Privacy preservation for federated learning in health care.

Artificial intelligence (AI) shows potential to improve health care by leveraging data to build models that can inform clinical workflows. However, access to large quantities of diverse data is needed to develop robust generalizable models. Data sharing across institutions is not always feasible due to legal, security, and privacy concerns.

Platelet factor 4 improves survival in a murine model of antibiotic-susceptible and methicillin-resistant peritonitis.

The complement receptor CR3, also known as integrin Mac-1 (CD11b/CD18), is one of the major phagocytic receptors on the surface of neutrophils and macrophages. We previously demonstrated that in its protein ligands, Mac-1 binds sequences enriched in basic and hydrophobic residues and strongly disfavors negatively charged sequences. The avoidance by Mac-1 of negatively charged surfaces suggests that the bacterial wall and bacterial capsule possessing net negative electrostatic charge may repel Mac-1 and that the cationic Mac-1 ligands can overcome this evasion by acting as opsonins.

PLATELET FACTOR 4 (PF4) IMPROVES SURVIVAL IN A MURINE MODEL OF ANTIBIOTIC-SUSCEPTIBLE AND METHICILLIN-RESISTANT PERITONITIS.

The complement receptor CR3, also known as integrin Mac-1 (CD11b/CD18), is one of the major phagocytic receptors on the surface of neutrophils and macrophages. We previously demonstrated that in its protein ligands, Mac-1 binds sequences enriched in basic and hydrophobic residues and strongly disfavors negatively charged sequences. The avoidance by Mac-1 of negatively charged surfaces suggests that the bacterial wall and bacterial capsule possessing net negative electrostatic charge may repel Mac-1 and that the cationic Mac-1 ligands can overcome this evasion by acting as opsonins.

The Brain Tumor Segmentation - Metastases (BraTS-METS) Challenge 2023: Brain Metastasis Segmentation on Pre-treatment MRI.

The translation of AI-generated brain metastases (BM) segmentation into clinical practice relies heavily on diverse, high-quality annotated medical imaging datasets. The BraTS-METS 2023 challenge has gained momentum for testing and benchmarking algorithms using rigorously annotated internationally compiled real-world datasets. This study presents the results of the segmentation challenge and characterizes the challenging cases that impacted the performance of the winning algorithms.

OpenFL: the open federated learning library.

Federated learning (FL) is a computational paradigm that enables organizations to collaborate on machine learning (ML) and deep learning (DL) projects without sharing sensitive data, such as patient records, financial data, or classified secrets.Open federated learning (OpenFL) framework is an open-source python-based tool for training ML/DL algorithms using the data-private collaborative learning paradigm of FL, irrespective of the use case. OpenFL works with training pipelines built with both TensorFlow and PyTorch, and can be easily extended to other ML and DL frameworks.

The federated tumor segmentation (FeTS) tool: an open-source solution to further solid tumor research.

De-centralized data analysis becomes an increasingly preferred option in the healthcare domain, as it alleviates the need for sharing primary patient data across collaborating institutions. This highlights the need for consistent harmonized data curation, pre-processing, and identification of regions of interest based on uniform criteria.Towards this end, this manuscript describes thederatedumoregmentation (FeTS) tool, in terms of software architecture and functionality.

The future of digital health with federated learning.

Data-driven machine learning (ML) has emerged as a promising approach for building accurate and robust statistical models from medical data, which is collected in huge volumes by modern healthcare systems. Existing medical data is not fully exploited by ML primarily because it sits in data silos and privacy concerns restrict access to this data. However, without access to sufficient data, ML will be prevented from reaching its full potential and, ultimately, from making the transition from research to clinical practice.

Federated learning in medicine: facilitating multi-institutional collaborations without sharing patient data.

Several studies underscore the potential of deep learning in identifying complex patterns, leading to diagnostic and prognostic biomarkers. Identifying sufficiently large and diverse datasets, required for training, is a significant challenge in medicine and can rarely be found in individual institutions. Multi-institutional collaborations based on centrally-shared patient data face privacy and ownership challenges.

Multi-Institutional Deep Learning Modeling Without Sharing Patient Data: A Feasibility Study on Brain Tumor Segmentation.

Deep learning models for semantic segmentation of images require large amounts of data. In the medical imaging domain, acquiring sufficient data is a significant challenge. Labeling medical image data requires expert knowledge.

Deposition of fibrinogen on the surface of in vitro thrombi prevents platelet adhesion.

The initial accumulation of platelets after vessel injury is followed by thrombin-mediated generation of fibrin which is deposited around the plug. While numerous in vitro studies have shown that fibrin is highly adhesive for platelets, the surface of experimental thrombi in vivo contains very few platelets suggesting the existence of natural anti-adhesive mechanisms protecting stabilized thrombi from platelet accumulation and continuous thrombus propagation. We previously showed that adsorption of fibrinogen on pure fibrin clots results in the formation of a nonadhesive matrix, highlighting a possible role of this process in surface-mediated control of thrombus growth.

Combined magnetic fields provide robust coverage for interbody and posterolateral lumbar spinal fusion sites.

Electromagnetic fields generated by spinal bone growth stimulation devices have been computationally modelled to determine coverage of the lumbar spinal vertebrae. The underlying assumption of these models was that the electric field, but not the magnetic field, was therapeutically relevant. However, there are no published studies examining the therapeutic coverage of spinal fusion sites by stimulators utilizing combined magnetic fields.

Effect of thrombin fragment (TP508) on myocardial ischemia reperfusion injury in a model of type 1 diabetes mellitus.

Background: We investigated the efficacy of novel thrombin fragment TP508 on ischemia-reperfusion injury using a porcine model of type 1 diabetes mellitus.

Methods And Results: Alloxan-induced diabetic male Yucatan swine underwent 60 minutes of mid-left anterior descending coronary artery occlusion, followed by 120 minutes of reperfusion. Fifty minutes into ischemia, animals received either placebo (DM; n=8) or TP508 as a bolus of 1 mg/kg followed by infusion at 2.

RGD-dependent binding of TP508 to integrin alphavbeta3 mediates cell adhesion and induction of nitric oxide.

TP508, a 23-amino acid RGD-containing synthetic peptide representing residues 508 to 530 of human prothrombin, mitigates the effects of endothelial dysfunction in ischaemic reperfusion injury. The objective of this study was to investigate whether TP508 binds to members of the integrin family of transmembrane receptors leading to nitric oxide synthesis. Immobilised TP508 supported adhesion of endothelial cells and alphavbeta3-expressing human embryonic kidney cells in a dose- and RGD-dependent manner.

Effect of dimerized thrombin fragment TP508 on acute myocardial ischemia reperfusion injury in hypercholesterolemic swine.

The thrombin-related peptide TP508 is a 23-amino acid monomer that represents a portion of the receptor binding domain in the thrombin molecule. TP508 is also known to readily convert to a dimer in an aqueous environment. In this study the dimeric form of TP508 was investigated in a porcine model of acute myocardial ischemia reperfusion injury (and compared with its monomer).

Internalization and intracellular trafficking of a PTD-conjugated anti-fibrotic peptide, AZX100, in human dermal keloid fibroblasts.

A challenge in advanced drug delivery is selectively traversing the plasma membrane, a barrier that prohibits the intracellular delivery of most peptide and nucleic acid-based therapeutics. A variety of short amino acid sequences termed protein transduction domains (PTDs) first identified in viral proteins have been utilized for over 20 years to deliver proteins nondestructively into cells, however, the mechanisms by which this occurs are varied and cell-specific. Here we describe the results of live cell imaging experiments with AZX100, a cell-permeable anti-fibrotic peptide bearing an "enhanced" PTD (PTD4).

The small heat shock protein, HSPB6, in muscle function and disease.

The small heat shock protein, HSPB6, is a 17-kDa protein that belongs to the small heat shock protein family. HSPB6 was identified in the mid-1990s when it was recognized as a by-product of the purification of HSPB1 and HSPB5. HSPB6 is highly and constitutively expressed in smooth, cardiac, and skeletal muscle and plays a role in muscle function.

Thrombin peptide TP508 stimulates cellular events leading to angiogenesis, revascularization, and repair of dermal and musculoskeletal tissues.

The thrombin peptide, TP508, also known as Chrysalin (OrthoLogic, Tempe, Arizona), is a twenty-three-amino-acid peptide that represents a portion of the receptor-binding domain of the native human thrombin molecule that has been identified as the binding site for a specific class of receptors on fibroblasts and other cells. Preclinical studies with this peptide have shown that it can accelerate tissue repair in both soft and hard tissues by mechanisms that appear to involve up-regulation of genes that initiate a cascade of healing events. These events include recruitment and activation of inflammatory cells, directed migration of cells (chemotaxis), cell proliferation, elaboration of extra-cellular matrix, and accelerated revascularization of the healing tissues.

Repair of rabbit segmental defects with the thrombin peptide, TP508.

The synthetic peptide, TP508 (Chrysalin), was delivered to rabbit segmental bone defects in biodegradable controlled-release PLGA microspheres to determine its potential efficacy for enhancing healing of non-critically and critically sized segmental defects. Non-critically sized radial defects were created in the forelimbs of New Zealand White rabbits, which were randomized into three treatment groups receiving 10, 50 and 100 microg doses of TP508 in the right radius and control microspheres (without TP508) in the left radius. Torsional testing of the radii at six weeks showed a significant increase in ultimate torque, failure torque, ultimate energy, failure energy, and stiffness when treated with TP508 compared to controls (p<0.

Prevalence of hepatitis B in selected Alaskan Eskimo villages.

Sera collected in 1973-1975 from 3053 residents of 12 selected Alaskan Eskimo villages were tested for evidence of hepatitis B virus infection. Overall, hepatitis B surface antigen (HBsAg) was found in 6.4% of those tested.

Serological testing for hepatitis B in male homosexuals: special emphasis on hepatitis B e antigen and antibody by radioimmunoassay.

Serological markers for hepatitis B virus in male homosexuals demonstrated a high prevalence of past and present infection. Seropositivity of 91% for hepatitis B e antigen or antibody was demonstrated by radioimmunoassay in hepatitis B surface antigen-positive specimens.

Experimental studies on the transmission of hepatitis B by mosquitoes.

Culex tarsalis and Aedes aegypti mosquitoes were fed on chimpanzees carrying hepatitis B surface antigen (HBS Ag) of known infectivity and pools were tested by radioimmunoassay daily for the presence of HBS Ag. HBS Ag continued to be detected at low levels in mosquito tissue after digestion of the blood meal. Inoculation of susceptible chimpanzees with macerated pools of A.