Diffusion Model Research Articles

CT synthesis with deep learning for MR-only radiotherapy planning: a review.

MR-only radiotherapy planning is beneficial from the perspective of both time and safety since it uses synthetic CT for radiotherapy dose calculation instead of real CT scans. To elevate the accuracy of treatment planning and apply the results in practice, various methods have been adopted, among which deep learning models for image-to-image translation have shown good performance by retaining domain-invariant structures while changing domain-specific details. In this paper, we present an overview of diverse deep learning approaches to MR-to-CT synthesis, divided into four classes: convolutional neural networks, generative adversarial networks, transformer models, and diffusion models. By comparing each model and analyzing the general approaches applied to this task, the potential of these models and ways to improve the current methods can be can be evaluated.

The Role of the Rat Prefrontal Cortex and Sex Differences in Decision-Making.

The prefrontal cortex is critical for decision-making across species, with its activity linked to choosing between options. Drift Diffusion Models (DDMs) are commonly employed to understand the neural computations underlying this behavior. Studies exploring the specific roles of regions of the rodent prefrontal cortex in controlling the decision process are limited. This study explored the role of the prelimbic cortex (PLC) in decision-making using a two-alternative forced-choice task. Rats first learned to report the location of a lateralized visual stimulus. The brightness of the stimulus indicated its reward value. Then, the rats learned to make choices between pairs of stimuli. Sex differences in learning were observed, with females responding faster and more selectively to high-value stimuli than males. DDM analysis found that males had decreased decision thresholds during initial learning, whereas females maintained a consistently higher drift rate. Pharmacological manipulations revealed that PLC inactivation reduced the decision threshold for all rats, indicating that less information was needed to make a choice in the absence of normal PLC processing. Mu opioid receptor stimulation of the PLC had the opposite effect, raising the decision threshold and reducing bias in the decision process towards high-value stimuli. These effects were observed without any impact on the rats' choice preferences. Our findings suggest that PLC has an inhibitory role in the decision process and regulates the amount of evidence that is required to make a choice. That is, PLC activity controls "when", but not "how", to act.Significance Statement This study reports causal evidence for a part of the rat prefrontal cortex, the prelimbic cortex, in controlling the amount of information needed to make a choice. Results were based on reversible inactivation using the GABA-A agonist muscimol and by stimulation of mu opioid receptors using intra-cortical infusions of the selective mu agonist DAMGO. We also found evidence for a sex difference in learning and performing a visually guided two-alternative forced-choice task. Drift Diffusion Models found that females had stable decision processes throughout learning, and showed a persistent bias against the lower value option. By contrast, males exhibited changes in their decision processes, notably reducing the amount of information needed to make choice over the period of early choice learning.

TDAD: Self-supervised industrial anomaly detection with a two-stage diffusion model

Visual anomaly detection has emerged as a highly applicable solution in practical industrial manufacturing, owing to its notable effectiveness and efficiency. However, it also presents several challenges and uncertainties. To address the complexity of anomaly types and the high cost associated with data annotation, this paper introduces a self-supervised learning framework called TDAD, based on a two-stage diffusion model. TDAD consists of three key components: anomaly synthesis, image reconstruction, and defect segmentation. It is trained end-to-end, with the goal of improving pixel-level segmentation accuracy of anomalies and reducing false detection rates. By synthesizing anomalies from normal samples, designing a diffusion model-based reconstruction network, and incorporating a multiscale semantic feature fusion module for defect segmentation, TDAD achieves state-of-the-art performance in image-level detection and anomaly localization on challenging and widely used datasets such as MVTec and VisA benchmarks.

ProT-Diff: A Modularized and Efficient Strategy for De Novo Generation of Antimicrobial Peptide Sequences by Integrating Protein Language and Diffusion Models.

Antimicrobial peptides (AMPs) are a promising solution for treating antibiotic-resistant pathogens. However, efficient generation of diverse AMPs without prior knowledge of peptide structures or sequence alignments remains a challenge. Here, ProT-Diff is introduced, a modularized deep generative approach that combines a pretrained protein language model with a diffusion model for the de novo generation of AMPs sequences. ProT-Diff generates thousands of AMPs with diverse lengths and structures within a few hours. After silico physicochemical screening, 45 peptides are selected for experimental validation. Forty-four peptides showed antimicrobial activity against both gram-positive or gram-negative bacteria. Among broad-spectrum peptides, AMP_2 exhibited potent antimicrobial activity, low hemolysis, and minimal cytotoxicity. An in vivo assessment demonstrated its effectiveness against a drug-resistant E. coli strain in acute peritonitis. This study not only introduces a viable and user-friendly strategy for de novo generation of antimicrobial peptides, but also provides potential antimicrobial drug candidates with excellent activity. It is believed that this study will facilitate the development of other peptide-based drug candidates in the future, as well as proteins with tailored characteristics.

BP-Diff: a conditional diffusion model for cuffless continuous BP waveform estimation using U-Net.

Continuous monitoring of blood pressure (BP) is crucial for daily healthcare. Although invasive methods provide accurate continuous BP measurements, they are not suitable for routine use. Photoplethysmography (PPG), a non-invasive technique that detects changes in blood volume within the microcirculation using light, shows promise for BP measurement. The primary goal of this study is to develop a novel cuffless method based on PPG for accurately estimating continuous BP.
Approach. We introduce BP-Diff, an end-to-end method for cuffless continuous BP waveform estimation utilizing a conditional diffusion probability model combined with a U-Net architecture. This approach takes advantage of the stochastic properties of diffusion models and the strong feature representation capabilities of U-Net. It integrates the continuous BP waveform as the initial status and uses the PPG signal and its derivatives as conditions to guide the training and sampling process.
Main results. BP-Diff was evaluated using both uncalibrated and calibrated schemes. The results indicate that, when uncalibrated, BP-Diff can accurately track BP dynamics, including peak and valley positions, as well as timing. After calibration, BP-Diff achieved highly accurate BP estimations. The mean absolute error (MAE) of the estimated BP waveforms, along with the systolic blood pressure (SBP), diastolic blood pressure (DBP), and mean arterial pressure (MAP) from the calibrated BP-Diff model, were 2.99 mmHg, 2.6 mmHg, 1.4 mmHg, and 1.44 mmHg, respectively. Consistency tests, including Bland-Altman analysis and Pearson correlation, confirmed its high reliability compared to reference BP. BP-Diff meets the American Association for Medical Instrumentation (AAMI) standards and has achieved a Grade A from the British Hypertension Society (BHS).
Significance. This study utilizes PPG signals to develop a novel cuffless continuous BP measurement method, demonstrating superiority over existing approaches. The method is suitable for integration into wearable devices, providing a practical solution for continuous BP monitoring in everyday healthcare.

Kinetics and thermodynamics investigations of efficient and eco-friendly removal of alizarin red S from water via acid-activated Dalbergia sissoo leaf powder and its magnetic iron oxide nanocomposite.

The present work aimed to highlight an efficient, readily accessible, and cost-effective adsorbent derived from Dalbergia sissoo (DS) leaf powder for removing the environmentally hazardous dye "alizarin red S" (ARS) from hydrous medium. A variant of the adsorbent is activated via sulfuric acid and composited with magnetic iron oxide nanoparticles (DSMNC). Both adsorbents are thoroughly characterized using techniques such as Fourier transform infrared spectroscopy, point of zero charge, energy-dispersive X-ray spectroscopy, and scanning electron microscopy, which show that they have a porous structure rich in active sites. Different adsorption conditions are optimized with the maximum removal efficiency of 76.63% for DS and 97.89% for DSMNC. The study was highlighted via the application of various adsorption isotherms, including Freundlich, Langmuir, Temkin, and Dubinin-Radushkevich, to adsorption data. Pseudo-first-order, pseudo-second-order, and intra-particle diffusion models were utilized to investigate the kinetics and mechanism of adsorption. The Freundlich model and pseudo-second-order kinetics exhibited the best fit, suggesting a combination of physical interactions, as confirmed by the D-R and Temkin models. The dominant adsorbate-adsorbent interactive interactions responsible for ARS removal were hydrogen bonding, dispersion forces, and noncovalent aromatic ring adsorbent pi-interactions. Thermodynamic parameters extracted from adsorption data indicated that the removal of the mutagenic dye "ARS" was exothermic and spontaneous on both DS and DSMNC, with DSMNC exhibiting higher removal efficiency.

Enhanced phosphorus adsorption using modified drinking water treatment residues: A comparative analysis of powder and alginate bead forms

This study provides an analysis of the phosphorus adsorption efficacy of three modified drinking water treatment residues (MDWTRs): MDWTR-P (powdered form), MDWTR-D2, and MDWTR-D5 (alginate bead-entrapped forms with bead diameters of 2 mm and 5 mm, respectively). The preparation process involved washing and drying the drinking water treatment residue, followed by grinding and sieving to achieve particle sizes below 90 μm. The residue was then incinerated at 600 °C in oxygen-limited conditions. Subsequently, the MDWTR was formulated into alginate beads by mixing with sodium alginate and FeCl3 solutions, resulting in spherical particles of specified diameters. The evaluation of surface area, pore volume, pore size, and CHN concentration revealed that MDWTR-D5 possesses the largest surface area (284.7 m2 g−1) and highest micropore volume (0.04 cm3 g−1), indicating a greater capacity for adsorption. SEM-EDS analysis demonstrated significant compositional changes post-treatment, particularly elevated phosphorus levels, confirming effective adsorption. Metal content analysis indicated high aluminum levels in MDWTR-P and increased iron content in MDWTR-D5. Toxicity Characteristic Leaching Procedure (TCLP) and in vitro bioaccessibility (IVBA) tests confirmed the non-hazardous nature of all MDWTRs, ensuring their safety for environmental applications. Kinetic analyses using pseudo-first-order, pseudo-second-order, and intraparticle diffusion models highlighted the superior performance of MDWTR-D5, with the highest equilibrium adsorption capacity and initial adsorption rate across all tested concentrations, suggesting both high efficiency and rapid adsorption potential. Further validation using Langmuir and Freundlich isotherms revealed MDWTR-D5's highest monolayer adsorption capacity (22.88 mg g−1) and Freundlich adsorption capacity parameter (6.97 mg g−1). Statistical analysis via one-way ANOVA confirmed significant differences in phosphorus concentrations among the MDWTRs samples (p-value <0.001), consistently underscoring MDWTR-D5's superior adsorption performance. These findings highlight MDWTR-D5's potential as an effective adsorbent for phosphorus removal in wastewater treatment, emphasizing its applicability in environmental remediation strategies.

Employing Deep Neural Networks and High‐Throughput Computing for the Recognition and Prediction of Vein‐Like Structures

In this investigation, convolutional neural networks (CNNs) are leveraged to engineer a simple segmentation and recognition algorithm specialized for the delineation of complex, network‐like morphologies—often termed “vein‐like structures (VLSs)”—in scanning electron microscopy (SEM) imagery. These intricate formations frequently appear during the nitriding treatment of medium‐ to high‐carbon alloy steels. To navigate the multifaceted characteristics of such architectures, CNN‐based methodologies are synergized with high‐throughput thermodynamic computations via Thermo‐Calc. This integration aims to quantify both the theoretical upper bounds and the actual values of the VLSs. By establishing deep neural network models for both theoretical upper bounds and actual measurements, the gap between thermodynamics and thermokinetics in the nitriding process is bridged. Applying this amalgamated predictive schema to 8Cr4Mo4V steel, a groundbreaking departure from conventional paradigms that exclusively depend on thermodynamic calculation‐based diffusion models is effectuated. The emergent model yields transformative implications for the metallurgical sector, paving the way for the refinement of future nitriding algorithms and enhancements in nitriding methodologies.

Universal Diffusion in Coulomb Crystals.

Diffusion coefficients for crystallized Coulomb plasmas are essential microphysics input for modeling white dwarf cores and neutron star crusts but are poorly understood. In this work we present a model for diffusion in Coulomb crystals. We show that melting and diffusion follow the same universal scaling such that diffusion is independent of screening. Our simulations show, contrary to prevailing wisdom, that the formation of vacancies is not suppressed by the large pressure. Rather, vacancy formation and hole diffusion is the dominant mode of self-diffusion in Coulomb crystals.

Facile synthesis of zeolitic‐imidazole framework‐67 (ZIF‐67) for the adsorption of indigo carmine dye

AbstractThis study aims to synthesize ZIF‐67 for the adsorption of indigo carmine (I.C.) dye. ZIF‐67 materials were synthesized under different conditions and the synthesis condition of ZIF‐67 giving the best adsorption efficiency was obtained. In batch adsorption studies, pH, adsorbent amount, contact time, initial dye concentration, and temperature effects were investigated. Adsorption data were applied to Langmuir, Freundlich, and Temkin isotherm models. The best correlation value was obtained in Langmuir isotherm (R2 = 0.999). The adsorption of I.C. dye with ZIF‐67 occurred in a monolayer with a homogeneous surface according to Langmuir isotherm. The highest adsorption capacity value of ZIF‐67 metal–organic framework for I.C. dye was found as 370 mg/g. Kinetic model studies were also performed and pseudo first‐order, pseudo second‐order, and intraparticle diffusion models were used. Pseudo second‐order was the best fitting kinetic model to the data (R2 = 0.997). Adsorption mechanisms of ZIF‐67 for I.C. dye include electrostatic interaction, π–π interaction, and hydrogen bonding. Thermodynamic studies showed that I.C. removal by ZIF‐67 is an exothermic and spontaneous process. ZIF‐67 was characterized by X‐ray diffraction (XRD), Fourier transform infrared spectroscopy (FT‐IR), field emission scanning electron microscope (FE‐SEM), and N2 adsorption–desorption. The surface area of synthesized ZIF‐67 was found 1493 m2 g−1. The salt concentration effect on I.C. adsorption by ZIF‐67 was also investigated, showing an increase in adsorption capacities with increasing salt concentration. The regeneration study was carried out and after 5 cycles, the reduction in efficiency was 21%.

AptaDiff: de novo design and optimization of aptamers based on diffusion models.

Aptamers are single-stranded nucleic acid ligands, featuring high affinity and specificity to target molecules. Traditionally they are identified from large DNA/RNA libraries using $in vitro$ methods, like Systematic Evolution of Ligands by Exponential Enrichment (SELEX). However, these libraries capture only a small fraction of theoretical sequence space, and various aptamer candidates are constrained by actual sequencing capabilities from the experiment. Addressing this, we proposed AptaDiff, the first in silico aptamer design and optimization method based on the diffusion model. Our Aptadiff can generate aptamers beyond the constraints of high-throughput sequencing data, leveraging motif-dependent latent embeddings from variational autoencoder, and can optimize aptamers by affinity-guided aptamer generation according to Bayesian optimization. Comparative evaluations revealed AptaDiff's superiority over existing aptamer generation methods in terms of quality and fidelity across four high-throughput screening data targeting distinct proteins. Moreover, surface plasmon resonance experiments were conducted to validate the binding affinity of aptamers generated through Bayesian optimization for two target proteins. The results unveiled a significant boost of $87.9\%$ and $60.2\%$ in RU values, along with a 3.6-fold and 2.4-fold decrease in KD values for the respective target proteins. Notably, the optimized aptamers demonstrated superior binding affinity compared to top experimental candidates selected through SELEX, underscoring the promising outcomes of our AptaDiff in accelerating the discovery of superior aptamers.

Mix-DDPM: Enhancing Diffusion Models through Fitting Mixture Noise with Global Stochastic Offset

Denoising diffusion probabilistic models (DDPM) have shown impressive performance in various domains as a class of deep generative models. In this article, we introduce the mixture noise-based DDPM (Mix-DDPM), which considers the Markov diffusion posterior as a Gaussian mixture model. Specifically, Mix-DDPM randomly selects a Gaussian component and then adds the chosen Gaussian noise, which can be demonstrated as a more efficient way to perturb the signals into a simple known distribution. We further define the reverse probabilistic model as a parameterized Gaussian mixture kernel. Due to the intractability in calculating the KL divergence between Gaussian mixture models, we derive a variational bound to maximize the likelihood, offering a concise formulation for optimizing the denoising model and valuable insights for designing the sampling strategies. Our theoretical derivation highlights that Mix-DDPM need only shift image which requires the inclusion of a global stochastic offset in both the diffusion and reverse processes , which can be efficiently implemented with just several lines of code. The global stochastic offset effectively fits a Gaussian mixture distribution enhancing the degrees of freedom of the entire diffusion model. Furthermore, we present three streamlined sampling strategies that interface with diverse fast dedicated solvers for diffusion ordinary differential equations, boosting the efficacy of image representation in the sampling phase and alleviating the issue of slow generation speed, thereby enhancing both efficiency and accuracy. Extensive experiments on benchmark datasets demonstrate the effectiveness of Mix-DDPM and its superiority over the original DDPM.

Noise suppression in photon-counting computed tomography using unsupervised Poisson flow generative models

Deep learning (DL) has proven to be important for computed tomography (CT) image denoising. However, such models are usually trained under supervision, requiring paired data that may be difficult to obtain in practice. Diffusion models offer unsupervised means of solving a wide range of inverse problems via posterior sampling. In particular, using the estimated unconditional score function of the prior distribution, obtained via unsupervised learning, one can sample from the desired posterior via hijacking and regularization. However, due to the iterative solvers used, the number of function evaluations (NFE) required may be orders of magnitudes larger than for single-step samplers. In this paper, we present a novel image denoising technique for photon-counting CT by extending the unsupervised approach to inverse problem solving to the case of Poisson flow generative models (PFGM)++. By hijacking and regularizing the sampling process we obtain a single-step sampler, that is NFE = 1. Our proposed method incorporates posterior sampling using diffusion models as a special case. We demonstrate that the added robustness afforded by the PFGM++ framework yields significant performance gains. Our results indicate competitive performance compared to popular supervised, including state-of-the-art diffusion-style models with NFE = 1 (consistency models), unsupervised, and non-DL-based image denoising techniques, on clinical low-dose CT data and clinical images from a prototype photon-counting CT system developed by GE HealthCare.

Deep diffusion learning of mutual-reflective structured light patterns for multi-body three-dimensional imaging

Simultaneous structured light imaging of multiple objects has become more demanding and widely in many scenarios involving robot operations in intelligent manufacturing. However, it is challenged by pattern aliasing caused by mutual reflection between high-reflective objects. To this end, we propose to learn clear fringe patterns from aliased mutual-reflective observations by diffusion models for achieving high-fidelity multi-body reconstruction in line with typical phase-shift algorithms. Regarding mutual reflection imaging as a formation of adding significant noise, we build a supervised generative learning framework based on diffusion models and then train a self-attention-based deep network with a U-Net-like skip-connected encoder-decoder architecture. We demonstrate the generalization capability of the trained model in fringe pattern recovery and its performance in phase and three-dimensional (3D) shape reconstruction. Both experimental results show that the proposed method has the expected feasibility and accuracy, heralding a promising solution for addressing the current challenge in various multi-body mutual-reflective 3D reconstruction tasks.

Breath-hold calibrated fMRI mapping of absolute cerebral metabolic rate of oxygen metabolism (CMRO2): An assessment of the accuracy and repeatability in a healthy adult population

Abstract We previously introduced a calibrated fMRI framework that utilises respiratory modulation with only a single gas (CO2) to map the grey matter (GM) cerebral metabolic rate of oxygen consumption (CMRO2). The method decouples and estimates the cerebral blood volume (CBV) and the oxygen extraction fraction (OEF) from a single measure of the maximum BOLD modulation. The method links the two parameters of interest with a model of oxygen diffusion from capillaries to mitochondria which incorporates the cerebral blood flow (CBF). Here, we apply this framework to gas-free breath-hold calibrated fMRI (bhc-fMRI), where simultaneous BOLD and ASL acquisitions are combined with modulation of arterial CO2 through repeated breath-holding. The accuracy and repeatability of the method is assessed in 33 healthy volunteers at rest and during continuous visual stimulation. Average GM OEF estimated from bhc-fMRI was 0.37 ± 0.04, indicating a small bias of 0.04 (with limits of agreement from -0.11 to 0.12) compared to the whole brain OEF of 0.32 ± 0.07 estimated from sagittal sinus using T2 Relaxation Under Spin Tagging (TRUST). The within-session repeatability of GM estimates were moderate to good for OEF, with ICC = 0.75 (0.56–0.87) and good to excellent for CMRO2, with ICC = 0.88 (0.74–0.94). An ROI analysis in the visual cortex found an average CBF increase of 16%, a CMRO2 increase of 12%, and an OEF decrease of 3% during the visual stimulation. The bhc-fMRI measurement of CMRO2 is simple to implement, has comparable accuracy and repeatability to existing gas-based methods, and is sensitive to modulations in metabolism during functional hyperaemia.

Multimodal Fusion for Talking Face Generation Utilizing Speech-Related Facial Action Units

Talking face generation is to synthesize a lip-synchronized talking face video by inputting an arbitrary face image and corresponding audio clips. The current talking face model can be divided into four parts: visual feature extraction, audio feature processing, multimodal feature fusion, and rendering module. For the visual feature extraction part, existing methods face the challenge of complex learning task with noisy features, this article introduces an attention-based disentanglement module to disentangle the face into Audio-face and Identity-face using speech-related facial action unit (AU) information. For the multimodal feature fusion part, existing methods ignore not only the interaction and relationship of cross-modal information but also the local driving information of the mouth muscles. This study proposes a novel generative framework that incorporates a dilated non-causal temporal convolutional self-attention network as a multimodal fusion module to enhance the learning of cross-modal features. The proposed method employs both audio- and speech-related facial AUs as driving information. Speech-related AU information can facilitate more accurate mouth movements. Given the high correlation between speech and speech-related AUs, we propose an audio-to-AU module to predict speech-related AU information. Finally, we present a diffusion model for the synthesis of talking face images. We verify the effectiveness of the proposed model on the GRID and TCD-TIMIT datasets. An ablation study is also conducted to verify the contribution of each component. The results of quantitative and qualitative experiments demonstrate that our method outperforms existing methods in terms of both image quality and lip-sync accuracy. Code is available at https://mftfg-au.github.io/Multimodal_Fusion/.

Flash combustion prepared Sm and Co doped Sr hexaferrite for environmental applications

AbstractNanotechnology is offering solutions to water contamination issues, as new techniques are needed to improve the removal of harmful compounds from water bodies. Despite previous reviews on this topic, nanotechnology is paving the way for more effective water treatment methods. Understanding the substitute influence of divalent Co2+ and rare earth elements Sm3+ on the structure, magnetic, and removal efficiency of hexagonal ferrites requires an understanding of a sequence of SrFe12O19, SrFe11.5Co0.5O19, Sr0.95Sm0.05Fe12O19, and Sr0.95Sm0.05Fe11.5Co0.5O19 M-type hexagonal ferrites were prepared using the flash technique. The XRD examination revealed that the crystallized material formed a single M-type hexagonal phase. The characteristics of M-type hexagonal ferrites include absorption bands with low wavenumbers in the FTIR curves between 400 to 1000 cm−1. There was a variation in magnetic characteristics with the replacement of Sm3+ and Co2+ doping, possibly due to the spin canting impact created by rare earth Sm3+ and Co2+ ions. The goal of the research is to explore the potential of doping magnetic hexaferrites and its influence in wastewater treatment. Various parameters, such as pH and contact duration, that influence the adsorption of lead ions from aqueous solutions were also examined. At pH 7 and 25 °C after 70min, the maximal removal efficiency of the Sr0.95Sm0.05Fe11.5Co0.5O19 was found to be 99%. Magnetic separation was carried out by applying an external magnetic field using a permanent magnet. The strong magnetization of the ferrites (51–58 emu/g) enabled the rapid separation of the magnetic particles from the solution, with over 95% of the ferrite particles being recovered within 10 to 70 min. The Freundlich isotherm model fitted all the isotherm data. Adsorption kinetics were explained by the pseudo-first-order, pseudo-second order, and intraparticle diffusion models. The investigated samples’ adsorption capacity remained efficient till 5 cycles.

Efficient image generation with Contour Wavelet Diffusion

The burgeoning field of image generation has captivated academia and industry with its potential to produce high-quality images, facilitating applications like text-to-image conversion, image translation, and recovery. These advancements have notably propelled the growth of the metaverse, where virtual environments constructed from generated images offer new interactive experiences, especially in conjunction with digital libraries. The technology creates detailed high-quality images, enabling immersive experiences. Despite diffusion models showing promise with superior image quality and mode coverage over GANs, their slow training and inference speeds have hindered broader adoption. To counter this, we introduce the Contour Wavelet Diffusion Model, which accelerates the process by decomposing features and employing multi-directional, anisotropic analysis. This model integrates an attention mechanism to focus on high-frequency details and a reconstruction loss function to ensure image consistency and accelerate convergence. The result is a significant reduction in training and inference times without sacrificing image quality, making diffusion models viable for large-scale applications and enhancing their practicality in the evolving digital landscape.

Porcine ex-vivo intestinal mucus has age-dependent blocking activity against transmissible gastroenteritis virus

Transmissible gastroenteritis virus (TGEV) causes high mortality in young piglets (< 3 days of age). With aging, the susceptibility/morbidity/mortality rates drop. We previously hypothesized that the age-related changes in the intestinal mucus could be responsible for this resistance. Hence, this study investigated the effect of porcine intestinal mucus from 3-day and 3-week-old pigs on the free mobility of the virulent TGEV Miller strain, and on the infection in swine testicle (ST) cells. Single particle tracking (SPT) revealed that TGEV had significantly higher diffusion coefficients in 3-day mucus compared to 3-week mucus. TGEV and charged and uncharged control nanoparticles diffused freely in 3-day mucus but were hindered by 3-week mucus in the diffusion model; TGEV mimicked the diffusion behavior of negatively charged carboxylated particles. Inoculation of ST cells with TGEV in the presence of 3-week mucus resulted in a significantly lower average number of infected cells (30.9 ± 11.9/5 fields) compared with 3-day mucus (84.6 ± 16.4/5 fields). These results show that 3-week mucus has a significant TGEV-blocking activity compared to 3-day mucus in free diffusion and infection of the underlying susceptible cells. Additionally, a label-free proteomics analysis revealed an increased expression of mucin 13, known for negatively regulating the tight junctions in intestinal epithelium, in 3-day-old pigs. In 3-week-old pigs, a higher expression of mucin 2, a type of secreted mucin which is known for inhibiting coronavirus infection, was observed. Concludingly, this study demonstrated a protective effect of 3-week mucus against viral infections.

Identity Diffuser: Preserving Abnormal Region of Interests While Diffusing Identity

To release medical images that can be freely used in downstream processes while maintaining their utility, it is necessary to remove personal features from the images while preserving the lesion structures. Unlike previous studies that focused on removing lesion structures while preserving the individuality of medical images, this study proposes and validates a new framework that maintains the lesion structures while diffusing individual characteristics. In this framework, we apply local differential privacy techniques to provide theoretical guarantees of privacy protection. Additionally, to enhance the utility of protected medical images, we perform denoising using a diffusion model on the noise-contaminated medical images. Numerous chest X-rays generated by the proposed method were evaluated by physicians, revealing a trade-off between the level of privacy protection and utility. In other words, it was confirmed that increasing the level of personal information protection tends to result in relatively lower utility. This study potentially enables the release of certain types of medical images that were previously difficult to share.