Modeling annotator preference and stochastic annotation error for medical image segmentation.

Manual annotation of medical images is highly subjective, leading to inevitable annotation biases. Deep learning models may surpass human performance on a variety of tasks, but they may also mimic or amplify these biases. Although we can have multiple annotators and fuse their annotations to reduce stochastic errors, we cannot use this strategy to handle the bias caused by annotators' preferences. In this paper, we highlight the issue of annotator-related biases on medical image segmentation tasks, and propose a Preference-involved Annotation Distribution Learning (PADL) framework to address it from the perspective of modeling an annotator's preference and stochastic errors so as to produce not only a meta segmentation but also the annotator-specific segmentation. Under this framework, a stochastic error modeling (SEM) module estimates the meta segmentation map and average stochastic error map, and a series of human preference modeling (HPM) modules estimate each annotator's segmentation and the corresponding stochastic error. We evaluated our PADL framework on two medical image benchmarks with different imaging modalities, which have been annotated by multiple medical professionals, and achieved promising performance on all five medical image segmentation tasks. Code is available at https://github.com/Merrical/PADL.

Interplay between COVID-19 and Secukinumab treatment in Spondylarthritis patients during the omicron surge: a retrospective cohort study.

The objective of this retrospective cohort study was to assess the relationship between Corona Disease 2019 (COVID-19) and Secukinumab treatment in patients with Spondylarthritis (SpA) in China during the omicron surge. Researchers retrieved 1018 medical records of Secukinumab-treated patients between January 2020 and January 2023 from the West China Hospital of Sichuan University. Out of these, 190 SpA patients from the rheumatology clinic were selected for the study. Guided phone questionnaires were administered by research staff to collect baseline characteristics, SpA disease status, and COVID-19 clinical outcomes. Cohabitants served as the control group and provided COVID-19 related data. Of the 190 potential SpA patients, 122 (66%) completed the questionnaire via phone, along with 259 cohabitants. 84.4% of SpA patients were diagnosed with Ankylosing Spondylitis (AS), and 15.6% were diagnosed with Psoriatic Arthritis (PsA). The rate of SARS-CoV-2 infection was 83.6% in the Secukinumab group and 88.8% in the cohabitants control group, with no significant difference (OR = 0.684, CI 0.366-1.275). One instance of severe COVID-19 was observed in the Secukinumab group, while two were identified in the cohabitants control group. Patients in the Secukinumab group had less time with fever caused by COVID-19 (p = 0.004). Discontinuing Secukinumab after SARS-CoV-2 infection did not significantly affect the course of COVID-19 or worsen SpA status according to our data. Our study suggests that administering Secukinumab to SpA patients does not increase their susceptibility to contracting SARS-CoV-2, and may have a positive effect on the course of SARS-CoV-2 infection.

COVID-19 in Systemic Lupus Erythematosus patients treated with belimumab: a retrospective clinical study.

BACKGROUND: Routine use of immunosuppressive agents in systemic lupus erythematosus (SLE) patients infected with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) potentially increases the risk of adverse outcomes. belimumab, a monoclonal antibody for the treatment of SLE, remains untested for its specific impact on coronavirus disease 2019 (COVID-19) symptoms in these patients. Here, this research investigated the effect of belimumab on COVID-19 symptoms in SLE patients infected with SARS-CoV-2. METHODS: This study enrolled SLE patients who underwent treatment with belimumab. After thorough screening based on the inclusion and exclusion criteria, data pertaining to COVID-19 for both the participants and their cohabitants were obtained through telephone follow-up. The potential impact of belimumab on COVID-19 was evaluated by comparing COVID-19 symptoms and medication use across various groups to investigate the association between belimumab treatment and COVID-19 in SLE. RESULTS: This study involved 123 SLE patients, of whom 89.4% tested positive for SARS-CoV-2. Among cohabitants of SLE patients, the SARS-CoV-2 positive rate was 87.2% (p = 0.543). Patients treated with belimumab exhibited a lower incidence of multiple COVID-19 symptoms than their cohabitating counterparts (p < 0.001). This protective effect was found to be partially related to the time of last belimumab administration. Among those with COVID-19, 30 patients opted to discontinue their anti-SLE drugs, and among them, 53% chose to discontinue belimumab. Discontinuing drugs did not increase the risk of hospitalization due to SARS-CoV-2 infection. CONCLUSION: This study concluded that treatment with belimumab did not increase susceptibility to COVID-19 and beneficially alleviated the symptoms of COVID-19.

Gut microbiota dysbiosis characterized by abnormal elevation of Lactobacillus in patients with immune-mediated necrotizing myopathy.

Aim: The gut microbiota plays an important role in human health. In this study, we aimed to investigate whether and how gut microbiota communities are altered in patients with immune-mediated necrotizing myopathy (IMNM) and provide new ideas to further explore the pathogenesis of IMNM or screen for its clinical therapeutic targets in the future. Methods: The gut microbiota collected from 19 IMNM patients and 23 healthy controls (HCs) were examined by using 16S rRNA gene sequencing. Alpha and beta-diversity analyses were applied to examine the bacterial diversity and community structure. Welch's t test was performed to identify the significantly abundant taxa of bacteria between the two groups. Spearman correlation analysis was performed to analyze the correlation between gut microbiota and clinical indicators. A receiver operator characteristic (ROC) curve was used to reflect the sensitivity and specificity of microbial biomarker prediction of IMNM disease. P < 0.05 was considered statistically significant. Results: Nineteen IMNM patients and 23 HCs were included in the analysis. Among IMNM patients, 94.74% (18/19) of them used glucocorticoids, while 57.89% (11/19) of them used disease-modifying antirheumatic drugs (DMARDs), and the disease was accessed by MITAX (18.26 ± 8.62) and MYOACT (20.68 ± 8.65) scores. Participants in the groups were matched for gender and age. The diversity of the gut microbiota of IMNM patients differed and decreased compared to that of HCs (Chao1, Shannon, and Simpson indexes: p < 0.05). In IMNM patients, the relative abundances of Bacteroides, Roseburia, and Coprococcus were decreased, while that of Lactobacillus and Streptococcus were relatively increased. Furthermore, in IMNM patients, Lactobacillus was positively correlated with the levels of anti-signal recognition particle (SRP) antibodies, anti-Ro52 antibodies, and erythrocyte sedimentation rate (ESR), while Streptococcus was positively correlated with anti-3-hydroxy-3-methylglutaryl-coenzyme A reductase (HMGCR) antibodies and C-reactive protein (CRP). Roseburia was negatively correlated with myoglobin (MYO), cardiac troponin T (cTnT), ESR, CRP, and the occurrence of interstitial lung disease (ILD). Bacteroides was negatively correlated with ESR and CRP, and Coprococcus was negatively correlated with ESR. Finally, the prediction model was built using the top five differential genera, which was verified using a ROC curve (area under the curve (AUC): 87%, 95% confidence interval: 73%-100%). Conclusion: We observed a characteristic compositional change in the gut microbiota with an abnormal elevation of Lactobacillus in IMNM patients, which was accompanied by changes in clinical indicators. This suggests that gut microbiota dysbiosis occurs in IMNM patients and is correlated with systemic autoimmune features.

Effects of Microecological Regulators on Rheumatoid Arthritis: A Systematic Review and Meta-Analysis of Randomized, Controlled Trials.

In this study, the available data from published randomized, controlled trials (RCTs) of the use of intestinal microecological regulators as adjuvant therapies to relieve the disease activity of rheumatoid arthritis (RA) are systematically compared. An English literature search was performed using PubMed, Embase, Scopus, Web of Science and the Cochrane Central Registry of Controlled Trials and supplemented by hand searching reference lists. Three independent reviewers screened and assessed the quality of the studies. Among the 2355 citations identified, 12 RCTs were included. All data were pooled using a mean difference (MD) with a 95% CI. The disease activity score (DAS) showed a significant improvement following microecological regulators treatment (MD (95% CI) of -1.01 (-1.81, -0.2)). A borderline significant reduction in the health assessment questionnaire (HAQ) scores was observed (MD (95% CI) of -0.11 (-0.21, -0.02)). We also confirmed the known effects of probiotics on inflammatory parameters such as the C-reactive protein (CRP) (MD -1.78 (95% CI -2.90, -0.66)) and L-1ß (MD -7.26 (95% CI -13.03, -1.50)). No significant impact on visual analogue scale (VAS) of pain and erythrocyte sedimentation rate (ESR) reduction was observed. Intestinal microecological regulators supplementation could decrease RA activity with a significant effect on DAS28, HAQ and inflammatory cytokines. Nevertheless, these findings need further confirmation in large clinical studies with greater consideration of the confounding variables of age, disease duration, and individual medication regimens.

Risk of Early Infection in Idiopathic Inflammatory Myopathies: Cluster Analysis Based on Clinical Features and Biomarkers.

Patients with idiopathic inflammatory myopathies (IIMs), referred to as myositis, are prone to infectious complications, which hinder the treatment of the disease and worsen the outcome of patients. The purpose of this study was to explore the different types of infectious complications in patients with myositis and to determine the predisposing factors for clinical reference. A retrospective study was conducted on 66 patients with IIM who were divided into different subpopulations by an unsupervised analysis of their clinical manifestations, laboratory features, and autoantibody characteristics. Combined with the incidence of infectious complications, the types of infectious pathogens and the sites of infection, the characteristics of infection, and susceptibility factors were explored. Three clusters with significantly different clinical characteristics and coinfection rates were identified (76.2% vs. 41.6% vs. 36.4%, p = 0.0139). Cluster 1 (n = 12) had a moderate risk of infection, with an infection rate of 41.6%. The patients in cluster 1 had a high probability of positive mechanic's hands, periungual erythema, anti-Ro52 antibody, and anti-Jo1 antibody. CD3 and CD4 were the highest among the three groups. Cluster 2 (n = 21) had a high risk of infection, and the incidence of infection was 76.2%. Almost all patients in this cluster had a rash, prominent clinical symptoms, and decreased WBC, PMN, LYM, CD3, and CD4 counts. Cluster 3 (n = 33) had a low risk of infection, with an infection rate of 36.4%. Compared with the other two clusters, cluster 3 (n = 33) lacked a typical rash but had a high ANA-positive rate. The patients in cluster 1 and cluster 3 were mainly infected by viruses, followed by bacterial infections. In cluster 2 patients, bacterial infections were the most prevalent. Fungal and Pneumocystis carinii were common causes of cluster 2 and 3 infections. In addition, the patients within a cluster often have a single infection, and pulmonary infections are the most common. We clustered the patients with IIM complicated with infection into three different types by their clinical symptoms and found that there were differences in the infection risk and infection types among the different cluster groups.

Domain and Content Adaptive Convolution Based Multi-Source Domain Generalization for Medical Image Segmentation.

The domain gap caused mainly by variable medical image quality renders a major obstacle on the path between training a segmentation model in the lab and applying the trained model to unseen clinical data. To address this issue, domain generalization methods have been proposed, which however usually use static convolutions and are less flexible. In this paper, we propose a multi-source domain generalization model based on the domain and content adaptive convolution (DCAC) for the segmentation of medical images across different modalities. Specifically, we design the domain adaptive convolution (DAC) module and content adaptive convolution (CAC) module and incorporate both into an encoder-decoder backbone. In the DAC module, a dynamic convolutional head is conditioned on the predicted domain code of the input to make our model adapt to the unseen target domain. In the CAC module, a dynamic convolutional head is conditioned on the global image features to make our model adapt to the test image. We evaluated the DCAC model against the baseline and four state-of-the-art domain generalization methods on the prostate segmentation, COVID-19 lesion segmentation, and optic cup/optic disc segmentation tasks. Our results not only indicate that the proposed DCAC model outperforms all competing methods on each segmentation task but also demonstrate the effectiveness of the DAC and CAC modules. Code is available at https://git.io/DCAC.

Learning From Ambiguous Labels for Lung Nodule Malignancy Prediction.

Lung nodule malignancy prediction is an essential step in the early diagnosis of lung cancer. Besides the difficulties commonly discussed, the challenges of this task also come from the ambiguous labels provided by annotators, since deep learning models have in some cases been found to reproduce or amplify human biases. In this paper, we propose a multi-view 'divide-and-rule' (MV-DAR) model to learn from both reliable and ambiguous annotations for lung nodule malignancy prediction on chest CT scans. According to the consistency and reliability of their annotations, we divide nodules into three sets: a consistent and reliable set (CR-Set), an inconsistent set (IC-Set), and a low reliable set (LR-Set). The nodule in IC-Set is annotated by multiple radiologists inconsistently, and the nodule in LR-Set is annotated by only one radiologist. Although ambiguous, inconsistent labels tell which label(s) is consistently excluded by all annotators, and the unreliable labels of a cohort of nodules are largely correct from the statistical point of view. Hence, both IC-Set and LR-Set can be used to facilitate the training of MV-DAR. Our MV-DAR contains three DAR models to characterize a lung nodule from three orthographic views and is trained following a two-stage procedure. Each DAR consists of three networks with the same architecture, including a prediction network (Prd-Net), a counterfactual network (CF-Net), and a low reliable network (LR-Net), which are trained on CR-Set, IC-Set, and LR-Set respectively in the pretraining phase. In the fine-tuning phase, the image representation ability learned by CF-Net and LR-Net is transferred to Prd-Net by negative-attention module (NA-Module) and consistent-attention module (CA-Module), aiming to boost the prediction ability of Prd-Net. The MV-DAR model has been evaluated on the LIDC-IDRI dataset and LUNGx dataset. Our results indicate not only the effectiveness of the MV-DAR in learning from ambiguous labels but also its superiority over present noisy label-learning models in lung nodule malignancy prediction.