Search

Convolutional Neural Networks (CNNs) have achieved state-of-the-art performance for medical image segmentation, yet need plenty of manual annotations for training. Semi-Supervised Learning (SSL) methods are promising to reduce the requirement of annotations, but their performance is still limited when the dataset size and the number of annotated images are small. Leveraging existing annotated datasets with similar anatomical structures to assist training has a potential for improving the model's performance. However, it is further challenged by the cross-anatomy domain shift due to the different appearance and even imaging modalities from the target structure. To solve this problem, we propose Contrastive Semi-supervised learning for Cross Anatomy Domain Adaptation (CS-CADA) that adapts a model to segment similar structures in a target domain, which requires only limited annotations in the target domain by leveraging a set of existing annotated images of similar structures in a source domain. We use Domain-Specific Batch Normalization (DSBN) to individually normalize feature maps for the two anatomical domains, and propose a cross-domain contrastive learning strategy to encourage extracting domain invariant features. They are integrated into a Self-Ensembling Mean-Teacher (SE-MT) framework to exploit unlabeled target domain images with a prediction consistency constraint. Extensive experiments show that our CS-CADA is able to solve the challenging cross-anatomy domain shift problem, achieving accurate segmentation of coronary arteries in X-ray images with the help of retinal vessel images and cardiac MR images with the help of fundus images, respectively, given only a small number of annotations in the target domain., Comment: 12 pages, 6 figures

Accurate segmentation of Organs-at-Risk (OAR) from Head and Neck (HAN) Computed Tomography (CT) images with uncertainty information is critical for effective planning of radiation therapy for Nasopharyngeal Carcinoma (NPC) treatment. Despite the state-of-the-art performance achieved by Convolutional Neural Networks (CNNs) for the segmentation task, existing methods do not provide uncertainty estimation of the segmentation results for treatment planning, and their accuracy is still limited by the low contrast of soft tissues in CT, highly imbalanced sizes of OARs and large inter-slice spacing. To address these problems, we propose a novel framework for accurate OAR segmentation with reliable uncertainty estimation. First, we propose a Segmental Linear Function (SLF) to transform the intensity of CT images to make multiple organs more distinguishable than existing simple window width/level-based methods. Second, we introduce a novel 2.5D network (named as 3D-SepNet) specially designed for dealing with clinic CT scans with anisotropic spacing. Thirdly, we propose a novel hardness-aware loss function that pays attention to hard voxels for accurate segmentation. We also use an ensemble of models trained with different loss functions and intensity transforms to obtain robust results, which also leads to segmentation uncertainty without extra efforts. Our method won the third place of the HAN OAR segmentation task in StructSeg 2019 challenge and it achieved weighted average Dice of 80.52% and 95 % Hausdorff Distance of 3.043 mm. Experimental results show that 1) our SLF for intensity transform helps to improve the accuracy of OAR segmentation from CT images; 2) With only 1/3 parameters of 3D UNet, our 3D-SepNet obtains better segmentation results for most OARs; 3) The proposed hard voxel weighting strategy used for training effectively improves the segmentation accuracy; 4) The segmentation uncertainty obtained by our method has a high correlation to mis-segmentations, which has a potential to assist more informed decisions in clinical practice. Our code is available at https://github.com/HiLab-git/SepNet .

We aimed to investigate the associations of serum alkaline phosphatase (ALP) levels with incident cardiovascular disease (CVD), coronary heart disease (CHD), and stroke, as well as their subtypes, among men and women in a prospective cohort study.A total of 11,408 men and 14,981 women were included to evaluate the associations between ALP levels and incident CVD. Participants were divided into four groups according to the quartiles of serum ALP levels in men and women separately. Cox proportional hazard models were used to estimate the hazard ratios (HRs) and 95% confidence intervals (CIs).During an average follow-up of 7.3 years, 7,015 incident CVDs (5,561 CHDs and 1,454 strokes) were documented. After adjustments for age, body mass index, smoking status, drinking status, diabetes, hyperlipidemia, hypertension, physical activity, aspirin usage, anticoagulants usage, menopausal status (women only), family history of CVD, estimated glomerular filtration rate, white blood cell counts, and admission batch and comparing the lowest quartile of ALP, the adjusted HRs (95% CIs) of participants in the highest quartile were 1.22 (1.11-1.34) for CVD, 1.14 (1.02-1.28) for CHD, 1.43 (1.18-1.73) for stroke, 1.31 (1.09-1.57) for acute coronary syndrome (ACS), 1.37 (1.11-1.70) for ischemic stroke, and 1.75 (1.10-2.79) for hemorrhagic stroke in men and 1.12 (1.01-1.23) for CVD, 1.10 (0.99-1.23) for CHD, 1.18 (0.92-1.51) for stroke, 1.23 (1.03-1.47) for ACS, 1.10 (0.83-1.45) for ischemic stroke, and 1.54 (0.90-2.65) for hemorrhagic stroke in women. The ALP-CVD associations remained significant even within the normal ranges of ALP levels (40-150 U/L). Moreover, linear dose-response relationships were found between ALP levels and incident CVD.Higher ALP levels, even within the normal range, were significantly associated with increased risks of CVD, in a dose-dependent manner. These findings suggested that regular monitoring of ALP levels may help in improving the early identification of the population at higher CVD risk.

Aim The aim of the present study was to investigate the associations of baseline and longitudinal changes in leukocyte counts with incident cardiovascular disease (CVD). Methods We conducted a prospective study to investigate the associations of baseline and 5-year changes in leukocyte counts with incident CVD and its subtypes in middle-aged and elderly Chinese. We estimated the hazard ratios (HRs) and 95% confidence intervals (CIs) for CVD using the Cox proportional-hazards models. Results In the analyses of baseline total leukocyte count of 26,655 participants, compared with the lowest quartile (＜4.71×109/L), participants in the fourth quartile (＞6.70×109/L) had 11% higher risk for CVD. Consistent with total leukocyte count, neutrophil count also exhibited a significant positive association with the risk of CVD. In the analyses of 5-year changes in total leukocyte count of 11,594 participants, the changes in leukocyte count were categorized into three groups, i.e., the decreased group (＜25%), stable group (25%-75%), and increased group (＞75%). Compared with participants in the stable group (-1.18 to 0.44×109/L), participants in the increased group (＞0.44×109/L) had 14% higher risk for CVD. We also observed significant positive associations of the changes in neutrophil and monocyte counts with the risk of CVD. Furthermore, the total leukocyte count in the second or third tertile at the first follow-up with a 5-year increase was related to higher CVD risk. Conclusion High baseline total leukocyte count and a 5-year increase in total leukocyte count were related to higher CVD risk.

PURPOSE Radiotherapy is one of the main treatments of nasopharyngeal cancer (NPC) and lung cancer. Accurate segmentation of organs at risks (OARs) in CT images is a key step in radiotherapy planning for NPC and lung cancer. However, the segmentation of OARs is influenced by the highly imbalanced size of organs, which often results in very poor segmentation results for small and difficult-to-segment organs. In addition, the complex morphological changes and fuzzy boundaries of OARs also pose great challenges to the segmentation task. In this paper, we propose a cross-layer attention fusion network (CLAF-CNN) to solve the problem of accurately segmenting OARs. METHODS In CLAF-CNN, we integrate the spatial attention maps of the adjacent spatial attention modules to make the segmentation targets more accurately focused, so that the network can capture more target-related features. In this way, the spatial attention modules in the network can be learned and optimized together. In addition, we introduce a new Top-K exponential logarithmic Dice loss (TELD-Loss) to solve the imbalance problem in OAR segmentation. The TELD-Loss further introduces a Top-K optimization mechanism based on Dice loss and exponential logarithmic loss, which makes the network pay more attention to small organs and difficult-to-segment organs, so as to enhance the overall performance of the segmentation model. RESULTS We validated our framework on the OAR segmentation datasets of the head and neck and lung CT images in the StructSeg 2019 challenge. Experiments show that the CLAF-CNN outperforms the state-of-the-art attention-based segmentation methods in the OAR segmentation task with average Dice coefficient of 79.65% for head and neck OARs and 88.39% for lung OARs. CONCLUSIONS This work provides a new network named CLAF-CNN which contains cross-layer spatial attention map fusion architecture and TELD-Loss for OAR segmentation. Results demonstrated that the proposed method could obtain accurate segmentation results for OARs, which has a potential of improving the efficiency of radiotherapy planning for nasopharynx cancer and lung cancer.

Radiotherapy is the main treatment method for nasopharynx cancer. Delineation of Gross Target Volume (GTV) from medical images is a prerequisite for radiotherapy. As manual delineation is time-consuming and laborious, automatic segmentation of GTV has a potential to improve the efficiency of this process. This work aims to automatically segment GTV of nasopharynx cancer from Computed Tomography (CT) images. However, it is challenged by the small target region, anisotropic resolution of clinical CT images, and the low contrast between the target region and surrounding soft tissues. To deal with these problems, we propose a 2.5D Convolutional Neural Network (CNN) to handle the different in-plane and through-plane resolutions. We also propose a spatial attention module to enable the network to focus on the small target, and use channel attention to further improve the segmentation performance. Moreover, we use a multi-scale sampling method for training so that the networks can learn features at different scales, which are combined with a multi-model ensemble method to improve the robustness of segmentation results. We also estimate the uncertainty of segmentation results based on our model ensemble, which is of great importance for indicating the reliability of automatic segmentation results for radiotherapy planning. Experiments with 2019 MICCAI StructSeg dataset showed that (1) Our proposed 2.5D network has a better performance on images with anisotropic resolution than the commonly used 3D networks. (2) Our attention mechanism can make the network pay more attention to the small GTV region and improve the segmentation accuracy. (3) The proposed multi-scale model ensemble achieves more robust results, and it can simultaneously obtain uncertainty information that can indicate potential mis-segmentations for better clinical decisions.

Domain generalizable model is attracting increasing attention in medical image analysis since data is commonly acquired from different institutes with various imaging protocols and scanners. To tackle this challenging domain generalization problem, we propose a Domain Composition and Attention-based network (DCA-Net) to improve the ability of domain representation and generalization. First, we present a domain composition method that represents one certain domain by a linear combination of a set of basis representations (i.e., a representation bank). Second, a novel plug-and-play parallel domain preceptor is proposed to learn these basis representations and we introduce a divergence constraint function to encourage the basis representations are as divergent as possible. Then, a domain attention module is proposed to learn the linear combination coefficients of the basis representations. The result of liner combination is used to calibrate the feature maps of an input image, which enables the model to generalize to different and even unseen domains. We validate our method on public prostate MRI dataset acquired from six different institutions with apparent domain shift. Experimental results show that our proposed model can generalizes well on different and even unseen domains and it outperforms state-of-the-art methods on the multi-domain prostate segmentation task. Code is available at https://github.com/HiLab-git/DCA-Net.

BackgroundTests to identify reversible airflow limitation are important in asthma diagnosis, but they are time-consuming and may be difficult for patients to cooperate. We aim to evaluate the predictive value of fractional exhaled nitric oxide (FeNO) and blood eosinophil (B-Eos) count in asthma diagnosis, and to distinguish patients who could avoid reversibility testing.MethodsWe screened 7463 suspected asthma cases between January 2014 and December 2019 in Chongqing, China, and identified 2349 patients with complete FeNO, B-Eos count, and spirometry data. Of these, 824 were diagnosed with asthma via a positive bronchial-provocation or bronchodilation test.ResultsWhen FeNO and B-Eos counts were used in combination, the area under the receiver operating characteristic curve (AUC) for diagnosing asthma increased (0.768 vs. 0.745 or 0.728; both P < 0.001). The odds ratio for having asthma increased progressively with a gradual increase in FeNO or B-Eos count (both P < 0.001). Further analysis of in-series combinations of different threshold values for these biomarkers indicated that moderately elevated biomarker levels (FeNO > 40 ppb and B-Eos > 300 cells/μl) support a diagnosis of asthma because diagnostic specificity was > 95% and the positive likelihood ratio (PLR) was > 10. This conclusion was verified when selecting the data from 2017 to 2019 as the verification cohort.ConclusionThe combination of FeNO and B-Eos count can improve diagnostic efficacy for asthma. Patients with moderately elevated biomarkers (FeNO > 40 ppb and B-Eos > 300 cells/μl) could be diagnosed with asthma.

Accurate segmentation of organ at risk (OARs) in head and neck CT images is critical for planning of radiotherapy of the nasopharynx cancer. In segmentation tasks, fully convolutional networks (FCNs) are widely used. Recently, as a kind of attention module, concurrent squeeze and excitation (scSE) blocks in FCNs are proved to have good performance. However, the attention feature maps generated by scSE blocks are isolated from each other, which doesn't help the network noticing the similarities among different feature maps. Consequently, we propose cross-layer spatial attention map fusion network (CSAF-CNN) to fuse different spatial attention maps to solve this problem. In addition, we introduce a top-k exponential logarithmic dice loss (TELD-Loss) in OARs segmentation, which effectively alleviates the serious sample imbalance problem of this task. We evaluate our framework in the head & neck CT scans of nasopharynx cancer patients in StructSeg 2019 challenge. We validate the effectiveness of the proposed method through ablation study, and achieve very competitive results.

The fully convolutional networks (FCNs) have achieved state-of-the-art performance in numerous medical image segmentation tasks. Most FCNs typically focus on fusing features in different levels to improve the learning ability to multi-scale features. In this paper, we explore an alternative direction to improve network performance by enhancing the encoding quality of high- and low-level features, so as to introduce two feature enhancement modules: (i) high-level feature enhancement module (HFE); (ii) low-level feature enhancement module (LFE). HFE utilizes attention mechanism to selectively aggregate the optimal feature information in high- and low-levels, enhancing the ability of high-level features to reconstruct accurate details. LFE aims to use global semantic information of high-level features to adaptively guide feature learning of bottom networks, so as to enhance the semantic consistency of high- and low-level features. We integrate HFE and LFE into a typical encoder-decoder network, and propose a novel medical image segmentation framework (HLF-Net). On two challenging datasets of skin lesion segmentation and spleen segmentation, we prove that the proposed modules and network can improve the performance considerably.

Accurate segmentation of organs-at-risks (OARs) from Computed Tomography (CT) image is a key step for efficient planning of radiation therapy for nasopharyngeal carcinoma (NPC) treatment. Convolutional Neural Networks (CNN) have recently become the state-of-the-art automated OARs image segmentation method. However, due to the low contrast of head and neck organism tissues in CT, the fully automatic segmentation may still need to be refined to become accurate and robust enough for clinical use. We propose a deep learning-based multi-organ interactive segmentation method to improve the results obtained by an automatic CNN and to reduce user interactions during refinement for higher accuracy. We use one CNN to obtain an initial automatic segmentation, on which user interactions are added to indicate mis-segmentation. Another CNN takes as input the user interactions with the initial segmentation and gives a refined result. We propose a dimension separate lightweight network that gives a faster and better dense predictions. In addition, we propose a mis-segmentation-based weighting strategy combined with loss functions to achieve more accurate segmentation. We validated the proposed framework in the context of 3D head and neck organism segmentation from CT images. Experimental results show our method achieves a large improvement from automatic CNNs, and obtains higher accuracy with fewer user interventions and less time compared with traditional interactive segmentation method.

Prospective studies on the relations between platelet count (PLT), mean platelet volume (MPV), platelet distribution width (PDW), and incident cardiovascular disease (CVD) were still limited. This study aimed to investigate the above-mentioned prospective relations in the middle-aged and older Chinese populations based on the Dongfeng-Tongji cohort. We included 31,751 participants who were free of coronary heart disease (CHD), stroke, cancer, or severely abnormal electrocardiogram at baseline. During a median follow-up of 5.9 years, we identified 5,683 incident CVD cases, including 4,423 CHD and 1,260 stroke cases. Cox proportional hazard models were used to estimate the hazard ratios (HRs) and 95% confident intervals (CIs) for the relation analyses. Compared with participants with 146 ≤ PLT ≤ 233 10E9/L, the adjusted HR (95% CI) of those with PLT < 146 10E9/L was 0.80 (0.68 to 0.95) for incident stroke. Compared with participants with 7.3 ≤ MPV ≤ 10.3 fl, the adjusted HRs (95% CIs) of those with MPV < 7.3 fl were 0.81 (0.75 to 0.88), 0.80 (0.73 to 0.88) and 0.84 (0.71 to 1.00) for incident CVD, CHD and stroke, respectively. Compared with participants with 13.2 ≤ PDW ≤ 18.1 %, the adjusted HRs (95% CIs) of those with PDW < 13.2 % were 0.80 (0.73 to 0.87) and 0.78 (0.70 to 0.86) for incident CVD and CHD, respectively. In conclusion, lower levels of PLT and MPV were significantly related to lower risk of stroke, while lower levels of MPV and PDW were significantly related to lower risks of CVD and CHD.

Adjunctive use of corticosteroids with anti-tuberculosis (TB) therapy has been reported to benefit people with tuberculous pleural effusion (TPE), while there is a paucity of data to support it as routine use. TPE can be subdivided into free-flowing type and loculated type. We evaluated the effects of adjunctive prednisolone therapy on functional sequelae, pleural thickening and pleural adhesions in patients with free-flowing TPE.This is a retrospective cohort study, conducted from Jan 2013 to Dec 2016 (ChiCTR-ORC-16009267). All the patients were diagnosed with TPE, and treated with standard 4-drug anti-TB chemotherapy regimen and complete drainage of the effusion. We compared the incidence of a composite of roentgenographic sequelae (pleural thickening of2 mm, or pleural adhesions and costophrenic angle 90°), or restrictive functional sequelae (FVC/FVC pre or TLC/TLC pre 80%) between those who received adjunctive prednisolone therapy and those who did not.The final cohorts consisted of 135 subjects. Of those, 56 received adjunctive prednisolone therapy and 79 did not. The incidence of a composite of roentgenographic sequelae or restrictive functional sequelae was significantly decreased in the prednisolone group as compared with the control group (51.8% vs. 75.9%; RR 2.83, 95% confidence interval, 1.27-6.31, P = 0.011). No serious side effects due to corticosteroid were noted.This study detected a significant association between adjunctive prednisolone therapy and decreased incidence of a composite of radiographic sequelae, or functional sequelae in HIV-negative, free-flowing type TPE patients treated with adjunctive prednisolone.