Noise-induced modality-specific pretext learning for pediatric chest X-ray image classification.

Introduction: Deep learning (DL) has significantly advanced medical image classification. However, it often relies on transfer learning (TL) from models pretrained on large, generic non-medical image datasets like ImageNet. Conversely, medical images possess unique visual characteristics that such general models may not adequately capture.

Automated quantification of SARS-CoV-2 pneumonia with large vision model knowledge adaptation.

Background: Large vision models (LVM) pretrained by large datasets have demonstrated their enormous capacity to understand visual patterns and capture semantic information from images. We proposed a novel method of knowledge domain adaptation with pretrained LVM for a low-cost artificial intelligence (AI) model to quantify the severity of SARS-CoV-2 pneumonia based on frontal chest X-ray (CXR) images.

Methods: Our method used the pretrained LVMs as the primary feature extractor and self-supervised contrastive learning for domain adaptation.

A pan-KRAS degrader for the treatment of KRAS-mutant cancers.

KRAS mutations are highly prevalent in a wide range of lethal cancers, and these mutant forms of KRAS play a crucial role in driving cancer progression and conferring resistance to treatment. While there have been advancements in the development of small molecules to target specific KRAS mutants, the presence of undruggable mutants and the emergence of secondary mutations continue to pose challenges in the clinical treatment of KRAS-mutant cancers. In this study, we developed a novel molecular tool called tumor-targeting KRAS degrader (TKD) that effectively targets a wide range of KRAS mutants.

Semantically redundant training data removal and deep model classification performance: A study with chest X-rays.

Deep learning (DL) has demonstrated its innate capacity to independently learn hierarchical features from complex and multi-dimensional data. A common understanding is that its performance scales up with the amount of training data. However, the data must also exhibit variety to enable improved learning.

The Transpeptidase Sortase A Binds Nucleic Acids and Mediates Mammalian Cell Labeling.

Wild-type sortase A is an important virulence factor displaying a diverse array of proteins on the surface of bacteria. This protein display relies on the transpeptidase activity of sortase A, which is widely engineered to allow protein ligation and protein engineering based on the interaction between sortase A and peptides. Here an unknown interaction is found between sortase A from Staphylococcus aureus and nucleic acids, in which exogenously expressed engineered sortase A binds oligonucleotides in vitro and is independent of its canonical transpeptidase activity.

Automatic Quantification of COVID-19 Pulmonary Edema by Self-supervised Contrastive Learning.

We proposed a self-supervised machine learning method to automatically rate the severity of pulmonary edema in the frontal chest X-ray radiographs (CXR) which could be potentially related to COVID-19 viral pneumonia. For this we use the modified radiographic assessment of lung edema (mRALE) scoring system. The new model was first optimized with the simple Siamese network (SimSiam) architecture where a ResNet-50 pretrained by ImageNet database was used as the backbone.

Uncovering the effects of model initialization on deep model generalization: A study with adult and pediatric chest X-ray images.

Model initialization techniques are vital for improving the performance and reliability of deep learning models in medical computer vision applications. While much literature exists on non-medical images, the impacts on medical images, particularly chest X-rays (CXRs) are less understood. Addressing this gap, our study explores three deep model initialization techniques: Cold-start, Warm-start, and Shrink and Perturb start, focusing on adult and pediatric populations.

Semantically Redundant Training Data Removal and Deep Model Classification Performance: A Study with Chest X-rays.

Deep learning (DL) has demonstrated its innate capacity to independently learn hierarchical features from complex and multi-dimensional data. A common understanding is that its performance scales up with the amount of training data. Another data attribute is the inherent variety.

Whole-Voltage-Range Solid-Solution Reaction in Layered Oxide Cathode of Sodium-Ion Batteries.

Layered manganese-based oxides (LMOs) are promising cathode materials for sodium-ion batteries (SIBs) due to their versatile structures. However, the Jahn-Teller effect of Mn induces severe distortion of MnO octahedra, and the resultant low symmetry is responsible for the gliding of MnO layers and then inferior multiple-phase transitions upon Na extraction/insertion. Here, hexagonal P2-Na Li Mn Ti O is synthesized through the incorporation of Li and Ti into the distorted orthorhombic P'2-Na MnO to function as a phase-transition-free oxide cathode.

Image Translation by Ad CycleGAN for COVID-19 X-Ray Images: A New Approach for Controllable GAN.

We propose a new generative model named adaptive cycle-consistent generative adversarial network, or Ad CycleGAN to perform image translation between normal and COVID-19 positive chest X-ray images. An independent pre-trained criterion is added to the conventional Cycle GAN architecture to exert adaptive control on image translation. The performance of Ad CycleGAN is compared with the Cycle GAN without the external criterion.

Clinical and Dosimetric Predictors for Postoperative Cardiopulmonary Complications in Esophageal Squamous Cell Carcinoma Patients Receiving Neoadjuvant Chemoradiotherapy and Surgery.

Background: Neoadjuvant chemoradiotherapy followed by esophagectomy is the standard treatment for patients with locally advanced esophageal squamous cell carcinoma (ESCC). This study explored correlations of clinical factors and dose-volume histogram (DVH) parameters with postoperative cardiopulmonary complications and predicted their risk by establishing a nomogram model.

Methods: Clinical and DVH parameters of ESCC patients who underwent trimodality treatment from 2002 to 2020 were collected.

Knockdown of PHLDA2 promotes apoptosis and autophagy of glioma cells through the AKT/mTOR pathway.

Pleckstrin homology like domain family A member 2 (PHLDA2) is an imprinted gene expressed in placenta and has been shown to be associated with tumor progression. However, the effect of PHLDA2 on glioma cell growth has not been reported yet. Data based on TCGA database showed that PHLDA2 was up-regulated in glioma tissues.

CycleGAN with Dynamic Criterion for Malaria Blood Cell Image Synthetization.

We present a cycle-consistent adversarial network (Cycle GAN) with dynamic criterion to synthesize blood cells parasitized by malaria plasmodia. The result shows 100% of the synthetic images are correctly classified by the pretrained classifier compared to 99.61% of the real images, 76.

Astrocytes constitute the major TNF-α-producing cell population in the infarct cortex in dMCAO rats receiving intravenous MSC infusion.

Recent studies report that inhibiting TNF-α might be a novel therapeutic strategy for managing brain ischemia. Our previous study reported that mesenchymal stem cell (MSC) transplantation could suppress TNF-α level in both serum and brain. However, the cell type(s) that contribute to the production of TNF-α during ischemia following MSC transplantation has not been well studied.

A SUMO1-Derived Peptide Targeting SUMO-Interacting Motif Inhibits α-Synuclein Aggregation.

The accumulation of α-synuclein amyloid fibrils in the brain is linked to Parkinson's disease and other synucleinopathies. The intermediate species in the early aggregation phase of α-synuclein are involved in the emergence of amyloid toxicity and considered to be the most neurotoxic. The N-terminal region flanking the non-amyloid-β component domain of α-synuclein has been implicated in modulating its aggregation.

Health-related quality of life in pilots of a Chinese commercial airline.

Health-related quality of life (HRQOL) is currently an important issue in the medical industry. However, data on HRQOL in commercial airline pilots are lacking. This study aimed to investigate HRQOL and its related factors in a sample of commercial pilots.

Bone marrow mesenchymal stem cell transplantation downregulates plasma level and the microglia expression of transforming growth factor β1 in the acute phase of cerebral cortex ischemia.

Background: Both bone marrow mesenchymal stem cell (BM-MSC) and transforming growth factor-β1 (TGF-β1) have a strong anti-inflammatory capacity in stroke. But their relationship has not been well addressed. In this study, we investigated how intravenous BM-MSC transplantation in rats effected the expression of TGF-β1 48 h post cerebral ischemia, and we analyzed the main cells that produce TGF-β1.

Peripheral Circulation and Astrocytes Contribute to the MSC-Mediated Increase in IGF-1 Levels in the Infarct Cortex in a dMCAO Rat Model.

Materials And Methods: Ischemic brain injury was induced by dMCAO in Sprague-Dawley rats. The transplantation group received MSC infusion 1 h after dMCAO. Expression of IGF-1 in GFAP+ astrocytes, Iba-1+ microglia/macrophages, CD3+ lymphocytes, Ly6C+ monocytes/macrophages, and neutrophil elastase (NE)+ neutrophils was examined to determine the contribution of these cells to the increase of IGF-1.

Optimized acupuncture treatment (acupuncture and intradermal needling) for cervical spondylosis-related neck pain: a multicenter randomized controlled trial.

Cervical spondylosis (CS)-related neck pain is difficult to treat because of its degenerative nature. The aim of this 9-center, single-blinded, randomized controlled trial was to evaluate the efficacy of optimized acupuncture for CS-related neck pain. Participants who met the inclusion criteria were randomized to optimized, shallow, and sham acupuncture groups (1:1:1).

A dual-ion battery has two sides: the effect of ion-pairs.

There is mutual interference between the positive and negative electrodes in a dual-ion battery. In LiTiO/graphite batteries, strong ion-pairs between Li and anions may have a detrimental effect on intercalation at both sides: Li into LiTiO, and anions into graphite.

Changes in Indoleamine 2,3-Dioxygenase 1 Expression and CD8+ Tumor-Infiltrating Lymphocytes after Neoadjuvant Chemoradiation Therapy and Prognostic Significance in Esophageal Squamous Cell Carcinoma.

Purpose: Despite good preclinical evidence, clinical data on the effect of neoadjuvant chemoradiation therapy (CRT) on expression of immune markers in esophageal cancer are limited. This study aimed to evaluate the changes in indoleamine 2,3-dioxygenase 1 (IDO1) expression, programmed cell death-ligand 1 (PD-L1) expression, and CD8+ tumor-infiltrating lymphocyte status after neoadjuvant CRT and the prognostic significance in esophageal squamous cell carcinoma (ESCC).

Methods And Materials: Between 2003 and 2017, 138 patients with ESCC who underwent neoadjuvant CRT and esophagectomy without achieving pathologic complete response were included for analysis.

Indoleamine 2,3-dioxygenase 1 and Programmed Cell Death-ligand 1 Co-expression Predicts Poor Pathologic Response and Recurrence in Esophageal Squamous Cell Carcinoma after Neoadjuvant Chemoradiotherapy.

This study aimed to investigate the impact of indoleamine 2,3-dioxygenase 1 (IDO1) expression, programmed cell death-ligand 1 (PD-L1) expression, CD8+ tumor-infiltrating lymphocyte (TIL) status, and their combination on pathologic complete response (pCR) and recurrence in esophageal squamous cell carcinoma (ESCC) treated with neoadjuvant chemoradiotherapy (CRT). Indoleamine 2,3-dioxygenase 1, PD-L1, and CD8+ TIL statuses were evaluated by immunohistochemical analysis on pre-CRT biopsies of 158 patients. Sixty-eight patients (43.

Intravenous Transplantation of Mesenchymal Stem Cells Reduces the Number of Infiltrated Ly6C Cells but Enhances the Proportions Positive for BDNF, TNF-1, and IL-1 in the Infarct Cortices of dMCAO Rats.

The resident microglial and infiltrating cells from peripheral circulation are involved in the pathological processes of ischemia stroke and may be regulated by mesenchymal stem/stromal cell (MSC) transplantation. The present study is aimed at differentiating the neurotrophic and inflammatory roles played by microglial vs. infiltrating circulation-derived cells in the acute phase in rat ischemic brains and explore the influences of intravenously infused allogeneic MSCs.