Convolutional Networks Can Learn to Generate Affinity Graphs for Image Segmentation

Brain hemorrhage segmentation in Computed Tomography (CT) scan images is challenging, due to low image contrast and large variations of hemorrhages in appearance. Unlike the previous approaches estimating the binary masks of hemorrhages directly, we newly introduce affinity graph, which is a graph representation of adjacent pixel connectivity to a U-Net segmentation network. The affinity graph can encode various regional features of the hemorrhages and backgrounds. Our segmentation network is trained in an end-to-end manner to learn the affinity graph as intermediate features and predict the hemorrhage boundaries from the graph. By learning the pixel connectivity using the affinity graph, we achieve better performance on the hemorrhage segmentation, compared to the conventional U-Net which just learns segmentation masks as targets directly. Experiments in this paper demonstrate that our model can provide higher Dice score and lower Hausdorff distance than the conventional U-Net training only segmentation map, and the model can also improve segmentation at hemorrhagic regions with blurry boundaries.

Retinal fundus images play a crucial role in the early detection of eye problems, aiding in timely diagnosis and treatment to prevent vision loss or blindness. With advancements in technology, Convolutional Neural Network (CNN) algorithms have emerged as effective tools for recognition, delineation, and classification tasks. This study proposes a comprehensive review of CNN algorithms used for retinal fundus image segmentation and classification. Our review follows a systematic approach, exploring diverse repositories to identify studies employing CNN to segment and classify retinal fundus images. Utilizing CNNs in the segmentation and classification of retinal fundus images can enhance the precision of segmentation outcomes and alleviate the sole dependence on human experts. This approach enables more accurate segmentation results, reducing the burden on human experts. A total of sixty-two studies are included in our review, analyzing aspects such as database usage and the advantages and disadvantages of the methods employed. The review provides valuable insights, limitations, observations, and future directions in the field. Despite certain limitations, the findings indicate that CNN algorithms consistently achieve high accuracies. The comprehensive examination of the included studies sheds light on the potential of CNN in retinal fundus image analysis.

Deep learning has achieved great success in the field of computer vision, and the precision in image classification and image detection has surpassed humans. Therefore, this paper combines deep learning and medical image segmentation, focusing on how to improve the accuracy and speed of segmentation algorithm of medical exercise rehabilitation image. Aiming at the shortcomings of traditional medical image recognition methods, a medical exercise rehabilitation image segmentation algorithm based on hierarchical features of convolutional neural networks is proposed, this paper calls it as hierarchical features of convolutional neural networks (HFCNN). The algorithm firstly samples the convolution output of multiple layers in the convolutional neural network to a unified scale and combines them to construct a hierarchical feature. This hierarchical feature combines the structural information of objects contained in the shallow layer of the network with the semantic information of objects contained in the deep layers of the network, so it has a strong ability to express. Secondly, the image can be segmented into multiple super pixels by the super pixel segmentation algorithm. The classifier is trained using the hierarchical features of the super pixel, and then the classification result is mapped back to the pixel. Finally, a fully connected conditional random field algorithm including one-potential potential energy and paired potential energy is constructed. The corresponding energy function is used to smooth the classification result of the pixel, and the regional consistency and continuity of the pixel mark are improved. Compared with many classical convolutional neural network algorithms, this algorithm not only accelerates the network convergence speed, shortens the training time, but also significantly improves the accuracy of segmentation algorithm of medical exercise rehabilitation image, showing good practical value.

Kidneys are most easily segmented by convolutional neural networks (CNN) on contrast enhanced CT (CECT) images, but their segmentation accuracy may be reduced when only non-contrast CT (NCCT) images are available. The purpose of this work was to investigate the improvement in segmentation accuracy when implementing a generative adversarial network (GAN) to create virtual contrast enhanced (vCECT) images from non-contrast inputs. A 2D cycleGAN model, incorporating an additional idempotent loss function to restrict the GAN from making unnecessary modifications to data already in the translated domain, was trained to generate virtual contrast enhanced images on 286 paired non-contrast and contrast enhanced inputs. A 3D CNN trained on contrast enhanced images was applied to segment the kidneys in a test set of 20 paired non-contrast and contrast enhanced images. The non-contrast images were converted to virtual contrast enhanced images, then kidneys in both image conditions were segmented by the CNN. Segmentation results were compared to analyst annotations on non-contrast images visually and by Dice Coefficient (DC). Segmentation on virtual contrast enhanced images were more complete with fewer extraneous detections compared to non-contrast images in 16/20 cases. Mean(±SD) DC was 0.88(±0.80), 0.90(±0.03), and 0.95(±0.05) for non-contrast, virtual contrast enhanced, and real contrast enhanced, respectively. Virtual contrast enhancement visually improved segmentation quality, poor performing cases had their performance improved resulting in an overall reduction in DC variation, and the minimum DC increased from 0.65 to 0.85. This work provides preliminary results demonstrating the potential effectiveness of using a GAN for virtual contrast enhancement to improve CNN-based kidney segmentation on non-contrast images.

Medical image segmentation is a fundamental task in medical image analysis, playing a crucial role in disease diagnosis, treatment planning, and clinical decision-making. Accurate segmentation of anatomical structures, such as blood vessels, organs, and lesions, is essential for reliable medical interpretations. However, existing segmentation models often face challenges in effectively capturing both local and global features within medical images. Traditional convolutional neural networks (CNNs) excel at extracting fine-grained local features but struggle to model long-range dependencies due to their limited receptive fields. Conversely, transformer-based models can capture global contextual relationships but often fail to preserve detailed local structures, leading to suboptimal segmentation performance. The purpose of this study is to develop SEFormer, a novel hybrid network that integrates SENet, ResNet, and Transformer to enhance medical image segmentation. SEFormer aims to effectively capture both local and global features, addressing the limitations of traditional CNNs and transformer-based models.SENet is incorporated to recalibrate feature maps by adaptively emphasizing informative features, thereby compensating for the global representation limitations of CNNs. Meanwhile, the ResNet backbone ensures deep feature extraction while maintaining computational efficiency. Additionally, the Transformer module is introduced to capture long-range dependencies and enhance contextual awareness, complementing the local feature extraction capabilities of CNNs. This hybrid approach allows SEFormer to balance fine-grained spatial details with broader contextual information, leading to more precise segmentation results.Furthermore, to mitigate information loss during feature extraction and ensure a more complete and hierarchical representation of image information, we introduce a feature pyramid structure inspired by multi-scale image pyramid models. By progressively increasing the receptive field across different scales and employing skip connections, SEFormer effectively fuses multi-scale local and global features at each stage of the pyramid. This hierarchical fusion mechanism ensures a richer and more robust feature representation, which is particularly beneficial for segmenting complex medical structures such as blood vessels. We evaluate SEFormer on the CHASE_DB1 dataset, a widely used benchmark for retinal vessel segmentation. Experimental results demonstrate that SEFormer outperforms existing state-of-the-art segmentation methods, achieving a 3.25% improvement in segmentation accuracy. Additionally, we conduct ablation studies to verify the contributions of different components within SEFormer. The results show that incorporating SENet enhances feature recalibration and channel attention, while the Transformer module significantly improves global context awareness. The feature pyramid structure further contributes to performance gains by ensuring a multi-scale representation of vascular structures. Compared to conventional CNN-based methods, our approach achieves better segmentation accuracy with improved robustness to variations in vessel thickness, noise, and image contrast. Furthermore, SEFormer maintains computational efficiency, making it suitable for real-world clinical applications where both accuracy and processing speed are critical. SEFormer provides an efficient and accurate approach for medical image segmentation by leveraging the complementary strengths of CNNs, SENet, and Transformers while integrating a feature pyramid structure to maximize feature representation. Future research directions include extending SEFormer to other medical imaging modalities, such as CT and MRI scans, as well as exploring further optimization techniques to reduce computational costs while maintaining high segmentation performance. Additionally, integrating SEFormer with active learning frameworks could further improve segmentation performance by leveraging expert-annotated data more efficiently.

With the increasingly important role of image segmentation in the field of computed tomography (CT) image segmentation, the requirements for image segmentation technology in related industries are constantly improving. When the hardware resources can fully meet the needs of the fast and high-precision image segmentation program system, the main means of how to improve the image segmentation effect is to improve the related algorithms. Therefore, this study has proposed a combination of genetic algorithm (GA) and Great Law (OTSU) algorithm to form an image segmentation algorithm-immune genetic algorithm (IGA) algorithm. The algorithm has improved the segmentation accuracy and efficiency of the original algorithm, which is beneficial to the more accurate results of CT image segmentation. The experimental results in this study have shown that the operating efficiency of the OTSU segmentation algorithm is up to 75%. The operating efficiency of the GA algorithm is up to 78%. The operating efficiency of the IGA algorithm is up to 92%. In terms of operating efficiency, the OTSU segmentation algorithm has more advantages. In terms of segmentation accuracy, the highest accuracy rate of OTSU segmentation algorithm is 45%. The accuracy of the GA algorithm is 80%. The highest accuracy of the IGA algorithm is 97%. The IGA algorithm is more powerful in terms of operating efficiency and accuracy. Therefore, the application of the IGA algorithm to CT image segmentation is beneficial to doctors to better judge the lesions and improve the diagnosis rate.

The paper investigates computer diagnostic systems, their architectures, methods, and algorithms used in their work to diagnose cancer, including breast, lung, brain, and other tumors. Traditional and neural network methods for image segmentation and classification are analyzed and compared, and diagnostic tools in medicine are analyzed. The key approaches to medical image processing are investigated, in particular, the analysis of segmentation methods based on U-Net networks and classification using convolutional neural networks. It is established that neural network methods outperform traditional approaches in terms of accuracy and efficiency in segmentation and classification tasks. The main advantages of using neural network architectures in computer diagnostic systems are revealed, in particular, the possibility of automating the diagnostic process and improving the accuracy of the results. Neural network-based solutions provide a more adaptive and scalable approach that can be trained and improved as new data becomes available, making them highly suitable for rapidly evolving industries such as medical diagnostics. Computer experiments on image preprocessing and segmentation were conducted. Thus, the effectiveness of U-Net networks for image segmentation tasks was established. An automatic diagnostic method based on U-Net and convolutional neural networks has been developed that reduces the diagnostic time due to parallel image processing. The paper presents a detailed scheme of the developed method. It includes the segmentation of immunohistochemical images, after which the quantitative characteristics and classification of histological images will be calculated, followed by the combination of all the results obtained to make a diagnosis. This approach provides a comprehensive analysis that combines structural and quantitative data, which helps to increase the reliability of diagnostic results The scientific novelty of the developed method of automatic diagnosis based on neural networks is the use of a parallel approach to performing segmentation and classification. The developed method can be used in computer diagnostic systems in medicine. The use of the developed method gives an increase in the speed of data processing and, accordingly, diagnosis. In addition, parallel execution contributes to more efficient use of computing resources.

Automatically generating summaries of user-generated videos is very useful but challenging. User-generated videos are unedited and usually only contain a long single shot which makes traditional video temporal segmentation methods such as shot boundary detection ineffective in producing meaningful video segments for summarization. To address this issue, we propose a novel temporal segmentation framework based on the clustering of joint visual and semantic affinity graph of the video frames. Based on a pre-trained deep convolutional neural network (CNN), we extract deep visual features of the frames to construct the visual affinity graph. We then construct the semantic affinity graph of the frames based on word embedding of the frames' semantic tags generated from an automatic image tagging algorithm. A dense neighbor method is then used to cluster the joint visual and semantic affinity graph to divide the video into sub-shot level segments and from which a summary of the video can be generated. Experimental results show that our approach outperforms state-of-the-art methods. Furthermore, we show that the method achieves results that are similar to those performed manually.

Brain cancer is a life-threatening disease that affects the brain. Detecting a tumor at an early stage becomes critical to saving lives. For the study of the brain and various medical images, magnetic resonance imaging and image segmentation algorithms have grown to be important medical diagnostic tools. Object detection and classification are important tasks in computer vision and image processing. Medicinal images play an important part in the diagnosis of tumors, but the detection of cancer faces many challenges. In this paper, the developed model has been suggested to classify CT-scan images of the brain at the beginning of production to have a tumor or recognition. It is built on a Deep Convolutional Neural Network (DCNN), and it has a particular architecture for detecting complex data features. Python has used to build the proposed model. The total dataset of the proposed network contains 4,500 images. The system used 3,600 images for training and 900 images for testing. The total number of parameters during the learning of the proposed model is 2,398,750. The highest accuracy is reached when the image size is 50x50 pixels and the number of epochs is 100. The suggested network has high performance to predict the tumor in different conditions. The work's experimental outcomes demonstrated the tumor's effectiveness with 98.5 % accuracy with in a very limited time. This model supports healthcare practitioners in making decisions about the presence or absence of disease. It is a more accurate, fast, and effective method to recognize the brain cancer based on Deep Learning technology.

Image segmentation of brain low-grade glioma (LGG) magnetic resonance imaging (MRI) contributes tremendously to diagnosis, classification and treatment of the disease. A tangible, accurate, reliable and fast image segmentation technique is demanded in clinical diagnosis and research. The emerging machine learning technique has been demonstrated its unique capability in the field of medical image processing, including medical image segmentation. Support vector machine (SVM) and convolutional neural network (CNN) are two widely used machine learning methods. In this work, image segmentation tools based on SVM and CNN are developed and evaluated for brain LGG MR image segmentation studies. The segmentation performance in terms of accuracy and cost is quantitatively analyzed and compared between the SVM and CNN techniques developed. Computed on the Google CoLab, each of the 109 SVM models represents an individual patient, is trained using a single image of that patient and takes a few seconds to complete. The CNN model is trained on a drastically larger dataset of 19,760 data augmented images and takes approximately 2 hours to obtain the most optimal result. The SVM models achieved an average and median accuracy of 0.937 and 0.976 respectively, precision of 0.456 and 0.535 respectively, recall of 0.878 and 0.906 respectively, and F1 score of 0.546 and 0.662 respectively. Although the CNN model required a significantly longer calculation time, it surpassed the SVM models in performance in LGG MR image segmentation, achieving an accuracy of 0.998, a precision of 0.999, a recall of 0.999 and an F1 score of 0.999. This study shows that SVM with appropriate filtering techniques is capable of obtaining reliable and fast segmentation of brain LGG MR images with sufficient accuracy and limited image data. CNN technique outperforms SVM in the accuracy of segmentation with requirements of significantly enlarged data set, long computation time and high-performance computer.

A Parkinson's disease-related nuclei segmentation network based on CNN-Transformer interleaved encoder with feature fusion.

For the purpose of analyzing the application of back propagation (BP) neural network model in retinal vascular segmentation of color fundus images, in this study, the traditional BP neural network was first improved by adopting the additional momentum method. Second, adaptive histogram, morphological background subtraction, Gauss preprocessing matching filter and Heisen matrix were used to enhance the image and extract the features. Third, the retinal vascular segmentation algorithm for color fundus images was constructed based on optimized BP neural network. Finally, the DRIVE and MESSIDOR data sets of color fundus images were introduced to compare the proposed algorithm with the convolutional neural network (CNN) and pulse-coupled neural network (PCNN) algorithms in terms of performance. In addition, the three algorithms were also compared in terms of the sensitivity (Se), specificity (Sp), accuracy (Ac), average operation time for each image and F1 value. The results show that the BP neural network algorithm proposed in this study shows obvious advantage over the other two algorithms in the segmentation of color fundus images. In the DRIVE data set, the Se (81.37%), Sp (90.55%) and Ac (95.82%) of BP algorithm are the highest among the three; in the MESSIDOR data set, the Se (85.22%), Sp (91.08%) and Ac (96.16%) of BP algorithm are also highest; in the DRIVE and MESSIDOR data sets, the operation time of BP algorithm is (28.46 ± 3.19 ms; 24.73 ± 4.81 ms, respectively), which are significantly less than the other two algorithms. Besides, the F1 value of the proposed algorithm is obviously higher than that the other two algorithms. As a result, it is concluded that compared with the CNN algorithm and the PCNN algorithm, the proposed algorithm is more effective in the retinal vascular segmentation of color fundus images; the capillaries have better connectivity, and the proposed algorithm can improve the segmentation Ac while reducing the operation time.

This study aimed to detect and diagnose the lung nodules as early as possible to effectively treat them, thereby reducing the burden on the medical system and patients. A lung computed tomography (CT) image segmentation algorithm was constructed based on the deep learning convolutional neural network (CNN). The clinical data of 69 patients with lung nodules diagnosed by needle biopsy and pathological comprehensive diagnosis at hospital were collected for specific analysis. The CT image segmentation algorithm was used to distinguish the nature and volume of lung nodules and compared with other computer aided design (CAD) software (Philips ISP). 69 patients with lung nodules were treated by radiofrequency ablation (RFA). The results showed that the diagnostic sensitivity of the CT image segmentation algorithm based on the CNN was obviously higher than that of the Philips ISP for solid nodules <5 mm (63 cases vs. 33 cases) (P < 0.05); it was the same result for the subsolid nodule <5 mm (33 case vs. 5 cases) (P < 0.05) that was slightly higher for solid and subsolid nodules with a diameter of 5–10 mm (37 cases vs. 28 cases) (P < 0.05). In addition, the CNN algorithm can reach all detection for calcified nodules and pleural nodules (7 cases; 5 cases), and the diagnostic sensitivities were much better than those of Philips ISP (2 cases; 3 cases) (P < 0.05). Patients with pulmonary nodules treated by RFA were in good postoperative condition, with a half-year survival rate of 100% and a one-year survival rate of 72.4%. Therefore, it could be concluded that the CT image segmentation algorithm based on the CNN could effectively detect and diagnose the lung nodules early, and the RFA could effectively treat the lung nodules.

Dental disease is a significant problem in humans and deep learning is increasingly being used in the field of dentistry. The purpose of this literature review is to identify dental problems such as tooth identification, caries, treated teeth, dental implants, and endodontic treatment using deep learning approaches in dental image analysis which help dentists in their decision-making process. Dental radiographs are essential for the diagnosis and detection of dental issues. The study focuses on the development and use of several image segmentation/ classification algorithms in the extraction of regions of interest from dental radiographs. To predict different forms of impacted teeth, a convolutional neural network is trained, validated, and tested using dental images with labelled images datasets. Our research suggests that Hybrid models such as CNN-SVM, CNN-KNN or CNN-LSTM or K-mean can be trained over mixed data sets to produce excellent results whereas compared to other image segmentation algorithms, UNet architecture performs better at segmenting dental Xray images.

Computer-aided diagnostics of cancer pathologies based on histological image segmentation is a promising area in the field of computer vision and machine learning. To date, the successes of neural networks in image segmentation in a number of tasks are comparable to human results and can even exceed them. The paper presents a fast algorithm of histological image segmentation based on the convolutional neural network U-Net. Using this approach allows to get better results in the tasks of medical image segmentation. The developed algorithm based on neural network AlexNet was used for the creation of the automatic markup of the histological image database. The neural network algorithms were trained and tested on the NVIDIA DGX-1 supercomputer using histological images. The results of the research show that the fast algorithm based on neural network U-Net can be successfully used for the histological image segmentation in real medical practice, which is confirmed by the high level of similarity of the obtained markup with the expert one.