Home sleep apnea tests (HSATs) have emerged as alternatives to in-laboratory polysomnography (PSG), but Type IV HSATs often show limited diagnostic performance. This study clinically validates SleepAI, a novel remote digital health system that applies AI algorithms to raw oximetry data for automated sleep staging and obstructive sleep apnea (OSA) diagnosis. SleepAI algorithms were trained on over 10,000 PSG recordings. The system consists of a wearable oximeter connected via Bluetooth to a mobile app transmitting raw data to a cloud-based platform for AI-driven analysis. Clinical validation was conducted in 53 subjects with suspected OSA, who used SleepAI for three nights at home and one night in a sleep centre alongside PSG. SleepAI's apnea-hypopnea index (AHI) estimates and three-class sleep staging (Wake, REM, NREM) were compared to PSG references. For OSA severity classification (non-OSA, mild, moderate, severe), SleepAI achieved an overall accuracy of 89%, with F1-scores of 1.0, 1.0, 0.9, and 0.88, respectively. The three-stage sleep classification achieved a Cohen's kappa of 0.75. Night-to-night AHI variability showed that 37.5% of participants experienced a one-level severity change across nights at home. No significant differences in sleep metrics were found between the first and subsequent nights at home, indicating no sleep disturbance by SleepAI. These findings support the SleepAI system as a promising and scalable alternative to existing Type IV HSATs, with the potential to address key clinical gaps by improving diagnostic accuracy and accessibility.

Self-supervised learning (SSL) has enabled Vision Transformers (ViTs) to learn robust representations from large-scale natural image datasets, enhancing their generalization across domains. In retinal imaging, foundation models pretrained on either natural or ophthalmic data have shown promise, but the benefits of in-domain pretraining remain uncertain. To investigate this, we benchmark six SSL-pretrained ViTs on seven digital fundus image (DFI) datasets totaling 70,000 expert-annotated images for the task of moderate-to-late age-related macular degeneration (AMD) identification. Our results show that iBOT pretrained on natural images achieves the highest out-of-distribution generalization, with AUROCs of 0.80-0.97, outperforming domain-specific models, which achieved AUROCs of 0.78-0.96 and a baseline ViT-L with no pretraining, which achieved AUROCs of 0.68-0.91. These findings highlight the value of foundation models in improving AMD identification and challenge the assumption that in-domain pretraining is necessary. Furthermore, we release BRAMD, an open-access dataset (n=587) of DFIs with AMD labels from Brazil.

The Hillel Yaffe Age Related Macular Degeneration (HYAMD) dataset is a longitudinal collection of 1,560 Digital Fundus Images (DFIs) from 325 patients examined at the Hillel Yaffe Medical Center (Hadera, Israel) between 2021 and 2024. The dataset includes an AMD cohort of 147 patients (aged 54-94) with varying stages of AMD and a control group of 190 diabetic retinopathy (DR) patients (aged 24-92). AMD diagnoses were based on comprehensive clinical ophthalmic evaluations, supported by Optical Coherence Tomography (OCT) and OCT angiography. Non-AMD DFIs were sourced from DR patients without concurrent AMD, diagnosed using macular OCT, fluorescein angiography, and widefield imaging. HYAMD provides gold-standard annotations, ensuring AMD labels were assigned following a full clinical assessment. Images were captured with a DRI OCT Triton (Topcon) camera, offering a 45 deg field of view and 1960 x 1934 pixel resolution. To the best of our knowledge, HYAMD is the first open-access retinal dataset from an Israeli sample, designed to support AMD identification using machine learning models.

The deformable in situ spatiotemporal bioelectronic chemical patch (DinST-BioC-Patch) plays a pivotal role in elucidating the pathogenesis of major diseases, yet it faces challenges in intricate bio-interface adaptability, sensitivity, and spatiotemporal signal processing. In this work, we introduce a bioelectronic chemical patch as initially reported and selected cardiac troponin I (cTnI) as the indicator to validate the practicability of DinST-BioC-Patch for in situ spatiotemporal chemical analysis. Excellent bio-adaptability and suitable mechanical property, ensure the feasibility of DinST-BioC-Patch on intricate bio-interfaces. Utilizing a double amplification and fluorinated amino acid-doping strategy, the doped organic electrochemical transistor array biosensors (7 mm × 7 mm with 64 channels), enabled DinST-BioC-Patch to achieve ultrasensitive detection of cTnI down to 8.71 pg/mL. For the spatiotemporal analysis, DinST-BioC-Patch precisely captured cTnI’s time-dependent dynamics (T = 0, 15, 30, and 45 min) across time and monitored its spatial distribution on the epicardium of the myocardial infarction area (0.7 cm × 0.7 cm). Notably, the DinST-BioC-Patch platform, in its initial demonstration, the spatiotemporal dynamics of biochemical cTnI in the epicardium of Sprague Dawley (SD) rats after myocardial infarction (MI) can be observed, indicating that biochemical cTnI spreads from the center of the infarction area to its periphery. It is anticipated that the DinST-BioC-Patch not only elucidate valuable insights into comprehensive biomolecular dynamics mechanisms but also serve as a platform for potential advancements in the fields of diagnosis, intervention, and prognostic monitoring of significant diseases.

The microvasculature, a complex network of small blood vessels, connects systemic circulation with local tissues, facilitating the nutrient and oxygen exchange that is critical for homeostasis and organ function. Engineering these structures is paramount for advancing tissue regeneration, disease modeling, and drug testing. However, replicating the intricate architecture of native vascular systems—characterized by diverse vessel diameters, cellular constituents, and dynamic perfusion capabilities—presents significant challenges. This complexity is compounded by the need to precisely integrate biomechanical, biochemical, and cellular cues. Recent breakthroughs in microfabrication, organoids, bioprinting, organ-on-a-chip platforms, and in vivo vascularization techniques have propelled the field toward faithfully replicating vascular complexity. These innovations not only enhance our understanding of vascular biology but also enable the generation of functional, perfusable tissue constructs. Here, we explore state-of-the-art technologies and strategies in microvascular engineering, emphasizing key advancements and addressing the remaining challenges to developing fully functional vascularized tissues.

Synthetic multicellular systems have great potential for performing complex tasks, including multi-signal detection and computation through cell-to-cell communication. However, engineering these systems is challenging, requiring precise control over the cell concentrations of distinct members and coordination of their activity. Here, we develop a bacterial consortia-based biosensor for Heme and Lactate, wherein members are coupled through a global shared quorum-sensing signal that simultaneously controls the activity of the diverse biosensing strains. The multicellular system incorporates a gene circuit that computes the minimum between each biosensor’s activity and the shared signal. We evaluate three consortia configurations: one where the shared signal is externally supplied, another directly produced via an inducible gene circuit, and a third generated through an incoherent feedforward loop (IFFL) gene circuit. Among these configurations, the IFFL system, which maintains the shared signal at low and stable levels over an extended period, demonstrates improved performance and robustness against perturbations in cell populations. Finally, we examine these coupled consortia to monitor Lactate and Heme in humanized fecal samples for diagnostics.

Deep neural networks have led to significant advancements in microscopy image generation and analysis. In single-molecule localization based super-resolution microscopy, neural networks are capable of predicting fluorophore positions from high-density emitter data, thus reducing acquisition time, and increasing imaging throughput. However, neural network-based solutions in localization microscopy require intensive human intervention and computation expertise to address the compromise between model performance and its generalization. For example, researchers manually tune parameters to generate training images that are similar to their experimental data; thus, for every change in the experimental conditions, a new training set should be manually tuned, and a new model should be trained. Here, we introduce AutoDS and AutoDS3D, two software programs for reconstruction of single-molecule super-resolution microscopy data that are based on Deep-STORM and DeepSTORM3D, that significantly reduce human intervention from the analysis process by automatically extracting the experimental parameters from the imaging raw data. In the 2D case, AutoDS selects the optimal model for the analysis out of a set of pre-trained models, hence, completely removing user supervision from the process. In the 3D case, we improve the computation efficiency of DeepSTORM3D and integrate the lengthy workflow into a graphic user interface that enables image reconstruction with a single click. Ultimately, we demonstrate superior performance of both pipelines compared to Deep-STORM and DeepSTORM3D for single-molecule imaging data of complex biological samples, while significantly reducing the manual labor and computation time.

Driven by recent advancements in nanofabrication techniques, single-molecule sensing and manipulations in nanofluidic devices are rapidly evolving. These sophisticated biosensors have already had significant impacts on basic research as well as on applications in molecular diagnostics. The nanoscale dimensions of these devices introduce new physical phenomena by confining the biomolecules in at least one dimension, creating effects such as biopolymer linearization, stretching, and separation by mass that are utilized to enhance the biomolecule sensing resolutions. At the same time, the suppressed diffusional motion allows for better single-molecule SNR (signal-to-noise ratio) sensing over time. In particular, nanofluidic devices based on nanochannels have been established as promising technologies for the linearization of ultralong genomic DNA molecules and for optical genome mapping, opening a window to directly observe and infer genome organization. More recently, nanochannels have shown promising capabilities for single-molecule protein sizing, separation, and identification. Consequently, this technology is attracting remarkable interest for applications in single-molecule proteomics. In this review, we discuss the recent advancements of nanochannel-based technologies, focusing on their applications for single-molecule sensing and the characterization of a wide range of biomolecules.

Humans excel at adjusting movements and acquiring new skills through feedback corrections and predictive control, yet how these feedback-feedforward computations evolve in the motor system remains unclear. We investigated this process by examining how humans learned a novel, continuous visuomotor mirror reversal (MR) tracking task over multiple days. Using a frequency-dependent system-identification approach and responses to cursor perturbations, we dissociated feedback-driven corrections from predictive feedforward adjustments. Our findings reveal two distinct learning pathways: early learning relies on rapid, corrective feedback at lower frequencies, while feedforward control gradually emerges at higher frequencies, compensating for feedback limitations. These findings suggest that motor learning involves a dynamic interplay between feedback and feedforward control, providing mechanistic insights into sensorimotor learning, with implications for optimizing motor skill acquisition and neurorehabilitation strategies

The future has always captivated human imagination, with efforts to assess disease prognosis dating back to ancient Egyptian times: “If the heart trembles, has little power and sinks, the disease is advancing … and death is near … ” (Papyrus Ebers, circa 1550 BC). However, the risks of relying on predictions were also acknowledged in antiquity: “…The prophecy has been taken from the prophets and given to the fools and babies instead … ” (Babylonian Talmud: Baba Bathra 12b). Recent advancements in medicine have significantly enhanced prognostic accuracy. The availability of comfortable and reliable wearable devices capable of measuring cardiac, neuronal, and other physiological signals, combined with sophisticated machine learning algorithms designed to interpret these signals in real time (Davoodi et al., 2024; Elul et al., 2024; Fira et al., 2024; Kerr et al., 2024), marks the dawn of a new era in preventive medicine—the age of real-time prediction.