A Study on the Effectiveness of Deep Learning-Based Anomaly Detection Methods for Breast Ultrasonography.

In the medical field, it is delicate to anticipate good performance in using deep learning due to the lack of large-scale training data and class imbalance. In particular, ultrasound, which is a key breast cancer diagnosis method, is delicate to diagnose accurately as the quality and interpretation of images can vary depending on the operator's experience and proficiency. Therefore, computer-aided diagnosis technology can facilitate diagnosis by visualizing abnormal information such as tumors and masses in ultrasound images. In this study, we implemented deep learning-based anomaly detection methods for breast ultrasound images and validated their effectiveness in detecting abnormal regions. Herein, we specifically compared the sliced-Wasserstein autoencoder with two representative unsupervised learning models autoencoder and variational autoencoder. The anomalous region detection performance is estimated with the normal region labels. Our experimental results showed that the sliced-Wasserstein autoencoder model outperformed the anomaly detection performance of others. However, anomaly detection using the reconstruction-based approach may not be effective because of the occurrence of numerous false-positive values. In the following studies, reducing these false positives becomes an important challenge.

Potential Therapeutic Skin Microbiomes Suppressing Staphylococcus aureus-Derived Immune Responses and Upregulating Skin Barrier Function-Related Genes via the AhR Signaling Pathway.

Disruption of the skin microbial balance can exacerbate certain skin diseases and affect prognosis and treatment. Changes in the distribution and prevalence of certain microbial species on the skin, such as Staphylococcus aureus (SA), can impact the development of severe atopic dermatitis (AD) or psoriasis (Pso). A dysfunctional skin barrier develops in AD and Pso due to SA colonization, resulting in keratinization and chronic or progressive chronic inflammation. Disruption of the skin barrier following SA colonization can elevate the production of T helper 2 (Th2)-derived cytokines, which can cause an imbalance in Th1, Th2, and Th17 cells. This study examined the ability of potential therapeutic skin microbiomes, such as Cutibacterium avidum R-CH3 and Staphylococcus hominis R9, to inhibit SA biofilm formation and restore skin barrier function-related genes through the activation of the aryl hydrocarbon receptor (AhR) and the nuclear factor erythroid-2-related factor 2 (Nrf2) downstream target. We observed that IL-4/IL-13-induced downregulation of FLG, LOR, and IVL induced by SA colonization could be reversed by dual AhR/Nrf2 activation. Further, OVOL1 expression may be modulated by functional microbiomes via dual AhR/Nrf2 activation. Our results suggest that our potential therapeutic skin microbiomes can prevent SA-derived Th2-biased skin barrier disruption via IL-13 and IL-4-dependent FLG deregulation, STAT3 activation, and AhR-mediated STAT6 expression.

Prediction of lung cancer using novel biomarkers based on microbiome profiling of bronchoalveolar lavage fluid.

There is an unmet need for biomarkers for the diagnosis of lung cancer and decision criteria for lung biopsy. We comparatively investigated the lung microbiomes of patients with lung cancer and benign lung diseases. Patients who underwent bronchoscopy at Chungnam National University Hospital between June 2021 and June 2022 were enrolled. Bronchoalveolar lavage fluid (BALF) was collected from 24 patients each with lung cancer and benign lung diseases. The samples were analyzed using 16S rRNA-based metagenomic sequencing. We found that alpha diversity and the beta diversity distribution (P = 0.001) differed significantly between patients with benign lung diseases and those with lung cancer. Firmicutes was the most abundant phylum in patients with lung cancer (33.39% ± 17.439), whereas Bacteroidota was the most abundant phylum in patients with benign lung disease (31.132% ± 22.505), respectively. In differential abundance analysis, the most differentially abundant microbiota taxon was unclassified_SAR202_clade, belonging to the phylum Chloroflexi. The established prediction model distinguished patients with benign lung disease from those with lung cancer with a high accuracy (micro area under the curve [AUC] = 0.98 and macro AUC = 0.99). The BALF microbiome may be a novel biomarker for the detection of lung cancer.