LuNet-LightGBM:An Effective Hybrid Approach for Lesion Segmentation and DR Grading

Diabetes problems can lead to an eye disease called Diabetic Retinopathy(DR),which permanently damages the blood vessels in the retina.If not treated early,DR becomes a significant reason for blindness.To identify the DR and determine the stages,medical tests are very labor-intensive,expensive,and timeconsuming.To address the issue,a hybrid deep and machine learning techniquebased autonomous diagnostic system is provided in this paper.Our proposal is based on lesion segmentation of the fundus images based on the LuNet network.Then a Refined Attention Pyramid Network(RAPNet)is used for extracting global and local features.To increase the performance of the classifier,the unique features are selected from the extracted feature set using Aquila Optimizer(AO)algorithm.Finally,the LightGBM model is applied to classify the input image based on the severity.Several investigations have been done to analyze the performance of the proposed framework on three publically available datasets(MESSIDOR,APTOS,and IDRiD)using several performance metrics such as accuracy,precision,recall,and f1-score.The proposed classifier achieves 99.29%,99.35%,and 99.31%accuracy for these three datasets respectively.The outcomes of the experiments demonstrate that the suggested technique is effective for disease identification and reliable DR grading.

Skin Lesion Segmentation and Classification Using Conventional and Deep Learning Based Framework

Background:In medical image analysis,the diagnosis of skin lesions remains a challenging task.Skin lesion is a common type of skin cancer that exists worldwide.Dermoscopy is one of the latest technologies used for the diagnosis of skin cancer.Challenges:Many computerized methods have been introduced in the literature to classify skin cancers.However,challenges remain such as imbalanced datasets,low contrast lesions,and the extraction of irrelevant or redundant features.Proposed Work:In this study,a new technique is proposed based on the conventional and deep learning framework.The proposed framework consists of two major tasks:lesion segmentation and classification.In the lesion segmentation task,contrast is initially improved by the fusion of two filtering techniques and then performed a color transformation to color lesion area color discrimination.Subsequently,the best channel is selected and the lesion map is computed,which is further converted into a binary form using a thresholding function.In the lesion classification task,two pre-trained CNN models were modified and trained using transfer learning.Deep features were extracted from both models and fused using canonical correlation analysis.During the fusion process,a few redundant features were also added,lowering classification accuracy.A new technique called maximum entropy score-based selection(MESbS)is proposed as a solution to this issue.The features selected through this approach are fed into a cubic support vector machine(C-SVM)for the final classification.Results:The experimental process was conducted on two datasets:ISIC 2017 and HAM10000.The ISIC 2017 dataset was used for the lesion segmentation task,whereas the HAM10000 dataset was used for the classification task.The achieved accuracy for both datasets was 95.6% and 96.7%, respectively, which was higher thanthe existing techniques.

A Saliency Based Image Fusion Framework for Skin Lesion Segmentation and Classification

Melanoma,due to its higher mortality rate,is considered as one of the most pernicious types of skin cancers,mostly affecting the white populations.It has been reported a number of times and is now widely accepted,that early detection of melanoma increases the chances of the subject’s survival.Computer-aided diagnostic systems help the experts in diagnosing the skin lesion at earlier stages using machine learning techniques.In thiswork,we propose a framework that accurately segments,and later classifies,the lesion using improved image segmentation and fusion methods.The proposed technique takes an image and passes it through two methods simultaneously;one is the weighted visual saliency-based method,and the second is improved HDCT based saliency estimation.The resultant image maps are later fused using the proposed image fusion technique to generate a localized lesion region.The resultant binary image is later mapped back to the RGB image and fed into the Inception-ResNet-V2 pre-trained model-trained by applying transfer learning.The simulation results show improved performance compared to several existing methods.

U-Net Based Dual-Pooling Segmentation of Bone Metastases in Thoracic SPECT Bone Scintigrams

In order to enhance the performance of the CNN-based segmentation models for bone metastases, this study proposes a segmentation method that integrates dual-pooling, DAC, and RMP modules. The network consists of distinct feature encoding and decoding stages, with dual-pooling modules employed in encoding stages to maintain the background information needed for bone scintigrams diagnosis. Both the DAC and RMP modules are utilized in the bottleneck layer to address the multi-scale problem of metastatic lesions. Experimental evaluations on 306 clinical SPECT data have demonstrated that the proposed method showcases a substantial improvement in both DSC and Recall scores by 3.28% and 6.55% compared the baseline. Exhaustive case studies illustrate the superiority of the methodology.

Rethinking the Dice Loss for Deep Learning Lesion Segmentation in Medical Images

Deep learning is widely used for lesion segmentation in medical images due to its breakthrough performance.Loss functions are critical in a deep learning pipeline,and they play important roles in segmenting performance.Dice loss is the most commonly used loss function in medical image segmentation,but it also has some disadvantages.In this paper,we discuss the advantages and disadvantages of the Dice loss function,and group the extensions of the Dice loss according to its improved purpose.The performances of some extensions are compared according to core references.Because different loss functions have different performances in different tasks,automatic loss function selection will be the potential direction in the future.

Automatic and Robust Segmentation of Multiple Sclerosis Lesions with Convolutional Neural Networks

The diagnosis of multiple sclerosis(MS)is based on accurate detection of lesions on magnetic resonance imaging(MRI)which also provides ongoing essential information about the progression and status of the disease.Manual detection of lesions is very time consuming and lacks accuracy.Most of the lesions are difficult to detect manually,especially within the grey matter.This paper proposes a novel and fully automated convolution neural network(CNN)approach to segment lesions.The proposed system consists of two 2D patchwise CNNs which can segment lesions more accurately and robustly.The first CNN network is implemented to segment lesions accurately,and the second network aims to reduce the false positives to increase efficiency.The system consists of two parallel convolutional pathways,where one pathway is concatenated to the second and at the end,the fully connected layer is replaced with CNN.Three routine MRI sequences T1-w,T2-w and FLAIR are used as input to the CNN,where FLAIR is used for segmentation because most lesions on MRI appear as bright regions and T1-w&T2-w are used to reduce MRI artifacts.We evaluated the proposed system on two challenge datasets that are publicly available from MICCAI and ISBI.Quantitative and qualitative evaluation has been performed with various metrics like false positive rate(FPR),true positive rate(TPR)and dice similarities,and were compared to current state-of-the-art methods.The proposed method shows consistent higher precision and sensitivity than other methods.The proposed method can accurately and robustly segment MS lesions from images produced by different MRI scanners,with a precision up to 90%.

Lesion Segmentation and Identification of Breast Tumor on Dynamic Contrast-Enhanced Magnetic Resonance Imaging

Dynamic contrast-enhanced magnetic resonance imaging(DCE-MRI) can show subtle lesion morphology, improve the display of lesion definitions, and objectively reflect the blood supply of breast tumors; it can also reflect different strengthening patterns of normal tissues and lesion areas after medical tracer injection. DCE-MRI has become an important basis for the clinical diagnosis of breast cancer. To DCE-MRI data acquired from several hospitals across multiple provinces, a series of in-silico computational methods were applied for lesion segmentation and identification of breast tumor in this paper. The image segmentation methods include Otsu segmentation of subtraction images, signal-interference-ratio segmentation method and an improved variational level set method,each has its own application scope. After that, the distribution of benign and malignant in lesion region is identified based on three-time-point theory. From the experiment, the analysis of DCE-MRI data of breast tumor can show the distribution of benign and malignant in lesion region, provide a great help for clinicians to diagnose breast cancer more expediently and lay a basis for medical diagnosis and treatment planning.

A network lightweighting method for difficult segmentation of 3D medical images

Currently,deep learning is widely used in medical image segmentation and has achieved good results.However,3D medical image segmentation tasks with diverse lesion characters,blurred edges,and unstable positions require complex networks with a large number of parameters.It is computationally expensive and results in high requirements on equipment,making it hard to deploy the network in hospitals.In this work,we propose a method for network lightweighting and applied it to a 3D CNN based network.We experimented on a COVID-19 lesion segmentation dataset.Specifically,we use three cascaded one-dimensional convolutions to replace a 3D convolution,and integrate instance normalization with the previous layer of one-dimensional convolutions to accelerate network inference.In addition,we simplify test-time augmentation and deep supervision of the network.Experiments show that the lightweight network can reduce the prediction time of each sample and the memory usage by 50%and reduce the number of parameters by 60%compared with the original network.The training time of one epoch is also reduced by 50%with the segmentation accuracy dropped within the acceptable range.

IoMT Enabled Melanoma Detection Using Improved Region Growing Lesion Boundary Extraction

The Internet ofMedical Things(IoMT)and cloud-based healthcare applications,services are beneficial for better decision-making in recent years.Melanoma is a deadly cancer with a highermortality rate than other skin cancer types such as basal cell,squamous cell,andMerkel cell.However,detection and treatment at an early stage can result in a higher chance of survival.The classical methods of detection are expensive and labor-intensive.Also,they rely on a trained practitioner’s level,and the availability of the needed equipment is essential for the early detection of Melanoma.The current improvement in computer-aided systems is providing very encouraging results in terms of precision and effectiveness.In this article,we propose an improved region growing technique for efficient extraction of the lesion boundary.This analysis and detection ofMelanoma are helpful for the expert dermatologist.The CNN features are extracted using the pre-trained VGG-19 deep learning model.In the end,the selected features are classified by SVM.The proposed technique is gauged on openly accessible two datasets ISIC 2017 and PH2.For the evaluation of our proposed framework,qualitative and quantitative experiments are performed.The suggested segmentation method has provided encouraging statistical results of Jaccard index 0.94,accuracy 95.7%on ISIC 2017,and Jaccard index 0.91,accuracy 93.3%on the PH2 dataset.These results are notably better than the results of prevalent methods available on the same datasets.The machine learning SVMclassifier executes significantly well on the suggested feature vector,and the comparative analysis is carried out with existing methods in terms of accuracy.The proposed method detects and classifies melanoma far better than other methods.Besides,our framework gained promising results in both segmentation and classification phases.

PLDMLT:Multi-Task Learning of Diabetic Retinopathy Using the Pixel-Level Labeled Fundus Images

In the field of medical images,pixel-level labels are time-consuming and expensive to acquire,while image-level labels are relatively easier to obtain.Therefore,it makes sense to learn more information(knowledge)from a small number of hard-to-get pixel-level annotated images to apply to different tasks to maximize their usefulness and save time and training costs.In this paper,using Pixel-Level Labeled Images forMulti-Task Learning(PLDMLT),we focus on grading the severity of fundus images for Diabetic Retinopathy(DR).This is because,for the segmentation task,there is a finely labeled mask,while the severity grading task is without classification labels.To this end,we propose a two-stage multi-label learning weakly supervised algorithm,which generates initial classification pseudo labels in the first stage and visualizes heat maps at all levels of severity using Grad-Cam to further provide medical interpretability for the classification task.A multitask model framework with U-net as the baseline is proposed in the second stage.A label update network is designed to alleviate the gradient balance between the classification and segmentation tasks.Extensive experimental results show that our PLDMLTmethod significantly outperforms other stateof-the-art methods in DR segmentation on two public datasets,achieving up to 98.897%segmentation accuracy.In addition,our method achieves comparable competitiveness with single-task fully supervised learning in the DR severity grading task.

Research progress in artificial intelligence assisted diabetic retinopathy diagnosis

Diabetic retinopathy(DR)is one of the most common retinal vascular diseases and one of the main causes of blindness worldwide.Early detection and treatment can effectively delay vision decline and even blindness in patients with DR.In recent years,artificial intelligence(AI)models constructed by machine learning and deep learning(DL)algorithms have been widely used in ophthalmology research,especially in diagnosing and treating ophthalmic diseases,particularly DR.Regarding DR,AI has mainly been used in its diagnosis,grading,and lesion recognition and segmentation,and good research and application results have been achieved.This study summarizes the research progress in AI models based on machine learning and DL algorithms for DR diagnosis and discusses some limitations and challenges in AI research.

Artificial intelligence can assist with diagnosing retinal vein occlusion

AIM:To assist with retinal vein occlusion(RVO)screening,artificial intelligence(AI)methods based on deep learning(DL)have been developed to alleviate the pressure experienced by ophthalmologists and discover and treat RVO as early as possible.METHODS:A total of 8600 color fundus photographs(CFPs)were included for training,validation,and testing of disease recognition models and lesion segmentation models.Four disease recognition and four lesion segmentation models were established and compared.Finally,one disease recognition model and one lesion segmentation model were selected as superior.Additionally,224 CFPs from 130 patients were included as an external test set to determine the abilities of the two selected models.RESULTS:Using the Inception-v3 model for disease identification,the mean sensitivity,specificity,and F1 for the three disease types and normal CFPs were 0.93,0.99,and 0.95,respectively,and the mean area under the curve(AUC)was 0.99.Using the DeepLab-v3 model for lesion segmentation,the mean sensitivity,specificity,and F1 for four lesion types(abnormally dilated and tortuous blood vessels,cotton-wool spots,flame-shaped hemorrhages,and hard exudates)were 0.74,0.97,and 0.83,respectively.CONCLUSION:DL models show good performance when recognizing RVO and identifying lesions using CFPs.Because of the increasing number of RVO patients and increasing demand for trained ophthalmologists,DL models will be helpful for diagnosing RVO early in life and reducing vision impairment.

Lesion region segmentation via weakly supervised learning

Background:Image-based automatic diagnosis of field diseases can help increase crop yields and is of great importance.However,crop lesion regions tend to be scattered and of varying sizes,this along with substantial intraclass variation and small inter-class variation makes segmentation difficult.Methods:We propose a novel end-to-end system that only requires weak supervision of image-level labels for lesion region segmentation.First,a two-branch network is designed for joint disease classification and seed region generation.The generated seed regions are then used as input to the next segmentation stage where we design to use an encoder-decoder network.Different from previous works that use an encoder in the segmentation network,the encoder-decoder network is critical for our system to successfully segment images with small and scattered regions,which is the major challenge in image-based diagnosis of field diseases.We further propose a novel weakly supervised training strategy for the encoder-decoder semantic segmentation network,making use of the extracted seed regions.Results:Experimental results show that our system achieves better lesion region segmentation results than state of the arts.In addition to crop images,our method is also applicable to general scattered object segmentation.We demonstrate this by extending our framework to work on the PASCAL VOC dataset,which achieves comparable performance with the state-of-the-art DSRG(deep seeded region growing)method.Conclusion:Our method not only outperforms state-of-the-art semantic segmentation methods by a large margin for the lesion segmentation task,but also shows its capability to perform well on more general tasks.

Channel Context and Dual-Domain Attention Based U-Net for Skin Lesion Attributes Segmentation

Skin melanoma is one of the most common malignant tumorsoriginating from melanocytes, and the incidence of the Chinese populationis showing a continuous increasing trend. Early and accurate diagnosisof melanoma has great significance for guiding clinical treatment.However, the symptoms of malignant melanoma are not obvious in theearly stage. It is difficult to be diagnosed with human observation. Meanwhile,it is easy to spread due to missed diagnosis. In order to accuratelydiagnose melanoma, end-to-end skin lesion attribute segmentation frameworkis presented in this paper. It is applied to facilitate the digitalizationprocess of attributes segmentation. The framework was improved on theU-Net construction that use the channel context feature fusion modulebetween the encoder and decoder to further merge context information. Adual-domain attention module is proposed to get more effective informationfrom the feature map. It shows that the proposed method effectivelysegments the lesion attributes and achieves good result in the ISIC2018task2 dataset.

An adaptive region growing algorithm for breast masses in mammograms

This study attempted to accurately segment the mammographic masses and distinguish malignant from benign tumors.An adaptive region growing algorithm with hybrid assessment function combined with maximum likelihood analysis and maximum gradient analysis was developed in this paper.In order to accommodate different situations of masses,the likelihood and the edge gradients of segmented masses were weighted adaptively by the use of information entropy.106 benign and 110 malignant tumors were included in this study.We found that the proposed algorithm obtained segmentation contour more accurately and delineated the tumor body as well as tumor peripheral regions covering typical mass boundaries and some spiculation patterns.Then the segmented results were evaluated by the classification accuracy.42 features including age,intensity,shape and texture were extracted from each segmented mass and support vector machine(SVM)was used as a classifier.The classification accuracy was evaluated using the area(A_(z))under the receiver operating characteristic(ROC)curve.It was found that the maximum likelihood analysis achieved an A_(z)value of 0.835,the maximum gradient analysis got an A_(z)value of 0.932 and the hybrid assessment function performed the best classification result where the value of A_(z)was 0.948.In addition,compared with traditional region growing algorithm,our proposed algorithm is more adaptive and provides a better performance for future works.

Clinical features and management of Crohn' s disease in Chinese patients

Background An increasing incidence of Crohn' s disease has been found in China in recent years. Our study has been focused on evaluating the diversity of the clinical manifestations of Crohn' s disease in order to improve early diagnostic accuracy and therapeutic efficacy.Methods Thirty patients with active Crohn's disease were enrolled and their clinical data, including diagnostic and therapeutic results, were analyzed. Endoscopy combined with histological examination of biopsy specimens provided characteristic features of the disease. Transabdominal bowel sonography (TABS) was used for detecting intestinal complications. Nutritional supportive therapy was given to 20 subjects with active cases of the disease.Results Most patients were young adults with a higher proportion of females to males (ratio: 1.14: 1). The disease affects any segment or a combination of segments along with the alimentary tract (from the mouth to the anus). In this study, the colon and small bowel were the major sites involved. Recurrent episodes of abdominal pain in the right lower quadrant and watery diarrhea were the most common symptoms. Granulomas were identifiable in nearly one-third (30.8%) of all biopsy specimens. In moderate cases of the disease, remission was achieved more quickly through the use of oral prednisone therapy than with SASP or 5-ASA. Beneficial effects on the host's nutritional status were observed. Immunosuppressives were used on an individual basis and showed variable therapeutic effects. Sixteen patients had surgery due to intestinal obstruction or failure to respond to drug therapies. Rapid improvement after surgery was reported.Conclusion Endoscopy (with biopsy) and TABS were both crucial procedures for diagnosis. SASP (or 5-ASA) and prednisone were effective as inductive therapies. Azathioprine has demonstrable benefits after induction therapy with prednisone. Surgery, as an alternative treatment, provided another effective choice in selected patients.