Modality-Specific Information Disentanglement From Multi-Parametric MRI for Breast Tumor Segmentation and Computer-Aided Diagnosis.

Breast cancer is becoming a significant global health challenge, with millions of fatalities annually. Magnetic Resonance Imaging (MRI) can provide various sequences for characterizing tumor morphology and internal patterns, and becomes an effective tool for detection and diagnosis of breast tumors. However, previous deep-learning based tumor segmentation methods from multi-parametric MRI still have limitations in exploring inter-modality information and focusing task-informative modality/modalities. To address these shortcomings, we propose a Modality-Specific Information Disentanglement (MoSID) framework to extract both inter- and intra-modality attention maps as prior knowledge for guiding tumor segmentation. Specifically, by disentangling modality-specific information, the MoSID framework provides complementary clues for the segmentation task, by generating modality-specific attention maps to guide modality selection and inter-modality evaluation. Our experiments on two 3D breast datasets and one 2D prostate dataset demonstrate that the MoSID framework outperforms other state-of-the-art multi-modality segmentation methods, even in the cases of missing modalities. Based on the segmented lesions, we further train a classifier to predict the patients' response to radiotherapy. The prediction accuracy is comparable to the case of using manually-segmented tumors for treatment outcome prediction, indicating the robustness and effectiveness of the proposed segmentation method. The code is available at https://github.com/Qianqian-Chen/MoSID.

Alpha-lipoic Acid Protects Against Chronic Alcohol Consumption-induced Cardiac Damage by the Aldehyde Dehydrogenase 2-associated PINK/Parkin Pathway.

ABSTRACT: Chronic alcohol intake contributes to high mortality rates due to ethanol-induced cardiac hypertrophy and contractile dysfunction, which are accompanied by increased oxidative stress and disrupted mitophagy. Alpha-lipoic acid (α-LA), a well-known antioxidant, has been shown to protect against cardiac hypertrophy and inflammation. However, little is known about its role and mechanism in the treatment of alcoholic cardiomyopathy. Here, we evaluated the role of α-LA in alcohol-induced cardiac damage by feeding mice a 4.8% (v/v) alcohol diet with or without α-LA for 6 w. Our results suggested that chronic alcohol consumption increased mortality, blood alcohol concentrations, and serum aldehyde levels, but a-LA attenuated the elevations in mortality and aldehydes. Chronic alcohol intake also induced cardiac dysfunction, including enlarged left ventricles, reduced left ventricular ejection fraction, enhanced cardiomyocyte size, and increased serum levels of brain natriuretic peptide, lactate dehydrogenase, and creatine kinase myocardial isoenzyme. Moreover, alcohol intake led to the accumulation of collagen fiber and mitochondrial dysfunction, the effects of which were alleviated by α-LA. In addition, α-LA intake also prevented the increase in reactive oxygen species production and the decrease in mitochondrial number that were observed after alcohol consumption. Chronic alcohol exposure activated PINK1/Parkin-mediated mitophagy. These effects were diminished by α-LA intake by the activation of aldehyde dehydrogenase 2. Our data indicated that α-LA helps protect cardiac cells against the effects of chronic alcohol intake, likely by inhibiting PINK1/Parkin-related mitophagy through the activation of aldehyde dehydrogenase 2.

Astragaloside IV Inhibits NLRP3 Inflammasome-Mediated Pyroptosis via Activation of Nrf-2/HO-1 Signaling Pathway and Protects against Doxorubicin-Induced Cardiac Dysfunction.

BACKGROUND: Doxorubicin (DOX) is an effective broad-spectrum antitumor drug, but its clinical application is limited due to the side effects of cardiac damage. Astragaloside IV (AS-IV) is a significant active component of Astragalus membranaceus that exerts cardioprotective effects through various pathways. However, whether AS-IV exerts protective effects against DOX-induced myocardial injury by regulating the pyroptosis is still unknown and is investigated in this study. METHODS: The myocardial injury model was constructed by intraperitoneal injection of DOX, and AS-IV was administered via oral gavage to explore its specific protective mechanism. Cardiac function and cardiac injury indicators, including lactate dehydrogenase (LDH), cardiac troponin I (cTnI), creatine kinase isoenzyme (CK-MB), and brain natriuretic peptide (BNP), and histopathology of the cardiomyocytes were assessed 4 weeks post DOX challenge. Serum levels of IL-1ß, IL-18, superoxide dismutase (SOD), malondialdehyde (MDA) and glutathione (GSH) and the expression of pyroptosis and signaling proteins were also determined. RESULTS: Cardiac dysfunction was observed after the DOX challenge, as evidenced by reduced ejection fraction, increased myocardial fibrosis, and increased BNP, LDH, cTnI, and CK-MB levels (p < 0.05, N = 3-10). AS-IV attenuated DOX-induced myocardial injury. The mitochondrial morphology and structure were also significantly damaged after DOX treatment, and these changes were restored after AS-IV treatment. DOX induced an increase in the serum levels of IL-1ß, IL-18, SOD, MDA and GSH as well as an increase in the expression of pyroptosis-related proteins (p < 0.05, N = 3-6). Besides, AS-IV depressed myocardial inflammatory-related pyroptosis via activation of the expressions of nuclear factor E2-related factor 2 (Nrf-2) and heme oxygenase 1 (HO-1) (p < 0.05, N = 3). CONCLUSIONS: Our results showed that AS-IV had a significant protective effect against DOX-induced myocardial injury, which may be associated with the activation of Nrf-2/HO-1 to inhibit pyroptosis.

Exosomes derived from human umbilical cord mesenchymal stem cells (HUCMSC-EXO) regulate autophagy through AMPK-ULK1 signaling pathway to ameliorate diabetic cardiomyopathy.

One of the main causes of severe diabetic heart failure and mortality is diabetic cardiomyopathy (DCM), a cardiovascular condition attributable to diabetes with a high incidence, a complicated and unexplained pathophysiology, and poor treatment results. Current findings have demonstrated that the onset of diabetic cardiomyopathy involves autophagy, inflammation, and mitochondrial damage. Myocardial autophagy behaves differently in different states,and one of the targets for the detection and treatment of cardiovascular illnesses like diabetic cardiomyopathy may be the control of autophagy. The role of human umbilical cord Mesenchymal stem cells-derived exosomes (HUCMSC-EXO) as a non-cellular system in the repair of cardiomyocytes, the evolution of diabetic cardiomyopathy and their cardioprotective effects are gradually being recognized. This study's objectives were to assess the therapeutic benefits of HUCMSC-EXO for diabetic cardiomyopathy and to look into their potential mechanisms of action. High-speed centrifugation was used to extract HUCMSC-EXO, and the shape of the exosomes was examined using transmission electron microscopy. Immunoblotting was used to determine the expression of CD9, CD63, and TSG101 molecules on the surface of the exosomes. A high-fat, high-sugar diet mixed with streptozotocin was used to build a rat model of type 2 diabetic cardiomyopathy. Cardiac function, ventricular wall thickness and cardiac histological changes were examined by cardiac ultrasound, serum BNP and histology. In cardiac myocytes, HUCMSC-EXO reduced the levels of autophagy-related protein expression. Additionally, immunoblotting supported our suspicion that this mechanism is strongly tied to the activation of the AMPK-ULK1 signaling pathway. So, we propose that it would be a good strategy to follow for treating diabetic cardiomyopathy. These findings offer both fresh concepts for building a model of diabetic cardiomyopathy and a creative theoretical framework for using HUCMSC-EXO to treat diabetic cardiomyopathy in a clinical setting.

Accurate recognition of colorectal cancer with semi-supervised deep learning on pathological images.

Machine-assisted pathological recognition has been focused on supervised learning (SL) that suffers from a significant annotation bottleneck. We propose a semi-supervised learning (SSL) method based on the mean teacher architecture using 13,111 whole slide images of colorectal cancer from 8803 subjects from 13 independent centers. SSL (~3150 labeled, ~40,950 unlabeled; ~6300 labeled, ~37,800 unlabeled patches) performs significantly better than the SL. No significant difference is found between SSL (~6300 labeled, ~37,800 unlabeled) and SL (~44,100 labeled) at patch-level diagnoses (area under the curve (AUC): 0.980 ± 0.014 vs. 0.987 ± 0.008, P value = 0.134) and patient-level diagnoses (AUC: 0.974 ± 0.013 vs. 0.980 ± 0.010, P value = 0.117), which is close to human pathologists (average AUC: 0.969). The evaluation on 15,000 lung and 294,912 lymph node images also confirm SSL can achieve similar performance as that of SL with massive annotations. SSL dramatically reduces the annotations, which has great potential to effectively build expert-level pathological artificial intelligence platforms in practice.