific cases.

Multiple Eye Disease Detection Using Deep Learning

Human eyes are susceptible to various abnormalities due to aging, trauma, and diseases like diabetes. Glaucoma, cataracts, macular degeneration, and diabetic retinopathy are the leading causes of blindness worldwide. It is crucial to detect and diagnose these eye diseases early to provide timely treatment and prevent vision loss. Multiple eye disease detection through the analysis of medical images can aid in this process. The steps involved in the detection of multiple eye diseases using deep learning include image acquisition, region of interest extraction, feature extraction, and disease classification or detection. In this study, we proposed a model using deep learning algorithms, ResNetand VGG16, to detect eye diseases such as uveitis, glaucoma, crossed eyes, bulging eyes, and cataracts. We achieved a 92% accuracy rate using ResNet50 and 79% accuracy using the VGG16 model. By automating the detection process, we can save time for doctors and increase the accuracy and detection rate. The proposed model can be integrated into the healthcare system to assist in early diagnosis and effective treatment of eye diseases.

Fusion of Quality Evaluation Metrics and Convolutional Neural Network Representations for ROI Filtering in LC-MS.

Region of interest (ROI) extraction is a fundamental step in analyzing metabolomic datasets acquired by liquid chromatography-mass spectrometry (LC-MS). However, noises and backgrounds in LC-MS data often affect the quality of extracted ROIs. Therefore, developing effective ROI evaluation algorithms is necessary to eliminate false positives meanwhile keep the false-negative rate as low as possible. In this study, a deep fused filter of ROIs (dffROI) was proposed to improve the accuracy of ROI extraction by combining the handcrafted evaluation metrics with convolutional neural network (CNN)-learned representations. To evaluate the performance of dffROI, dffROI was compared with peakonly (CNN-learned representation) and five handcrafted metrics on three LC-MS datasets and a gas chromatography-mass spectrometry (GC-MS) dataset. Results show that dffROI can achieve higher accuracy, better true-positive rate, and lower false-positive rate. Its accuracy, true-positive rate, and false-positive rate are 0.9841, 0.9869, and 0.0186 on the test set, respectively. The classification error rate of dffROI (1.59%) is significantly reduced compared with peakonly (2.73%). The model-agnostic feature importance demonstrates the necessity of fusing handcrafted evaluation metrics with the convolutional neural network representations. dffROI is an automatic, robust, and universal method for ROI filtering by virtue of information fusion and end-to-end learning. It is implemented in Python programming language and open-sourced at https://github.com/zhanghailiangcsu/dffROI under BSD License. Furthermore, it has been integrated into the KPIC2 framework previously proposed by our group to facilitate real metabolomic LC-MS dataset analysis.