Continuous vital sign monitoring ensures early detection, prevents intensive care unit (ICU) admissions, and improves patient outcomes. Continuous heart rate (HR) monitoring methods often require direct skin contact, which can lead to patient discomfort. The rising popularity of ballistocardiography (BCG) offers a promising, noncontact solution for continuous vital sign monitoring with improved patient comfort. This study aims to develop and validate a novel HR measurement algorithm leveraging convolutional neural networks (CNNs) and BCG signals for accurate, noncontact, and continuous HR monitoring. By integrating time-domain peak detection with short-time Fourier transform and CNN models, the proposed approach seeks to enhance HR measurement accuracy across diverse health care settings. The study follows the Food and Drug Administration (FDA)'s Good Machine Learning Practice guidelines and evaluates the algorithm's robustness, generalizability, and clinical applicability through extensive testing on a diverse dataset, ensuring improved patient comfort and early detection of clinical deterioration. The proposed algorithm combines time-domain peak detection with short-time Fourier transform and CNNs to enhance HR measurement from BCG signals. The CNN model developed was trained on 129,976 data points from 373 participants (HR range: 36-230 bpm), including ICU patients, and was tuned on 75,970 data points from 192 participants (HR range: 46-169 bpm), with HR obtained from clinical-grade electrocardiography devices to improve generalizability. The algorithm was tested on 70,211 data points from 205 participants, including ICU patients, across 5 independent studies to demonstrate robust performance against diverse settings, demographics, and comorbidities. The methodology is in compliance with the FDA's Good Machine Learning Practice for Medical Device Development: Guiding Principles. The algorithm achieved a mean absolute error of under 3 bpm and a detection rate exceeding 80%, underscoring its robustness. The Bland-Altman analysis indicates high accuracy with a minimal bias of 0.25 and limits of agreement within 8.59 bpm. Additionally, the Pearson correlation coefficient of 0.97 from the Deming regression further demonstrates strong alignment with reference HR measurements, reinforcing its precision and reliability for clinical applications. This CNN-based algorithm presents a robust solution for contactless HR monitoring, addressing the limitations of prior methods in noise management and adaptability. Its demonstrated accuracy, particularly in real-world, noisy clinical environments, highlights its potential for broad application in patient monitoring and improved comfort.
In this paper, we present a modified nonmonotone line search algorithm that employs a variable parameter to control the degree of nonmonotonicity. This modification enhances both the probability of identifying the global minimum and the rate of convergence. Within the framework of alternating nonnegative least squares (ANLS), we propose a hybrid algorithm that employs either the modified nonmonotone projected Barzilai-Borwein method and the block coordinate descent method to address the subproblems in each iteration. To further accelerate convergence, we integrate a technique that allows for a larger step size. Under mild assumptions, we establish the global convergence of the algorithm. Numerical experiments conducted on both synthetic and real datasets demonstrate that the proposed algorithm is efficient for nonnegative matrix factorization (NMF) and outperforms other state-of-the-art methods.
Age assessment plays an important role in forensic sciences to aid in criminal or civil matters. Third molars continue developing during the legal age of adulthood. Dental age estimation based on radiographic examination is an operator-dependent procedure. The aim was to test the performance of supervised machine learning (SML) to classify individuals according to the threshold of 18 years using the third molar index (I3m). Panoramic radiographs (n = 597) of 52% males and 48% females between 13 and 26 years were used. Three Convolutional Neural Networks were compared to segment the third molar. Several machine learning algorithms were tested with a 10-fold cross-validation approach to identify the best age classification algorithm. The SML age estimation model was compared to the expert. Attention U-Net achieved the highest segmentation scores, and the k-nearest neighbors (KNN) showed the best scores of classification algorithms. The SML (KNN) scored sensitivity = 77.77%, specificity = 96%, and AUC = 0.87 in males. In females, sensitivity = 77.4%, specificity = 86.7%, and AUC = 0.82. The expert scores were: sensitivity = 70.4%, specificity = 100%, AUC = 0.85 for male, and for female: sensitivity = 54.8%, specificity = 93.3%, AUC = 0.74. SML classified males and females below and above the legal age with high accuracy compared to manual methods. The presented work suggests that AI, using pixel-level analysis, can detect subtle cues beyond human perception. This signals a possible departure from traditional age estimation indices.
Background and ObjectivesArtificial intelligence (AI) is increasingly used in healthcare decision-making to improve clinical efficiency and patient risk assessment. However, emerging evidence suggests these tools may disadvantage older adults from lower socioeconomic backgrounds, potentially worsening health disparities. This scoping review examines how AI-driven healthcare decision tools affect health outcomes among socioeconomically disadvantaged older adults.MethodsFollowing the Arksey and O'Malley framework and the PRISMA-ScR guidelines, I searched PubMed, CINAHL, Embase, Scopus, and Web of Science for peer-reviewed studies published between 2000 and 2025. Studies were screened using a Population, Concept, and Context (PCC) framework.ResultsThirty-three studies met inclusion criteria, revealing four themes: algorithmic bias in training data, socioeconomic and digital access barriers, differential outcomes by income and race, and governance gaps in AI deployment.DiscussionCurrent AI healthcare tools risk reinforcing socioeconomic health disparities among older adults, highlighting the need for equitable design, policy oversight, and regulatory reforms.
Optically Induced Dielectrophoresis (ODEP) has been widely used in biomedical applications such as cell sorting and cell capture because of its operational flexibility and low cellular damage. However, existing automated ODEP methods often lack effective control of non-target cells, which may reduce manipulation performance in multicellular environments. To address this problem, this study proposes an automated cell manipulation method integrating ODEP, image processing, static optical traps and the A-star algorithm. Cells are first identified and localized from microscopic images. Non-target cells are then constrained by static optical traps and treated as static obstacles during path planning. Based on the detected cell positions, obstacle avoiding paths are generated and converted into executable optical patterns. Experiments were performed using yeast cells under a frequency of 1 kHz, a voltage of 2 V, and a light spot velocity of 5 μm/s. The results showed that the target cells followed the planned obstacle-avoiding paths and reached the designated destinations, while the non-target cells remained confined within their corresponding static optical-trap regions. The success rates of repeated single-cell directed transport and two-cell convergence experiments were approximately 90% and 80%, respectively. Non-target cells showed mean displacements of 0.68 μm (n = 30, single-cell) and 3.74 μm (n = 14, two-cell). Although larger in the two-cell experiment, none escaped optical traps or interfered with target manipulation. This work demonstrates the feasibility of combining static optical confinement with automated path planning for cell manipulation in multicellular fields of view and provides a basis for further studies involving denser and more complex cellular environments.
A patient presents with symptomatic bradycardia and high-grade atrioventricular block and receives a dual-chamber pacemaker with left bundle branch area pacing lead. Post procedure, there is concern for dislodgement of the ventricular lead after an electrocardiogram shows a missing ventricular paced beat. After additional device interrogation and review of the device manufacturer algorithms, it was determined that what was thought to be loss-of-capture or loss-of-sensing was coincidental artifact and normal device function. The highest risk for pacemaker lead dislodgement is immediately after implantation, and a 12-lead electrocardiogram should be reviewed post implantation. Understanding device manufacturer algorithms can help interpret unusual artifacts and findings in paced rhythms.
Background: Identifying patients at risk of opioid overdose in healthcare settings is critical, yet evidence on predictive models and their performance to predict imminent opioid overdose remains limited.Objective: We compared classical and Machine Learning (ML) survival models to predict 30-day overdose risk following a first opioid-related diagnosis to determine whether algorithmic complexity improves clinical decision support.Methods: We conducted a prospective cohort study using longitudinal Electronic Health Record (EHR) data from the All of Us Research Program (v8; May 2018-October 2023). Adults with first recorded opioid-related diagnosis were followed 30 days for incident overdose. We compared Cox proportional hazards, Weibull Accelerated Failure Time (AFT) (parametric), and four ML survival models (Elastic-Net Cox, Random Survival Forest, Gradient-Boosted Survival Trees, Survival SVM).Results: Among 14,737 individuals, 560 overdoses occurred within 30 days (cumulative incidence 3.8%, 95% CI 3.5-4.1). Prior overdose (aHR 5.77, 95% CI 4.64-7.17) and opioid misuse (aHR 3.72, 95% CI 2.94-4.70) were the strongest predictors and consistently drove risk across models. Test-set C-indices ranged from 0.708 to 0.745, with Weibull AFT and Survival SVM performing best, although performance was similar across models. Calibration was acceptable, with predicted risks closely aligned with observed 30-day risks across low and moderate risk strata, and high-risk groups demonstrated substantially lower overdose-free survival.Conclusions: Patients with prior overdose or opioid misuse diagnosis represent a high-yield target for proactive clinical intervention. Simple, interpretable approaches may be sufficient for identifying high-risk patients, as added algorithmic complexity did not meaningfully improve imminent overdose prediction.
Ferroptosis has been increasingly implicated in the pathophysiology of atrial fibrillation (AF). Pentoxifylline (PTX), a methylxanthine derivative, has shown potential therapeutic benefits in cardiovascular diseases; however, its unique role in ferroptosis-associated AF remains unclear. This study aimed to elucidate the molecular mechanisms through which PTX may exert therapeutic effects on ferroptosis-related AF, using a multifaceted approach integrating network pharmacology, bioinformatics, and experimental validation. Two transcriptomic datasets, GSE41177 and GSE79768, were retrieved from the Gene Expression Omnibus (GEO) to identify differentially expressed genes (DEGs) in AF. Ferroptosis-related genes (FRGs) were collected from the FerrDb database. PTX-associated targets were predicted using Super-PRED and SwissTargetPrediction. Overlapping DEGs and predicted PTX targets were intersected with FRGs to identify potential pharmacological targets. Candidate genes were further refined through protein-protein interaction (PPI) network construction and five topological algorithms (Degree, Maximum Neighborhood Component [MNC], Maximal Clique Centrality [MCC], Edge Percolated Component [EPC], and Closeness). Genes exhibiting consistent expression patterns in both GEO datasets were defined as key genes. Diagnostic value was assessed using receiver operating characteristic (ROC) curves. The immune infiltration landscape and correlations with key genes were analyzed via the CIBERSORT algorithm and Spearman correlation. Molecular docking was performed, and PyMOL was used to assess binding affinities between PTX and key gene-encoded proteins. In addition, regulatory networks involving non-coding RNAs and key genes were predicted. Single-cell RNA sequencing (scRNA-seq) was applied to determine cell-type-specific gene expression. Finally, the therapeutic effects of PTX and the underlying ferroptosis-related molecular pathways were evaluated both in an acetylcholine (ACh)-CaCl2-induced AF rat model and in angiotensin II (AngII)-stimulated HL-1 cardiomyocytes. From a total of 10,511 DEGs and 315 predicted PTX targets, 87 overlapping pharmacological targets were identified. Thirteen of these overlapped with known FRGs. Among them, PIK3CA and TLR4 emerged as key genes of interest based on PPI network centrality and consistent expression across datasets. Immune profiling revealed significant differences in six immune cell types between AF and control samples, with activated dendritic cells and follicular helper T cells negatively correlated with key gene expression. Molecular docking indicated favorable binding affinities between PTX and both PIK3CA (-4.33 kcal/mol) and TLR4 (-3.72 kcal/mol). Further analysis identified 23 microRNAs (miRNAs) predicted to target PIK3CA and TLR4, of which 21 miRNAs interacted with 22 long non-coding RNAs (lncRNAs), suggesting a complex regulatory network. scRNA-seq analysis revealed enriched PIK3CA expression in mast cells and elevated TLR4 expression in neutrophils and monocytes/macrophages, suggesting involvement of immune-related mechanisms. In vivo, PTX significantly reduced AF susceptibility, shortened AF duration, and attenuated atrial structural remodeling. In AngII-stimulated HL-1 cardiomyocytes, PTX markedly suppressed intracellular Fe2+ accumulation. In both models, these protective effects coincided with downregulation of TLR4 and upregulation of PIK3CA, implicating modulation of ferroptosis-related pathways as the underlying mechanism. This study identifies PIK3CA and TLR4 as pivotal genes in the ferroptosis-associated molecular network of AF and as potential therapeutic targets of PTX. These findings support the potential of PTX to mitigate AF by regulating ferroptosis through these targets, providing a preclinical mechanistic basis for the potential value of PTX in AF management.
Coronary artery disease (CAD) is a common cardiovascular disorder strongly associated with glutamine metabolism. This study seeks to identify novel glutamine metabolism-related gene markers in CAD. Based on GEO datasets and glutamine metabolism-related genes, hub genes were pinpointed through WGCNA combined with LASSO, SVM-RFE, and random forest algorithms. Immune cell infiltration was estimated with the CIBERSORT algorithm. To confirm the expression and roles of hub genes, ox-LDL-stimulated HUVECs were analyzed in culture. MYBPC3 and TRIM47 emerged as candidate diagnostic signatures for CAD. Associations between MYBPC3/TRIM47 and immune cell infiltration were revealed. Upregulation of MYBPC3 and TRIM47 in CAD was confirmed in both GSE113079 and ox-LDL-treated HUVECs. Knockdown of TRIM47 or MYBPC3 promoted ox-LDL-induced proliferation of HUVECs, alleviated cellular apoptosis, and suppressed the expression of pro-inflammatory cytokines (IL-1β, IL-6, TNF-α) as well as M1 macrophage polarization. MYBPC3 and TRIM47 may serve as candidate diagnostic signatures for CAD.
Artificial intelligence (AI) is increasingly evolving from a research technology into a tool for everyday clinical practice. While early applications primarily focused on medical image analysis, generative AI systems and large language models are now available for a wide range of clinical and administrative tasks. These include medical documentation, literature review, guideline-based knowledge management, patient communication, and workflow optimization. At the same time, diagnostic and therapeutic applications continue to evolve. AI-assisted systems support radiological and pathological image interpretation, risk stratification, and clinical decision-making processes. Despite considerable opportunities, important limitations remain. AI hallucinations, algorithmic bias, data protection requirements, and regulatory considerations necessitate continuous human oversight and critical evaluation. Therefore, the long-term success of AI will depend not only on technological performance but also on its responsible integration into existing healthcare structures. This review provides a practice-oriented overview of current and future AI applications in urology and discusses opportunities, limitations, and prerequisites for safe implementation in clinical practice and hospital care. Künstliche Intelligenz (KI) entwickelt sich zunehmend von einer Forschungstechnologie zu einem Werkzeug des klinischen Alltags. Während frühe Anwendungen v. a. auf die Analyse medizinischer Bilddaten fokussiert waren, stehen heute generative KI-Systeme und große Sprachmodelle für vielfältige Aufgaben in Klinik und Praxis zur Verfügung. Hierzu zählen die Erstellung medizinischer Dokumente, die Unterstützung bei Literaturrecherchen, die Aufbereitung von Leitlinienwissen, die Patientenkommunikation sowie die Optimierung administrativer Prozesse. Parallel hierzu entwickeln sich diagnostische und therapeutische Anwendungen weiter. KI-gestützte Verfahren unterstützen die Auswertung radiologischer und pathologischer Daten, die Risikostratifizierung sowie die klinische Entscheidungsfindung. Trotz erheblicher Potenziale bestehen relevante Herausforderungen. Halluzinationen generativer KI-Systeme, algorithmische Verzerrungen, Datenschutzanforderungen sowie regulatorische Vorgaben erfordern eine kritische Bewertung und kontinuierliche ärztliche Kontrolle. Der nachhaltige Erfolg von KI wird daher nicht allein von der Leistungsfähigkeit einzelner Modelle abhängen, sondern maßgeblich von deren verantwortungsvoller Implementierung in bestehende Versorgungsstrukturen. Der vorliegende Beitrag gibt einen praxisorientierten Überblick über aktuelle und zukünftige Einsatzmöglichkeiten von KI in der Urologie und diskutiert Chancen, Grenzen und Voraussetzungen für einen sicheren Einsatz in Klinik und Praxis.
Obstructive sleep apnea (OSA) is a significant medical and social problem with worldwide importance. The following research article aims to demonstrate that nurse involvement can contribute to the identification of OSA risk using an online survey. For the study, an online questionnaire was developed comprising 15 questions focused on the main symptoms of sleep apnea. The survey is intended for patients aged 30 and older, regardless of sex. The survey card was posted online for 1 month. According to the International Classification of Sleep Disorders, there are over 100 disorders. Insufficient and poor-quality sleep can cause several health disorders: fatigue, inability to concentrate, irritability, changes in metabolism, sexual disorders, depression, and anxiety. Sleep apnea is often characterized by loud snoring followed by periods of breathing cessation. Eventually, the decrease or pause in breathing signals the person to wake up, and the awakening is accompanied by loud snoring or gasping. The nurse developed an algorithm/protocol to determine the risk of OSA and to specify the professional behavior she should have toward the patient. Nurses can administer the online survey and, if there is an established risk for OSA, refer patients to sleep medicine specialists for investigation and treatment, following a developed algorithm/protocol. Sleep apnea is a complex medical problem, and the nurse can be part of the team.
The paper presents a clinical case of appendiceal mucinous adenocarcinoma with invasion into the right ovary, imitating an ovarian tumor. The primary tumor was diagnosed only after a pathological evaluation of the surgical material with differential immunohistochemical assay after laparotomy and removal of the ovarian tumor and appendectomy. Difficulties in establishing an accurate diagnosis were caused by extensive invasion of the appendiceal tumor into the ovary. This disease requires increased oncological alertness and a clear algorithm for differential diagnosis of mucinous tumors of the appendix and ovary. Приведено клиническое наблюдение муцинозной аденокарциномы червеобразного отростка с прорастанием в правый яичник, имитирующей опухоль яичника. Первичный очаг опухоли диагностирован только после прижизненного патолого-анатомического исследования операционного материала с выполнением дифференциального иммуногистохимического исследования после лапаротомии и удаления опухоли яичника и аппендэктомии. Трудности в постановке точного диагноза вызваны обширным прорастанием опухоли аппендикса в яичник. Данное заболевание требует повышенной онкологической настороженности и четкого алгоритма дифференциальной диагностики муцинозных опухолей червеобразного отростка и яичника.
This article addresses the issue of reduced-order modeling and finite-dimensional guaranteed cost output tracking (GCOT) control design with the fuselage's elastic vibration suppression for a flexiblevariable wingspan aircraft (FVWA). The rigid-flexible coupled system dynamics of the FVWA are represented by ordinary differential equations (ODEs) and the beam equation. Via an integration of deep learning (DL) and sparse identification of nonlinear dynamics (SINDy), a reduced-order modeling framework that employs DL for dimensionality reduction and utilizes SINDy for governing dynamics identification is proposed to surmount the difficulty caused by the rigid-flexible complex dynamics. Then, a finite-dimensional fuzzy flight control algorithm is developed via the Takagi-Sugeno (T-S) fuzzy control technique and the integral control to achieve the goal of flight command tracking and the fuselage's elastic vibration suppression for the FVWA. Meanwhile, an upper bound of the quadratic flight performance index is also provided. Finally, numerical simulation results are presented to illustrate the effectiveness and superiority of the theoretical method.
The search for efficient solid-state hydrogen storage materials is essential for advancing clean energy technologies. In this work, Li-based perovskite hydrides (LiBH3, LiCuH3, LiMgH3, and LiSiH3) are studied using first-principles calculations. All compounds show negative formation energies, indicating thermodynamic stability. LiBH3, LiCuH3, and LiMgH3 are dynamically stable, while LiSiH3 is unstable due to imaginary phonon modes. Mechanical analysis confirms stability for all systems, with LiCuH3 and LiMgH3 showing higher rigidity. Electronic properties reveal metallic behavior for LiBH3, LiCuH3, and LiSiH3, while LiMgH3 is semiconducting with a 2.59 eV band gap. Hydrogen storage capacities are 14.56 wt% for LiBH3, 4.11 wt% for LiCuH3, 8.82 wt% for LiMgH3, and 7.95 wt% for LiSiH3. LiCuH3 also shows the lowest desorption temperature (382.7 K), indicating favorable kinetics. All calculations were performed within the framework of density functional theory (DFT) using the Quantum ESPRESSO package. The exchange-correlation energy was treated using the generalized gradient approximation in the Perdew-Burke-Ernzerhof (GGA-PBE) form, and electron-ion interactions were described using ultrasoft pseudopotentials. Valence configurations included Li (2s1), H (1s1), B (2s22p1), Cu (3d9.54s1.5), Mg (3s2), and Si (3s23p2). Convergence tests determined a plane-wave cutoff energy of 55 Ry and a Monkhorst-Pack k-point mesh of 9 × 9 × 9, ensuring total energy convergence within 1-2 meV/atom. Structural optimizations were carried out using the Broyden-Fletcher-Goldfarb-Shanno (BFGS) algorithm with convergence thresholds of 1 × 10⁻6 eV/atom for energy and 1 × 10⁻3 eV/Å for forces. Phonon dispersion, mechanical properties, and optical responses were evaluated using the thermo_pw package integrated with Quantum ESPRESSO.
This work presents a series-stacked 2-deck RRAM crossbar array with hierarchical multi-mode operation enabled by a shared mid electrode. A 2 × 16 × 16 stacked crossbar array is fabricated, where the mid electrode electrically connects two vertically integrated resistive switching layers, simultaneously serving as the top electrode of the lower layer and the bottom electrode of the upper layer. This architecture enables bias-selective access to independent single-layer operation (1F and 2F) as well as electrically coupled serial operation (1F + 2F) within the same cell footprint. The shared mid electrode plays a key role in controlling voltage distribution and interlayer interaction, expanding the operational space beyond simple density scaling. Using an incremental step pulse with verify algorithm (ISPVA), all modes exhibit stable multilevel conductance modulation up to 6-bit resolution (64 states) with clear state separation and retention exceeding 104 s, while maintaining endurance over 100 switching cycles. System-level evaluation using a VGG-based CNN for CIFAR-10 classification achieves inference accuracies of 93.37%, 93.38%, and 93.37% for the 1F, 2F, and 1F + 2F modes, respectively. The serial configuration also enables logic-in-memory functionality and demonstrates strong physical unclonable function (PUF) characteristics with near-ideal uniformity (∼50%) and high entropy.
Parkinson's disease (PD) is a neurodegenerative disorder increasingly associated with gut microbiota alterations, yet the mechanisms by which microbial metabolites influence PD remain unclear. Here, we applied an integrative computational and experimental strategy to identify key gut microbial metabolites and host genes potentially involved in PD. Differentially abundant gut microbes were obtained from the gutMDisorder database and their corresponding metabolites from gutMGene, with predicted protein targets generated using the Similarity Ensemble Approach. Transcriptomic data from PD brain tissues were analyzed to identify differentially expressed genes, which were intersected with metabolite targets, followed by enrichment and protein-protein interaction analyses. Three machine learning algorithms were applied for gene prioritization, while molecular docking evaluated metabolite-gene binding affinities and ProTox3.0 predicted toxicity and blood-brain barrier permeability. In vitro assays further assessed the functional effects of 3-indolepropionic acid in a rotenone-induced SH-SY5Y cell model. Our analyses identified 44 PD-associated microbial taxa linked to 77 metabolites and 905 predicted target genes, with 29 overlapping differentially expressed genes enriched in synaptic signaling and dopaminergic pathways. Dopamine receptor D2 (DRD2) emerged as a central hub gene, with strong docking interactions predicted for two indole metabolites, 3-(1H-indol-3-yl)propanoate and 3-indolepropionic acid. Functional validation showed that 3-indolepropionic acid improved cell viability, reduced apoptosis, and preserved DRD2 expression under neurotoxic stress. Together, these findings suggest that specific gut microbial metabolites may modulate host dopaminergic signaling via DRD2, offering new insights into the microbiota-brain axis and potential targets for further PD research.
Accurate and scalable discrimination of closely related genetic variants at single-nucleotide resolution remains challenge in molecular diagnostics, particularly under multiplexed conditions. Herein, we present a mechanistically guided Color-Coding Strategy for Multiple Variants Based on Single-Mutation-Responsive Strand-Displacement Padlock Probes that enables combinatorial color-coded identification of genetic variants with minimal parallel reactions. By integrating multiple isothermal amplification reactions (n) with multichannel strand displacement probes (m), the system establishes a theoretical n-dimensional coding framework capable of resolving up to ∏ i = 1 n m i variants in a single assay. Each variant is uniquely encoded by an "n-color codon" and automatically decoded via a programmable algorithm. Using SARS-CoV-2 as a model, we identified up to 15 variants using only three reaction tubes across RNA, synthetic DNA, pseudovirus, and clinical nasopharyngeal swabs (n = 76). In clinical evaluation, the assay achieved 100% positive agreement with RT-qPCR for SARS-CoV-2-positive specimens and successfully assigned variant identities to 72 of 76 samples (94.7%). The platform achieves a reduced reaction number, experimental complexity, and cost compared to conventional approaches. Owing to its modular and programmable design, this strategy is readily adaptable to emerging variants, demonstrating its potential for pathogen surveillance, and genetic variant analysis.
Wide-complex tachycardia poses diagnostic and therapeutic challenges, particularly when class IC antiarrhythmic toxicity mimics ventricular tachycardia (VT). An 85-year-old woman with a dual-chamber pacemaker, paroxysmal atrial fibrillation on flecainide, and chronic kidney disease presented with dyspnea and a wide-complex tachycardia initially diagnosed as VT. Intravenous amiodarone was administered. Initial device interrogation revealed ventricular sensing at 496 ms cycle length without atrial tracking. Ventricular overdrive pacing demonstrated that the atrial cycle length did not accelerate to the pacing rate, confirming an independent atrial tachycardia at 551 ms and excluding atrioventricular reentrant tachycardia and VT with retrograde conduction. The Ventricular Intrinsic Preference algorithm subsequently demonstrated intact atrioventricular conduction, proving the wide QRS was aberrancy. Recognition of potential amiodarone-flecainide interaction prompted drug cessation and sodium bicarbonate therapy, leading to clinical recovery. This case demonstrates how pacemaker diagnostics can differentiate flecainide-induced aberrancy from VT and highlights the important drug-drug interaction between amiodarone and flecainide via CYP450 inhibition. In patients with pacemakers presenting with wide-complex tachycardia, device interrogation and knowledge of programmed parameters are helpful to differentiate supraventricular tachycardia with aberrancy from VT.
Breast implant-associated anaplastic large cell lymphoma is a mature CD30-positive T-cell lymphoma that develops around the implant as a peri-implant fluid collection confined by a fibrous capsule, and less commonly as a tumor with infiltrative growth. The disease was recognized as a distinct clinico-pathological entity in 2022 in the 5th edition of the WHO Classification of Tumors of Hematopoietic and Lymphoid Tissues. This work analyzed current understanding of the disease: epidemiological data, etiopathogenetic aspects, and diagnostic algorithms. The article discusses the association of the disease with textured implants, predominantly manufactured by Allergan Biocell (USA). The main pathogenetic mechanisms are reviewed: chronic inflammation, the role of bacterial biofilms, and JAK/STAT signaling pathway activation. Current approaches to diagnosis and staging are described. The necessity of establishing a national registry of BIA-ALCL cases in the Russian Federation and developing domestic clinical guidelines for optimizing patient management is emphasized. A description of a clinical case of BIA-ALCL is presented, one of the few registered cases in the Russian Federation. Имплант-ассоциированная анапластическая крупноклеточная лимфома молочной железы — это фенотипически зрелая CD30-позитивная T-клеточная лимфома, которая возникает вокруг импланта в виде скопления клеточной суспензии, ограниченного фиброзной капсулой, и реже — как солидная опухоль с инфильтрирующим ростом. В 5-м издании Классификации ВОЗ опухолей гемопоэтической и лимфоидной тканей опухоль рассматривается как отдельное заболевание в группе анапластических крупноклеточных лимфом. В работе проанализированы современные представления о заболевании: эпидемиологические сведения, этиопатогенетические аспекты и алгоритмы диагностики заболевания. В статье обсуждается связь заболевания с текстурированными имплантами преимущественно производства фирмы Allergan Biocell (США). Рассмотрены основные механизмы патогенеза: хроническое воспаление, роль бактериальных биопленок и активация JAK/STAT сигнального пути. Описаны современные подходы к диагностике и определению стадии заболевания. Отмечена необходимость создания национального регистра случаев имплант-ассоциированной анапластической крупноклеточной лимфомы в РФ и разработки отечественных клинических рекомендаций для оптимизации ведения пациенток. Представлено описание клинического случая имплант-ассоциированной анапластической крупноклеточной лимфомы молочной железы, одного из немногих, зарегистрированных на территории РФ.
The accurate deconvolution of bulk transcriptomes typically confounds stable physical cell identities with dynamic physiological states, limiting our understanding of complex microenvironmental processes such as ovarian aging. To overcome this, we present DeepMCD, an end-to-end multi-task deep learning framework designed to simultaneously deconvolve cell-type proportions and programmed cell death (PCD) compositional fractions from standard bulk RNA-seq data. By mapping high-dimensional expression profiles into a shared, tokenized latent space, DeepMCD employs a Transformer-based cross-task attention mechanism to explicitly leverage cellular morphological context for calibrating PCD predictions. Concurrently, an adaptive uncertainty-weighting loss ensures balanced optimization, effectively mitigating negative transfer. Extensive benchmarking demonstrates that DeepMCD significantly outperforms state-of-the-art single-task algorithms. Through rigorous ablation and interpretability analyses, we computationally substantiate the biological premise that functional death states are heavily reliant on specific cell-type contexts. Applying DeepMCD to real-world mouse ovarian aging cohorts, we reconstructed a cell-type-specific PCD landscape, bypassing the need for costly single-cell sequencing. Specifically, we identified an age-associated increase in inflammatory and lytic death modalities, together with a strong association between macrophage enrichment and pyroptosis during ovarian aging. Crucially, the DeepMCD-derived PCD fractions exhibit profound divergent correlations with core ovarian fibrosis-related genes. These computationally extracted signatures may serve as cost-effective candidate digital biomarkers for evaluating ovarian fibrosis and reproductive senescence. Ultimately, DeepMCD provides a highly interpretable, robust, and scalable computational tool for bulk RNA-seq data decoding.