A survey of loss function of medical image segmentation algorithms / 生物医学工程学杂志

Medical image segmentation based on deep learning has become a powerful tool in the field of medical image processing. Due to the special nature of medical images, image segmentation algorithms based on deep learning face problems such as sample imbalance, edge blur, false positive, false negative, etc. In view of these problems, researchers mostly improve the network structure, but rarely improve from the unstructured aspect. The loss function is an important part of the segmentation method based on deep learning. The improvement of the loss function can improve the segmentation effect of the network from the root, and the loss function is independent of the network structure, which can be used in various network models and segmentation tasks in plug and play. Starting from the difficulties in medical image segmentation, this paper first introduces the loss function and improvement strategies to solve the problems of sample imbalance, edge blur, false positive and false negative. Then the difficulties encountered in the improvement of the current loss function are analyzed. Finally, the future research directions are prospected. This paper provides a reference for the reasonable selection, improvement or innovation of loss function, and guides the direction for the follow-up research of loss function.

Colorectal polyp segmentation method based on fusion of transformer and cross-level phase awareness / 生物医学工程学杂志

In order to address the issues of spatial induction bias and lack of effective representation of global contextual information in colon polyp image segmentation, which lead to the loss of edge details and mis-segmentation of lesion areas, a colon polyp segmentation method that combines Transformer and cross-level phase-awareness is proposed. The method started from the perspective of global feature transformation, and used a hierarchical Transformer encoder to extract semantic information and spatial details of lesion areas layer by layer. Secondly, a phase-aware fusion module (PAFM) was designed to capture cross-level interaction information and effectively aggregate multi-scale contextual information. Thirdly, a position oriented functional module (POF) was designed to effectively integrate global and local feature information, fill in semantic gaps, and suppress background noise. Fourthly, a residual axis reverse attention module (RA-IA) was used to improve the network's ability to recognize edge pixels. The proposed method was experimentally tested on public datasets CVC-ClinicDB, Kvasir, CVC-ColonDB, and EITS, with Dice similarity coefficients of 94.04%, 92.04%, 80.78%, and 76.80%, respectively, and mean intersection over union of 89.31%, 86.81%, 73.55%, and 69.10%, respectively. The simulation experimental results show that the proposed method can effectively segment colon polyp images, providing a new window for the diagnosis of colon polyps.

Segmentation of prostate region in magnetic resonance images based on improved V-Net / 生物医学工程学杂志

Magnetic resonance (MR) imaging is an important tool for prostate cancer diagnosis, and accurate segmentation of MR prostate regions by computer-aided diagnostic techniques is important for the diagnosis of prostate cancer. In this paper, we propose an improved end-to-end three-dimensional image segmentation network using a deep learning approach to the traditional V-Net network (V-Net) network in order to provide more accurate image segmentation results. Firstly, we fused the soft attention mechanism into the traditional V-Net's jump connection, and combined short jump connection and small convolutional kernel to further improve the network segmentation accuracy. Then the prostate region was segmented using the Prostate MR Image Segmentation 2012 (PROMISE 12) challenge dataset, and the model was evaluated using the dice similarity coefficient (DSC) and Hausdorff distance (HD). The DSC and HD values of the segmented model could reach 0.903 and 3.912 mm, respectively. The experimental results show that the algorithm in this paper can provide more accurate three-dimensional segmentation results, which can accurately and efficiently segment prostate MR images and provide a reliable basis for clinical diagnosis and treatment.

A meta-learning based method for segmentation of few-shot magnetic resonance images / 生物医学工程学杂志

When applying deep learning algorithms to magnetic resonance (MR) image segmentation, a large number of annotated images are required as data support. However, the specificity of MR images makes it difficult and costly to acquire large amounts of annotated image data. To reduce the dependence of MR image segmentation on a large amount of annotated data, this paper proposes a meta-learning U-shaped network (Meta-UNet) for few-shot MR image segmentation. Meta-UNet can use a small amount of annotated image data to complete the task of MR image segmentation and obtain good segmentation results. Meta-UNet improves U-Net by introducing dilated convolution, which can increase the receptive field of the model to improve the sensitivity to targets of different scales. We introduce the attention mechanism to improve the adaptability of the model to different scales. We introduce the meta-learning mechanism, and employ a composite loss function for well-supervised and effective bootstrapping of model training. We use the proposed Meta-UNet model to train on different segmentation tasks, and then use the trained model to evaluate on a new segmentation task, where the Meta-UNet model achieves high-precision segmentation of target images. Meta-UNet has a certain improvement in mean Dice similarity coefficient (DSC) compared with voxel morph network (VoxelMorph), data augmentation using learned transformations (DataAug) and label transfer network (LT-Net). Experiments show that the proposed method can effectively perform MR image segmentation using a small number of samples. It provides a reliable aid for clinical diagnosis and treatment.

An Intelligent Method for Optimizing Large-volume Medical Image Transmission / 中国医疗器械杂志

In order to alleviate the conflict between medical supply and demand, and to improve the efficiency of medical image transmission, this study proposes an intelligent method for large-volume medical image transmission. This method extracts and generates keyword pairs by analyzing medical diagnostic reports, and uses a 3D-UNet to segment original image data into various sub-area based on its anatomy structure. Then, the sub-areas are scored through keyword pairs and preset scoring criteria, and transmitted to user frontend in the order of prioritization score. Experiments show that this method can fulfill physicians' requirements of radiology reading and diagnosis with only ten percent of data transmitted, which efficiently optimized traditional transmission procedures.

Encephalic ventricular volumetry in multislice computed tomography images in adults with normal cognitive functions

Introducción: Debido a la necesidad de un diagnóstico precoz de los trastornos neurodegenerativos, se ha intentado armonizar los criterios diagnósticos mediante métodos morfométricos basados en técnicas de neuroimagen, pero aún no se han obtenido resultados concluyentes. Objetivo: Determinar el volumen ventricular debido a su amplio uso como marcador de atrofia cerebral e identificar el efecto del sexo sobre estas estructuras, según el tipo de cráneo, estimado a partir de técnicas de imagen de tomografía computarizada multicorte. Métodos: Se desarrolló un estudio observacional y descriptivo en 30 sujetos con funciones neurocognitivas y exploración neuropsiquiátrica normales, con edades comprendidas entre 45 y 54 años, a los que se les realizó una tomografía computarizada multicorte simple de cráneo. Se utilizó un método de segmentación de imágenes basado en la homogeneidad. Resultados: Los volúmenes ventriculares mostraron una correlación significativa y positiva entre ellos, excepto entre el tercer y cuarto ventrículo y el tercero y el volumen ventricular derecho. Los estadísticos del modelo lineal multivariante aplicado mostraron que sólo eran significativos en función del sexo y del tipo de cráneo. No se encontraron diferencias significativas con respecto al sexo en ningún volumen, excepto en el tercer ventrículo (p= 0,01). Lo mismo ocurrió por tipo de cráneo (p= 0,005). Conclusiones: El método de morfometría del sistema ventricular encefálico a partir de imágenes de Tomografía Computarizada / Segmentación por homogeneidad, permitió cuantificar los cambios volumétricos cerebrales asociados al envejecimiento normal y puede ser utilizado como biomarcador de la relación entre la estructura cerebral y las funciones cognitivas.

Introduction: Due to the need for an early diagnosis of neurodegenerative disorders, attempts have been made to harmonize diagnostic criteria using morphometric methods based on neuroimaging techniques, but conclusive results have not yet been obtained. Objective: To determine the ventricular volume due to its wide use as a marker of cerebral atrophy and to identify the effect of sex on these structures, according to the type of skull, estimated from multislice computed tomography imaging techniques. Methods: An observational and descriptive study was developed in 30 subjects with normal neurocognitive functions and neuropsychiatric examination, aged between 45 and 54 years, who underwent a simple multislice CT scan of the skull. An image segmentation method based on homogeneity was used. Results: The ventricular volumes showed a significant and positive correlation between them, except between the third and fourth ventricles and the third and the right ventricular volume. The statistics in the multivariate linear model applied showed that they were only significant in terms of sex and type of skull. No significant differences were found regarding sex in any volume except in the third ventricle (p= 0.01). The same occurred by type of skull (p= 0.005). Conclusions: The morphometry method of the encephalic ventricular system from Computed Tomography images / Segmentation by homogeneity, allowed to quantify the cerebral volumetric changes associated with normal aging and can be used as a biomarker of the relationship between brain structure and cognitive functions.

Automatic segmentation of organs at risk in head and neck carcinoma from radiation therapy using multi-scale fusion and attention based mechanisms / 中华放射肿瘤学杂志

Objective:To develop a multi-scale fusion and attention mechanism based image automatic segmentation method of organs at risk (OAR) from head and neck carcinoma radiotherapy.Methods:We proposed a new OAR segmentation method for medical images of heads and necks based on the U-Net convolution neural network. Spatial and channel squeeze excitation (csSE) attention block were combined with the U-Net, aiming to enhance the feature expression ability. We also proposed a multi-scale block in the U-Net encoding stage to supplement characteristic information. Dice similarity coefficient (DSC) and 95% Hausdorff distance (HD) were used as evaluation criteria for deep learning performance.Results:The segmentation of 22 OAR in the head and neck was performed according to the medical image computing computer assisted intervention (MICCAI) StructSeg2019 dataset. The proposed method improved the average segmentation accuracy by 3%-6% compared with existing methods. The average DSC in the segmentation of 22 OAR in the head and neck was 78.90% and the average 95%HD was 6.23 mm.Conclusion:Automatic segmentation of OAR from the head and neck CT using multi-scale fusion and attention mechanism achieves high segmentation accuracy, which is promising for enhancing the accuracy and efficiency of radiotherapy in clinical practice.

Research on con-beam CT images segmentation method / 中华放射医学与防护杂志

Image-guided radiation therapy (IGRT) is a visual image-guided radiotherapy technique that has many advantages such as increasing the dose of tumor target area and reducing the dose of normal organ exposure. Cone beam CT (CBCT) is one of the most commonly used medical images in IGRT, and the rapid and accurate targeting of CBCT and the segmentation of dangerous organs are of great significance for radiotherapy. The current research method mainly includes partitioning method based on registration and segmentation method based on deep learning. This study reviews the CBCT image segmentation method, existing problems and development directions.

Laplace-Based Interpolation Method in Reduction of Metal Artifact in Computed Tomography Imaging

@#Introduction: Metal artifacts can degrade the image quality of computed tomography (CT) images which lead to errors in diagnosis. This study aims to evaluate the performance of Laplace interpolation (LI) method for metal artifacts reduction (MAR) in CT images in comparison with cubic spline (CS) interpolation. Methods: In this study, the proposed MAR algorithm was developed using MATLAB platform. Firstly, the virtual sinogram was acquired from CT image using Radon transform function. Then, dual-adaptive thresholding detected and segmented the metal part within the CT sinogram. Performance of the two interpolation methods to replace the missing part of segmented sinogram were evaluated. The interpolated sinogram was reconstructed, prior to image fusion to obtain the final corrected image. The qualitative and quantitative evaluations were performed on the corrected CT images (both phantom and clinical images) to evaluate the effectiveness of the proposed MAR technique. Results: From the findings, LI method had successfully replaced the missing data on both simple and complex thresholded sinogram as compared to CS method (p-value = 0.17). The artifact index was significantly reduced by LI method (p-value = 0.02). For qualitative analysis, the mean scores by radiologists for LI-corrected images were higher than original image and CS-corrected images. Conclusion: In conclusion, LI method for MAR produced better results as compared to CS interpolation method, as it worked more effective by successfully interpolated all the missing data within sinogram in most of the CT images.

A generative adversarial network-based unsupervised domain adaptation method for magnetic resonance image segmentation / 生物医学工程学杂志

Intelligent medical image segmentation methods have been rapidly developed and applied, while a significant challenge is domain shift. That is, the segmentation performance degrades due to distribution differences between the source domain and the target domain. This paper proposed an unsupervised end-to-end domain adaptation medical image segmentation method based on the generative adversarial network (GAN). A network training and adjustment model was designed, including segmentation and discriminant networks. In the segmentation network, the residual module was used as the basic module to increase feature reusability and reduce model optimization difficulty. Further, it learned cross-domain features at the image feature level with the help of the discriminant network and a combination of segmentation loss with adversarial loss. The discriminant network took the convolutional neural network and used the labels from the source domain, to distinguish whether the segmentation result of the generated network is from the source domain or the target domain. The whole training process was unsupervised. The proposed method was tested with experiments on a public dataset of knee magnetic resonance (MR) images and the clinical dataset from our cooperative hospital. With our method, the mean Dice similarity coefficient (DSC) of segmentation results increased by 2.52% and 6.10% to the classical feature level and image level domain adaptive method. The proposed method effectively improves the domain adaptive ability of the segmentation method, significantly improves the segmentation accuracy of the tibia and femur, and can better solve the domain transfer problem in MR image segmentation.

Image segmentation of skin lesions based on dense atrous spatial pyramid pooling and attention mechanism / 生物医学工程学杂志

The skin is the largest organ of the human body, and many visceral diseases will be directly reflected on the skin, so it is of great clinical significance to accurately segment the skin lesion images. To address the characteristics of complex color, blurred boundaries, and uneven scale information, a skin lesion image segmentation method based on dense atrous spatial pyramid pooling (DenseASPP) and attention mechanism is proposed. The method is based on the U-shaped network (U-Net). Firstly, a new encoder is redesigned to replace the ordinary convolutional stacking with a large number of residual connections, which can effectively retain key features even after expanding the network depth. Secondly, channel attention is fused with spatial attention, and residual connections are added so that the network can adaptively learn channel and spatial features of images. Finally, the DenseASPP module is introduced and redesigned to expand the perceptual field size and obtain multi-scale feature information. The algorithm proposed in this paper has obtained satisfactory results in the official public dataset of the International Skin Imaging Collaboration (ISIC 2016). The mean Intersection over Union (mIOU), sensitivity (SE), precision (PC), accuracy (ACC), and Dice coefficient (Dice) are 0.901 8, 0.945 9, 0.948 7, 0.968 1, 0.947 3, respectively. The experimental results demonstrate that the method in this paper can improve the segmentation effect of skin lesion images, and is expected to provide an auxiliary diagnosis for professional dermatologists.

Segmentation of core infarct in acute ischemic stroke in diffusion weighted imaging using cascaded VB-Net / 中华放射学杂志

Objective:To explore the detection and segmentation of ischemic core infarct volume of the acute stroke in diffusion weighted imaging (DWI) images using cascaded VB-Net.Methods:MRI data of 1 500 patients (2 456 lesions) with acute ischemic stroke in Henan Provincial People′s Hospital from December 2016 to December 2018 were retrospectively analyzed. Firstly, manual segmentation of ischemic core was performed on DWI images (b=1 000 s/mm 2), and then all data were divided into training set, validation set and independent test set by 8∶1∶1. Then, the cascaded VB-Net was constructed, and the core infarct was automatically detected and segmented in the test set. Interclass correlation coefficient (ICC) was used to evaluate the consistency of volume size measured by manual segmentation and cascaded VB-Net. The patients were divided into large ischemic core lesion group (ischemic core volume ≥10 ml) and small ischemic core lesion group (ischemic core volume<10 ml), and the Dice coefficient difference between the two groups was compared using Mann-Whitney U test. Results:In independent test set, cascaded model had the detection rate of 94.6% (243/257) with Dice coefficient of 0.76 (0.68, 0.84). The agreement of cacade VB-Net segmented [4.19(1.21,14.13)ml] and manual segmented ischemic core infarct volume [4.08(1.19,17.92)ml] was high (ICC=0.97, P<0.001). There was no significant difference in Dice coefficient between large and small lesion groups [0.76 (0.69, 0.85), 0.76 (0.67, 0.84), Z=-0.44, P=0.657]. Conclusions:The cascaded VB-Net model provided a tool to realize automatic detection, segmentation, and calculation of ischemic core infarct volume. It has good segmentation accuracy and high consistency with manual segmentation, which can provide an auxiliary decision-making tool for the selection of treatment plans.

Infrared Imaging Meibomian Gland Segmentation System Based on Deep Learning / 中国医疗器械杂志

In order to better assist doctors in the diagnosis of dry eye and improve the ability of ophthalmologists to recognize the condition of meibomian gland, a meibomian gland image segmentation and enhancement method based on Mobile-U-Net network was proposed. Firstly, Mobile-Net is used as the coding part of U-Net for down sampling, and then features are extracted and fused with the features in decoder to guide image segmentation. Secondly, the segmentation of meibomian gland region is enhanced to assist doctors to judge the condition. Thirdly, a large number of meibomian gland images are collected to train and verify the semantic segmentation network, and the clarity evaluation index is used to verify the meibomian gland enhancement effect. The experimental results show that the similarity coefficient of the proposed method is stable at 92.71%, and the image clarity index is better than the similar dry eye detection instruments on the market.

White blood segmentation based on dual path and atrous spatial pyramid pooling / 生物医学工程学杂志

The count and recognition of white blood cells in blood smear images play an important role in the diagnosis of blood diseases including leukemia. Traditional manual test results are easily disturbed by many factors. It is necessary to develop an automatic leukocyte analysis system to provide doctors with auxiliary diagnosis, and blood leukocyte segmentation is the basis of automatic analysis. In this paper, we improved the U-Net model and proposed a segmentation algorithm of leukocyte image based on dual path and atrous spatial pyramid pooling. Firstly, the dual path network was introduced into the feature encoder to extract multi-scale leukocyte features, and the atrous spatial pyramid pooling was used to enhance the feature extraction ability of the network. Then the feature decoder composed of convolution and deconvolution was used to restore the segmented target to the original image size to realize the pixel level segmentation of blood leukocytes. Finally, qualitative and quantitative experiments were carried out on three leukocyte data sets to verify the effectiveness of the algorithm. The results showed that compared with other representative algorithms, the proposed blood leukocyte segmentation algorithm had better segmentation results, and the mIoU value could reach more than 0.97. It is hoped that the method could be conducive to the automatic auxiliary diagnosis of blood diseases in the future.

Research on segmentation of pterygium lesions based on convolutional neural networks / 国际眼科杂志(Guoji Yanke Zazhi)

@#AIM: To study the precise segmentation of pterygium lesions using the convolutional neural networks from artificial intelligence.<p>METHODS: The network structure of Phase-fusion PSPNet for the segmentation of pterygium lesions is proposed based on the PSPNet model structure. In our network, the up-sampling module is connected behind the pyramid pooling module, which gradually increase the sampling based on the principle of phased increase. Therefore, the information loss is reduced, it is suitable for segmentation tasks with fuzzy edges. The experiments conducted on the dataset provided by the Affiliated Eye Hospital of Nanjing Medical University, which includes 517 ocular surface photographic images of pterygium were divided into training set(330 images), validation set(37 images)and test set(150 images), which the training set and the validation set images are used for training, and the test set images are only used for testing. Comparing results of intelligent segmentation and expert annotation of pterygium lesions.<p>RESULTS: Phase-fusion PSPNet network structure for pterygium mean intersection over union(MIOU)and mean average precision(MPA)were 86.31% and 91.91%, respectively, and pterygium intersection over union(IOU)and average precision(PA)were 77.64% and 86.10%, respectively.<p>CONCLUSION: Convolutional neural networks can segment pterygium lesions with high precision, which is helpful to provide an important reference for doctors' further diagnosis of disease and surgical recommendations, and can also visualize the pterygium intelligent diagnosis.

Research on segmentation of pterygium lesions based on convolutional neural networks / 国际眼科杂志(Guoji Yanke Zazhi)

@#AIM: To study the precise segmentation of pterygium lesions using the convolutional neural networks from artificial intelligence.<p>METHODS: The network structure of Phase-fusion PSPNet for the segmentation of pterygium lesions is proposed based on the PSPNet model structure. In our network, the up-sampling module is connected behind the pyramid pooling module, which gradually increase the sampling based on the principle of phased increase. Therefore, the information loss is reduced, it is suitable for segmentation tasks with fuzzy edges. The experiments conducted on the dataset provided by the Affiliated Eye Hospital of Nanjing Medical University, which includes 517 ocular surface photographic images of pterygium were divided into training set(330 images), validation set(37 images)and test set(150 images), which the training set and the validation set images are used for training, and the test set images are only used for testing. Comparing results of intelligent segmentation and expert annotation of pterygium lesions.<p>RESULTS: Phase-fusion PSPNet network structure for pterygium mean intersection over union(MIOU)and mean average precision(MPA)were 86.31% and 91.91%, respectively, and pterygium intersection over union(IOU)and average precision(PA)were 77.64% and 86.10%, respectively.<p>CONCLUSION: Convolutional neural networks can segment pterygium lesions with high precision, which is helpful to provide an important reference for doctors' further diagnosis of disease and surgical recommendations, and can also visualize the pterygium intelligent diagnosis.

A computed tomography image segmentation algorithm for improving the diagnostic accuracy of rectal cancer based on U-net and residual block / 生物医学工程学杂志

As an important basis for lesion determination and diagnosis, medical image segmentation has become one of the most important and hot research fields in the biomedical field, among which medical image segmentation algorithms based on full convolutional neural network and U-Net neural network have attracted more and more attention by researchers. At present, there are few reports on the application of medical image segmentation algorithms in the diagnosis of rectal cancer, and the accuracy of the segmentation results of rectal cancer is not high. In this paper, a convolutional network model of encoding and decoding combined with image clipping and pre-processing is proposed. On the basis of U-Net, this model replaced the traditional convolution block with the residual block, which effectively avoided the problem of gradient disappearance. In addition, the image enlargement method is also used to improve the generalization ability of the model. The test results on the data set provided by the "Teddy Cup" Data Mining Challenge showed that the residual block-based improved U-Net model proposed in this paper, combined with image clipping and preprocessing, could greatly improve the segmentation accuracy of rectal cancer, and the Dice coefficient obtained reached 0.97 on the verification set.

Application of artificial intelligence in the target delineation of radiotherapy / 国际肿瘤学杂志

Artificial intelligence is the use of computer algorithms to copy or simulate human behavior, giving machines human-like ability. With the rapid development of radiotherapy technology, artificial intelligence has great potential value in all stages of radiotherapy. Image segmentation is the premise of target delineation using artificial intelligence. The commonly used methods in clinic mainly include automatic segmentation based on deep learning and atlas library. The technology of artificial intelligence in organs at risk delineation is relatively mature, which can significantly shorten the delineation time and improve the efficiency. The delineation of tumor targets has achieved some success, the accuracy still needs to be further improved. Artificial intelligence technology makes the target delineation more and more efficient, and the consistency and repeatability have been significantly improved. It is expected to provide more accurate and individualized treatment for patients.

Application of MRI segmentation in tumor radiotherapy / 中华放射肿瘤学杂志

Magnetic resonance imaging (MRI) is a technology with no radiation and high resolution of soft tissues. Therefore, MRI-guided radiotherapy has become a hot spot in the field of radiotherapy. It is of great importance to accurately delineate the targets in radiation oncology. Currently, the delineation of targets is mostly completed by manual segmentation, which is time-consuming, subjective and inconsistent. Automatic segmentation can improve the efficiency and consistency without sacrificing the accuracy of segmentation. In this article, the automatic segmentation methods of MRI applied in radiotherapy were reviewed. The goals, challenges and methods of automatic segmentation for different radiotherapy sites including prostate, nasopharyngeal carcinoma, brain tumors and other organs were analyzed and discussed.

Research progress of MR imaging for prediction of CT imaging / 中国辐射卫生

Medical images can provide clinicans with accurate and comprehensive patients’ information. Morphological or functional abnormalities caused by various diseases can be manifested in many aspects. Although MR images and CT images can highlight the medical image data of different tissue structures of patients, single MR images or CT images cannot fully reflect the complexity of diseases. Using MR image to predict CT image is one of the cross-modal prediction of medical images. In this paper, the methods of MR image prediction for CTmage are classified into four categoriesincluding registration based on atlas, based on image segmentationmethod, based on learning method and based on deep learning method. In our research, we concluded that the method based on deep learning should bemore promoted in the future by compering the existing problems and future development of MR image predicting CT image method.