IEEE Trans Pattern Anal Mach Intell. 2025 Jan 9;PP. doi: 10.1109/TPAMI.2025.3527498. Online ahead of print.

ABSTRACT

Existing methods for integerized training speed up deep learning by using low-bitwidth integerized weights, activations, gradients, and optimizer buffers. However, they overlook the issue of full-precision latent weights, which consume excessive memory to accumulate gradient-based updates for optimizing the integerized weights. In this paper, we propose the first latent weight quantization schema for general integerized training, which minimizes quantization perturbation to training process via residual quantization with optimized dual quantizer. We leverage residual quantization to eliminate the correlation between latent weight and integerized weight for suppressing quantization noise. We further propose dual quantizer with optimal nonuniform codebook to avoid frozen weight and ensure statistically unbiased training trajectory as full-precision latent weight. The codebook is optimized to minimize the disturbance on weight update under importance guidance and achieved with a three-segment polyline approximation for hardware-friendly implementation. Extensive experiments show that the proposed schema allows integerized training with lowest 4-bit latent weight for various architectures including ResNets, MobileNetV2, and Transformers, and yields negligible performance loss in image classification and text generation. Furthermore, we successfully fine-tune Large Language Models with up to 13 billion parameters on one single GPU using the proposed schema.

PMID:40030978 | DOI:10.1109/TPAMI.2025.3527498

IEEE Trans Neural Netw Learn Syst. 2025 Jan 10;PP. doi: 10.1109/TNNLS.2024.3525264. Online ahead of print.

ABSTRACT

This work focuses on the safety of learning-based control for unknown nonlinear system, considering the stability of learned dynamics and modeling mismatch between the learned dynamics and the true one. A learning-based scheme imposing the stability constraint is proposed in this work for modeling and stable control of unknown nonlinear system. Specifically, a linear representation of unknown nonlinear dynamics is established using the Koopman theory. Then, a deep learning approach is utilized to approximate embedding functions of Koopman operator for unknown system. For the safe manipulation of proposed scheme in the real-world applications, a stable constraint of learned dynamics and Lipschitz constraint of embedding functions are considered for learning a stable model for prediction and control. Moreover, a robust predictive control scheme is adopted to eliminate the effect of modeling mismatch between the learned dynamics and the true one, such that the stabilization of unknown nonlinear system is achieved. Finally, the effectiveness of proposed scheme is demonstrated on the tethered space robot (TSR) with unknown nonlinear dynamics.

PMID:40030974 | DOI:10.1109/TNNLS.2024.3525264

IEEE J Biomed Health Inform. 2025 Jan 10;PP. doi: 10.1109/JBHI.2024.3524642. Online ahead of print.

ABSTRACT

To improve the performance of object recognition under artificial prosthetic vision, this study proposes a two-stage method. The first stage is to extract the saliency and edge Mask of the object (SMP, EMP). Then, the irregular visual information of the object is processed using Irregularity Correction (IC). We design eye-hand coordination tasks and simulate artificial vision with retinal prostheses to validate strategy effectiveness, and select direct pixelation (DP) as a control group. Each subject retained a phosphene map in the same stochastic pattern in all his/her trails. The real-time experimental results showed that the deep saliency-based optimization strategies improved the performance of the subjects when completing tasks, in terms of head movement, recognition accuracy, and response time, and counts for successful small-objects recognition. The subjects have the smallest-scale average head movement (76.53 deg ± 20.75 deg), higher average objects recognition accuracy (91.18% ± 2.52%), and less time for finishing the task (35.71 s ± 8.66 s) and better successful search times of the small target objects (1.35 ± 0.33) under the SMP strategy. When integrating with IC, subjects' average performances have further improved to 63.39 ± 15.38 deg, 94.22% ± 3.94%, 25.76 s ± 6.24 s and 1.05 ± 0.30 respectively, which also significantly outperformed the DP condition. These results indicated that when utilizing the deep-learning-based saliency detection and IC processing, subjects could shorten the searching process and were able to discern the target objects more reliably. This work could be informative to future prosthetic devices considering implementation with the technique of artificial intelligence.

PMID:40030970 | DOI:10.1109/JBHI.2024.3524642

IEEE J Biomed Health Inform. 2025 Jan 7;PP. doi: 10.1109/JBHI.2025.3526837. Online ahead of print.

ABSTRACT

Remote photoplethysmography (rPPG) has recently attracted much attention due to its non-contact measurement convenience and great potential in health care and computer vision applications. Early rPPG studies were mostly developed on self-collected uncompressed video data, which limited their application in scenarios that require long-distance real-time video transmission, and also hindered the generation of large-scale publicly available benchmark datasets. In recent years, with the popularization of high-definition video and the rise of telemedicine, the pressure of storage and real-time video transmission under limited bandwidth have made the compression of rPPG video inevitable. However, video compression can adversely affect rPPG measurements. This is due to the fact that conventional video compression algorithms are not specifically proposed to preserve physiological signals. Based on this, we propose a video compression scheme specifically designed for rPPG application. The proposed approach consists of three main strategies: 1) facial ROI-based computational resource reallocation; 2) rPPG signal preserving bit resource reallocation; and 3) temporal domain up- and down-sampling coding. UBFC-rPPG, ECG-Fitness, and a self-collected dataset are used to evaluate the performance of the proposed method. The results demonstrate that the proposed method can preserve almost all physiological information after compressing the original video to 1/60 of its original size. The proposed method is expected to promote the development of telemedicine and deep learning techniques relying on large-scale datasets in the field of rPPG measurement.

PMID:40030966 | DOI:10.1109/JBHI.2025.3526837

IEEE J Biomed Health Inform. 2025 Jan 7;PP. doi: 10.1109/JBHI.2025.3526727. Online ahead of print.

ABSTRACT

Cardiovascular disease is a leading global cause of death, requiring accurate heart segmentation for diagnosis and surgical planning. Deep learning methods have been demonstrated to achieve superior performances in cardiac structures segmentation. However, there are still limitations in 3D whole heart segmentation, such as inadequate spatial context modeling, difficulty in capturing long-distance dependencies, high computational complexity, and limited representation of local high-level semantic information. To tackle the above problems, we propose a lightweight Pyramid Pooling Transformer-CNN (P2TC) network for accurate 3D whole heart segmentation. The proposed architecture comprises a dual encoder-decoder structure with a 3D pyramid pooling Transformer for multi-scale information fusion and a lightweight large-kernel Convolutional Neural Network (CNN) for local feature extraction. The decoder has two branches for precise segmentation and contextual residual handling. The first branch is used to generate segmentation masks for pixel-level classification based on the features extracted by the encoder to achieve accurate segmentation of cardiac structures. The second branch highlights contextual residuals across slices, enabling the network to better handle variations and boundaries. Extensive experimental results on the Multi-Modality Whole Heart Segmentation (MM-WHS) 2017 challenge dataset demonstrate that P2TC outperforms the most advanced methods, achieving the Dice scores of 92.6% and 88.1% in Computed Tomography (CT) and Magnetic Resonance Imaging (MRI) modalities respectively, which surpasses the baseline model by 1.5% and 1.7%, and achieves state-of-the-art segmentation results. Our code will be released via https://github.com/Countdown229/P2TC.

PMID:40030965 | DOI:10.1109/JBHI.2025.3526727

IEEE J Biomed Health Inform. 2025 Jan 10;PP. doi: 10.1109/JBHI.2025.3527403. Online ahead of print.

ABSTRACT

Adenoid hypertrophy is one of the most common upper respiratory tract disorders during childhood, leading to a range of symptoms such as nasal congestion, mouth breathing and obstructive sleep apnea. Current diagnostic methods, including computerized tomography scans and nasal endoscopy, are invasive or involve ionizing radiation, rendering them unsuitable for long-term assessments. To address these clinical challenges, this paper proposes a novel deep learning approach for the non-invasive detection of adenoid hypertrophy using heartlung sounds. Firstly, we established a heart-lung sound database with corresponding labels indicating adenoid size. Subsequently, we employed three different deep learning tasks to explore the association between heart-lung sounds and adenoid size. In particular, it includes binary classification to distinguish between normal and abnormal cases, four-grade classification to assess the severity of adenoid hypertrophy, and regression models to predict the actual size of the adenoids. The experimental results demonstrate that the deep learning models can effectively predict the condition of adenoid hypertrophy based on heart-lung sounds. In resource-constrained clinical environments, the proposed methods for adenoid hypertrophy automatic detection provide a simple and non-invasive approach, which can reduce healthcare costs and facilitate remote self-screening.

PMID:40030964 | DOI:10.1109/JBHI.2025.3527403

IEEE J Biomed Health Inform. 2025 Jan 7;PP. doi: 10.1109/JBHI.2025.3526806. Online ahead of print.

ABSTRACT

In recent years, the deployment of supervised machine learning techniques for segmentation tasks has significantly increased. Nonetheless, the annotation process for extensive datasets remains costly, labor-intensive, and error-prone. While acquiring sufficiently large datasets to train deep learning models is feasible, these datasets often experience a distribution shift relative to the actual test data. This problem is particularly critical in the domain of medical imaging, where it adversely affects the efficacy of automatic segmentation models. In this work, we introduce DiffuSeg, a novel conditional diffusion model developed for medical image data, that exploits any labels to synthesize new images in the target domain. This allows a number of new research directions, including the segmentation task that motivates this work. Our method only requires label maps from any existing datasets and unlabelled images from the target domain for image diffusion. To learn the target domain knowledge, a feature factorization variational autoencoder is proposed to provide conditional information for the diffusion model. Consequently, the segmentation network can be trained with the given labels and the synthetic images, thus avoiding human annotations. Initially, we apply our method to the MNIST dataset and subsequently adapt it for use with medical image segmentation datasets, such as retinal fundus images for vessel segmentation and MRI images for heart segmentation. Our approach exhibits significant improvements over relevant baselines in both image generation and segmentation accuracy, especially in scenarios where annotations for the target dataset are unavailable during training. An open-source implementation of our approach can be released after reviewing.

PMID:40030962 | DOI:10.1109/JBHI.2025.3526806

Proc Natl Acad Sci U S A. 2025 Mar 11;122(10):e2417943122. doi: 10.1073/pnas.2417943122. Epub 2025 Mar 3.

ABSTRACT

Sperm morphogenesis is a tightly regulated differentiation process, disruption of which leads to sperm malfunction and male infertility. Here, we show that Tex38 knockout (KO) male mice are infertile. Tex38 KO spermatids exhibit excess retention of residual cytoplasm around the head, resulting in abnormal sperm morphology with backward head bending. TEX38 interacts and colocalizes with ZDHHC19, a testis-enriched acyltransferase catalyzing protein S-palmitoylation, at the plasma membrane of spermatids. ZDHHC19 and TEX38 are each downregulated in mouse testes lacking the other protein. TEX38 stabilizes and localizes ZDHHC19 to the plasma membrane of cultured cells and vice versa, consolidating their interdependence. Mice deficient in ZDHHC19 or harboring a C142S mutation that disables the palmitoyltransferase activity of ZDHHC19 display phenotypes resembling those of Tex38 KO mice. Strikingly, ZDHHC19 palmitoylates ARRDC5, an arrestin family protein regulating sperm differentiation. Overall, our findings indicate that TEX38 forms a stable complex with ZDHHC19 at the plasma membrane of spermatids, which governs downstream S-palmitoylation of proteins essential for morphological transformation of spermatids.

PMID:40030029 | DOI:10.1073/pnas.2417943122

PLoS Biol. 2025 Mar 3;23(3):e3003030. doi: 10.1371/journal.pbio.3003030. Online ahead of print.

ABSTRACT

In Drosophila, two interacting adhesion protein families, Defective proboscis responses (Dprs) and Dpr interacting proteins (DIPs), coordinate the assembly of neural networks. While intercellular DIP::Dpr interactions have been well characterized, DIPs and Dprs are often co-expressed within the same cells, raising the question as to whether they also interact in cis. We show, in cultured cells and in vivo, that DIP-α and DIP-δ can interact in cis with their ligands, Dpr6/10 and Dpr12, respectively. When co-expressed in cis with their cognate partners, these Dprs regulate the extent of trans binding, presumably through competitive cis interactions. We demonstrate the neurodevelopmental effects of cis inhibition in fly motor neurons and in the mushroom body. We further show that a long disordered region of DIP-α at the C-terminus is required for cis but not trans interactions, likely because it alleviates geometric constraints on cis binding. Thus, the balance between cis and trans interactions plays a role in controlling neural development.

PMID:40029885 | DOI:10.1371/journal.pbio.3003030

Expert Opin Drug Discov. 2025 Mar 3. doi: 10.1080/17460441.2025.2474661. Online ahead of print.

ABSTRACT

INTRODUCTION: Identifying treatments that can alter the natural history of amyotrophic lateral sclerosis (ALS) is challenging. For years, drug discovery in ALS has relied upon traditional approaches with limited success. Drug repurposing, where clinically approved drugs are reevaluated for other indications, offers an alternative strategy that overcomes some of the challenges associated with de novo drug discovery. Whilst not a new concept, the potential of drug repurposing in ALS is yet to be fully realized.

AREAS COVERED: In this review, the authors discuss the challenge of drug discovery in ALS and specifically examine the potential of drug repurposing for the identification of new effective treatments. The authors consider a broad range of approaches, from screening in experimental models to computational approaches, and outline some general principles for pre-clinical and clinical research to help bridge the translational gap. Literature was reviewed from original publications, press releases and clinical trials.

EXPERT OPINION: Despite the remaining challenges, drug repurposing offers the opportunity to improve therapeutic options for ALS patients. Nevertheless, stringent pre-clinical research will be necessary to identify the most promising compounds while innovative experimental medicine studies will also be paramount to bridge the aforementioned translational gap. The authors further highlight the importance of combining expertise across academia, industry and wider stakeholders, which will be key in the successful delivery of repurposed therapies to the clinic.

PMID:40029669 | DOI:10.1080/17460441.2025.2474661

Med Sci (Paris). 2025 Feb;41(2):173-179. doi: 10.1051/medsci/2025016. Epub 2025 Mar 3.

ABSTRACT

The aim of the MITOMICS project is to establish a clinical database of patients diagnosed with mitochondrial diseases, combined with a « multiomics » integrated approach in order to gain a better understanding of the molecular mechanisms underlying these diseases, and ultimately, to offer better patient care. The MITOMICS project thus contributes to the consolidation of a French "mitochondrial medicine", a notion that deserves to be examined. With the upcoming launch of the fourth national plan for rare diseases, it is an example of the study and management of rare and ultrarare diseases in France. This article traces the emergence of mitochondrial medicine since the early 1960s. It presents its main characteristics (genocentrism, strong techno-dependence), as well as its major technical and theoretical limitations, with a view to developing personalized mitochondrial medicine for the years to come.

PMID:40028956 | DOI:10.1051/medsci/2025016

Clin Pharmacol Ther. 2025 Mar 3. doi: 10.1002/cpt.3623. Online ahead of print.

ABSTRACT

A clinical trial was carried out to investigate the pharmacogenetics of single-dose pravastatin and pitavastatin pharmacokinetics in 173 and 164 healthy participants. Additionally, 96 participants were included from previous pharmacogenetic studies with pravastatin. In a genome-wide meta-analysis of pravastatin including all 269 participants, SLCO1B1 c.521T>C (rs4149056) was associated with increased AUC0-∞ (P = 9.8 × 10-12). Similarly, SLCO1B1 c.521T>C was genome-wide significantly associated with increased AUC0-∞ of pitavastatin (P = 9.7 × 10-15). Candidate gene analyses suggested that participants with increased function SLCO1B1 variants had decreased pravastatin exposure. Furthermore, decreased function CYP2C9 variants may increase pitavastatin and pitavastatin lactone exposure. Compared to participants with normal function SLCO1B1 genotype, the AUC0-∞ of pravastatin was 140% (90% confidence interval: 86-210%; P = 4.7 × 10-8) and 37% (20-56%; P = 1.1 × 10-4) greater in participants with poor and decreased function SLCO1B1 genotype, respectively, while participants with highly increased function SLCO1B1 genotype had a 60% (39-75%; P = 6.0 × 10-4) lower AUC0-∞. The AUC0-∞ of pitavastatin was 153% (100-222%; P = 1.6 × 10-9) and 35% (8-69%; P = 0.027) greater in participants with poor and decreased function SLCO1B1 genotype, respectively, than in those with normal function SLCO1B1 genotype. Participants with intermediate metabolizer CYP2C9 genotype had 18% (3-34%; P = 0.046) greater AUC0-∞ of pitavastatin than those with normal metabolizer CYP2C9 genotype. This study demonstrates the important role of SLCO1B1 in pravastatin and pitavastatin pharmacokinetics and suggests that CYP2C9 variants also affect the pharmacokinetics of pitavastatin.

PMID:40029062 | DOI:10.1002/cpt.3623

Heliyon. 2025 Feb 8;11(4):e42572. doi: 10.1016/j.heliyon.2025.e42572. eCollection 2025 Feb 28.

ABSTRACT

BACKGROUND: Volvariella volvacea is a highly nutritious edible mushroom grown mainly in Southeast Asian countries. However, the low yield of V. volvacea has discouraged farmers from engaging in its production.

OBJECTIVE: The study was conducted to observe the improvement of V. volvacea yield depending on various physiochemical parameters of V. volvacea growth.

METHODS: The parameters tested in this study include the weight of the substrate, i.e., 2 kg (W1) and 6 kg (W2); the surface area of the substrate: A1 (1218 cm2), A2 (1530 cm2) and A3 (2000 cm2); and four different substrate formulations (F1, F2, F3 and F4).

RESULTS: Substrate weight and surface area were found to be important, but not critical, factors in determining fruiting bodies formation, total fungal mass, and BE rate. However, the formulation media showed a significant contribution that could help in the induction of fruiting bodies. According to the results, the culture medium with a mixture of EFB substrate and black soil showed the highest BE percentage of 17.75 % (at optimised substrate weights = 2 kg).

CONCLUSION: The results of this study can be used as a reference for further studies to improve the cultivation of V. volvacea, especially when EFB fibres are used as the main substrate. Future studies to identify genes involved in the formation of fruiting bodies are strongly recommended.

PMID:40028610 | PMC:PMC11869036 | DOI:10.1016/j.heliyon.2025.e42572

Cardiovasc Ther. 2025 Feb 21;2025:5711316. doi: 10.1155/cdr/5711316. eCollection 2025.

ABSTRACT

Background: Cardiovascular diseases (CVDs) encompass a group of diseases that affect the heart and/or blood vessels, making them the leading cause of global mortality. In our study, we performed proteome-wide Mendelian randomization (MR) and colocalization analyses to identify novel therapeutic protein targets for CVDs and evaluate the potential drug-related protein side effects. Methods: We conducted a comprehensive proteome-wide MR study to assess the causal relationship between plasma proteins and the risk of CVDs. Summary-level data for 4907 circulating protein levels were extracted from a large-scale protein quantitative trait loci (pQTL) study involving 35,559 individuals. Additionally, genome-wide association study (GWAS) data for CVDs were extracted from the UK Biobank and the Finnish database. Colocalization analysis was utilized to identify causal variants shared between plasma proteins and CVDs. Finally, we conducted a comprehensive phenome-wide association study (PheWAS) using the R10 version of the Finnish database. This study was aimed at examining the potential drug-related protein side effects in the treatment of CVDs. A total of 2408 phenotypes were included in the analysis, categorized into 44 groups. Results: The research findings indicate the following associations: (1) In coronary artery disease (CAD), the plasma proteins A4GNT, COL6A3, KLC1, CALB2, KPNA2, MSMP, and ADH1B showed a positive causal relationship (p-fdr < 0.05). LAYN and GCKR exhibited a negative causal relationship (p-fdr < 0.05). (2) In chronic heart failure (CHF), PLG demonstrated a positive causal relationship (p-fdr < 0.05), while AZGP1 displayed a negative causal relationship (p-fdr < 0.05). (3) In ischemic stroke (IS), ALDH2 exhibited a positive causal relationship (p-fdr < 0.05), while PELO showed a negative causal relationship (p-fdr < 0.05). (4) In Type 2 diabetes (T2DM), the plasma proteins MCL1, SVEP1, PIP4K2A, RFK, HEXIM2, ALDH2, RAB1A, APOE, ANGPTL4, JAG1, FGFR1, and MLN demonstrated a positive causal relationship (p-fdr < 0.05). PTPN9, SNUPN, VAT1, COMT, CCL27, BMP7, and MSMP displayed a negative causal relationship (p-fdr < 0.05). Colocalization analysis conclusively identified that AZGP1, ALDH2, APOE, JAG1, MCL1, PTPN9, PIP4K2A, SNUPN, and RAB1A share a single causal variant with CVDs (PPH3 + PPH4 > 0.8). Further phenotype-wide association studies have shown some potential side effects of these nine targets (p-fdr < 0.05). Conclusions: This study identifies plasma proteins with significant causal associations with CVDs, providing a more comprehensive understanding of potential therapeutic targets. These findings contribute to our knowledge of the underlying mechanisms and offer insights into potential avenues for treatment.

PMID:40026415 | PMC:PMC11870767 | DOI:10.1155/cdr/5711316

Iran J Med Sci. 2025 Feb 1;50(2):98-111. doi: 10.30476/ijms.2024.101852.3450. eCollection 2025 Feb.

ABSTRACT

BACKGROUND: Next-Generation Sequencing (NGS) methods specifically Whole-Exome Sequencing (WES) have demonstrated promising findings with a high accuracy of 91%-99% in Pharmacogenomics (PGx). A PGx-based panel can be utilized to minimize adverse drug reactions (ADRs) and maximize the treatment efficacy. Remarkably, Cancer Pain Management (CPM) is a cutting-edge concept in modern medicine. Thus, this study aimed to investigate the WES results by a PGx-based panel containing genes involved in Pain, Anti-inflammatory, and Immunomodulating agents (PAIma) signaling pathways.

METHODS: A total of 200 unrelated Iranians (100 western and 100 northern) were included. 100 WES results were analyzed through the PAIma panel. After DNA extraction, 100 samples were genotyped by Multiplex-Amplification-Refractory Mutation System (ARMS) PCR. A primary in silico investigation performed on 128 candidate genes through Protein-Protein Interactions (PPIs) and Gene-miRNA Interactions (GMIs) via the STRING database, and miRTargetLink2, respectively. Additionally, Enrichment Analysis (EA) was applied to find the unknown interplays among these three major pathways by Enrichr.

RESULTS: 55,590 annotations through 21 curated pathways were filtered, 900 variants were found, and 128 genes were refined. Finally, 54 candidate variants (48 non-synonymous single nucleotide variants (nsSNVs), 2 stop-gained, 1 frameshift, and 3 splicing) remained.

CONCLUSION: Conclusively, six potentially actionable variants including rs1695 (GSTP1), rs628031 (SLC22A1), rs17863778 (UGT1A7), rs16947 (CYP2D6), rs2257401 (CYP3A7), and rs2515641 (CYP2E1) had the most deviations among Iranians, compared with the reference genome, which should be genotyped for drug prescribing. Remarkably, PPIs, GMIs, and EA revealed potential risks of carcinogenesis and cancer phenotypes resulting from PAIma pathways genes.

PMID:40026294 | PMC:PMC11870856 | DOI:10.30476/ijms.2024.101852.3450

FEBS J. 2025 Mar 3. doi: 10.1111/febs.70050. Online ahead of print.

ABSTRACT

Cystic fibrosis (CF) is a genetic disease caused by mutations in the CF transmembrane conductance regulator (CFTR) gene, leading to chronic, unresolved inflammation of the airways due to uncontrolled recruitment of polymorphonuclear leukocytes (PMNs). Evidence indicates that CFTR loss-of-function, in addition to promoting a pro-inflammatory phenotype, is associated with an increased risk of developing cancer, suggesting that CFTR can exert tumor-suppressor functions. Three-dimensional (3D) in vitro culture models, such as the CF lung airway-on-a-chip, can be suitable for studying PMN recruitment, as well as events of cancerogenesis, that is epithelial cell invasion and migration, in CF. To gather insight into the pathobiology of CFTR loss-of-function, we generated CFTR-knockout (KO) clones of the 16HBE14o- human bronchial cell line by CRISPR/Cas9 gene editing, and performed a comparative proteomic analysis of these clones with their wild-type (WT) counterparts. Systematic signaling pathway analysis of CFTR-KO clones revealed modulation of inflammation, PMN recruitment, epithelial cell migration, and epithelial-mesenchymal transition. Using a latest-generation organ-on-a-chip microfluidic platform, we confirmed that CFTR-KO enhanced PMN recruitment and epithelial cell invasion of the endothelial layer. Thus, a dysfunctional CFTR affects multiple pathways in the airway epithelium that ultimately contribute to sustained inflammation and cancerogenesis in CF.

PMID:40029006 | DOI:10.1111/febs.70050

Phys Occup Ther Pediatr. 2025 Mar 3:1-16. doi: 10.1080/01942638.2025.2469567. Online ahead of print.

ABSTRACT

AIMS: To evaluate the effects of telerehabilitation (TG) compared with an unsupervised home exercise training program (HG) on muscle function, physical activity (PA), and sleep in children with cystic fibrosis (CF).

METHODS: Thirty children with CF (mean age = 10.2 ± 1.9 years) were randomly allocated to TG or HG. The exercise protocol was applied thrice a week for six weeks in the TG via Skype. The same exercises were sent in an exercise booklet to the HG, and phone contact was made once a week. Muscle function (one-minute sit-to-stand (1-min STS), sit-up, pushup, squat, and plank tests)), PA (Physical Activity Questionnaire for Older Children), and sleep (Epworth Sleepiness Scale (ESS) and Pediatric Sleep Questionnaire (PSQ)) were assessed before and after the 6-week study period.

RESULTS: The 1-min STS significantly improved in the TG compared with the HG (p ≤ .001, ηp2 = 0.474). The sit-up (p = .005, ηp2 = 0.247), pushup (p = .002, ηp2 = 0.180), squat (p = .002, ηp2 = 0.284), and plank (p < .001, ηp2 = 0.360) test scores were significantly improved in the TG compared to the HG. No significant changes between groups were seen for PA (p = .261, ηp2 = 0.045), ESS (p = .160, ηp2 = 0.069), or PSQ (p = .763, ηp2 = 0.003).

CONCLUSION: Children who received TG improved muscle function more than children who received an HG. The effectiveness of longer term TG programs should be investigated in children with CF.

PMID:40028780 | DOI:10.1080/01942638.2025.2469567

Genet Med Open. 2025 Jan 7;3:101962. doi: 10.1016/j.gimo.2025.101962. eCollection 2025.

ABSTRACT

PURPOSE: Newborn screening identifies rare diseases that result from the recessive inheritance of pathogenic variants in both copies of a gene. Long-read genome sequencing (LRS) is used for identifying and phasing genomic variants, but further efforts are needed to develop LRS for applications using low-yield DNA samples.

METHODS: In this study, genomic DNA with high molecular weight was obtained from 2 cystic fibrosis patients, comprising a whole-blood sample (CF1) and a newborn dried blood spot sample (CF2). Library preparation and genome sequencing (30-fold coverage) were performed using 20 ng of DNA input on both the PacBio Revio system and the Illumina NovaSeq short-read sequencer. Single-nucleotide variants, small indels, and structural variants were identified for each data set.

RESULTS: Our results indicated that the genotype concordance between long- and short-read genome sequencing data was higher for single-nucleotide variants than for small indels. Both technologies accurately identified known pathogenic variants in the CFTR gene (CF1: p.(Met607_Gln634del), p.(Phe508del); CF2: p.(Phe508del), p.(Ala455Glu)) with complete concordance for the polymorphic poly-TG and consecutive poly-T tracts. Using PacBio read-based haplotype phasing, we successfully determined the allelic phase and identified compound heterozygosity of pathogenic variants at genomic distances of 32.4 kb (CF1) and 10.8 kb (CF2).

CONCLUSION: Haplotype phasing of rare pathogenic variants from minimal DNA input is achieved through LRS. This approach has the potential to eliminate the need for parental testing, thereby shortening the time to diagnosis in genetic disease screening.

PMID:40027236 | PMC:PMC11869909 | DOI:10.1016/j.gimo.2025.101962

Obstet Med. 2025 Feb 26:1753495X251319588. doi: 10.1177/1753495X251319588. Online ahead of print.

ABSTRACT

BACKGROUND: Elexacaftor/tezacaftor/ivacaftor (ETI) was made available to eligible women in September 2020 by NHS Scotland.

METHODS: Retrospective data collection for the 13 pregnancies in women taking ETI from the West of Scotland Adult Cystic Fibrosis Unit, September 2020-December 2023.

RESULTS: Mean pre-pregnancy FEV1 was 2.26L, 70% predicted (range 1.25-3.19); (38-86% predicted). Mean FEV1 post-pregnancy was 2.29L, 71% predicted (range 1.49-3.40); (45-92% predicted). The mean age at conception (29 years) and mean percentage predicted FEV1 (70%) were higher than in other UK studies. Two pregnancies resulted in miscarriage, the remaining 11 pregnancies resulted in a live birth. Seven women had a pulmonary exacerbation of CF during pregnancy. Three of four women with FEV1 < 60% predicted had uncomplicated pregnancies with no pulmonary exacerbations.

CONCLUSION: We demonstrate that people with CF and varying spectrums of lung disease who take CFTR modulators can have uncomplicated pregnancies with positive lung function outcomes.

PMID:40027072 | PMC:PMC11866333 | DOI:10.1177/1753495X251319588

Curr Med Sci. 2025 Mar 3. doi: 10.1007/s11596-025-00017-3. Online ahead of print.

ABSTRACT

OBJECTIVE: To develop and validate a deep neural network (DNN) model for diagnosing Parkinson's Disease (PD) using handwritten spiral and wave images, and to compare its performance with various machine learning (ML) and deep learning (DL) models.

METHODS: The study utilized a dataset of 204 images (102 spiral and 102 wave) from PD patients and healthy subjects. The images were preprocessed using the Histogram of Oriented Gradients (HOG) descriptor and augmented to increase dataset diversity. The DNN model was designed with an input layer, three convolutional layers, two max-pooling layers, two dropout layers, and two dense layers. The model was trained and evaluated using metrics such as accuracy, sensitivity, specificity, and loss. The DNN model was compared with nine ML models (random forest, logistic regression, AdaBoost, k-nearest neighbor, gradient boost, naïve Bayes, support vector machine, decision tree) and two DL models (convolutional neural network, DenseNet-201).

RESULTS: The DNN model outperformed all other models in diagnosing PD from handwritten spiral and wave images. On spiral images, the DNN model achieved accuracies of 41.24% over naïve Bayes, 31.24% over decision tree, and 27.9% over support vector machine. On wave images, the DNN model achieved accuracies of 40% over naïve Bayes, 36.67% over decision tree, and 30% over support vector machine. The DNN model demonstrated significant improvements in sensitivity and specificity compared to other models.

CONCLUSIONS: The DNN model significantly improves the accuracy of PD diagnosis using handwritten spiral and wave images, outperforming several ML and DL models. This approach offers a promising diagnostic tool for early PD detection and provides a foundation for future work to incorporate additional features and enhance detection accuracy.

PMID:40029495 | DOI:10.1007/s11596-025-00017-3