Colorectal cancer remains a pressing challenge in global health, necessitating advanced biological models and analytical methodologies. Tumor organoids (tumoroids) have emerged as a compelling platform for cancer research, owing to their capacity to replicate the genetic and structural complexity of human tissues. However, extracting meaningful gene regulatory insights from bulk ribonucleic acid (RNA) sequencing data derived from tumoroids remains nontrivial due to cellular heterogeneity and temporal variation. We propose, for the first time, a comprehensive Bayesian framework to model gene expression dynamics throughout the developmental trajectory of colorectal tumoroids. We introduce a nonparametric Dirichlet process mixture model (DPMM) to cluster genes based on temporal expression patterns and a sparse regression scheme, incorporating Horseshoe+ priors, to construct gene regulatory networks (GRNs) among identified clusters. The proposed approach demonstrates robust performance in capturing high-dimensional relationships, enabling elucidation of key regulatory mechanisms in tumor progression. Our results offer valuable insights for personalized treatment and underscore the utility of Bayesian methods in complex biological systems.

In real-world graphs, node attributes are often incomplete due to acquisition costs or privacy restrictions, reducing representation quality and harming downstream predictions in graph neural networks (GNNs). A common remedy is feature-propagation-based imputation. However, cold-start effects arising from attribute resetting and low-degree nodes impede effective propagation and convergence in these methods. To address these challenges, we propose AttriReBoost (ARB), a propagation-based method that mitigates cold-start issues in attribute-missing graphs. ARB enhances global feature propagation (FP) by redefining initial boundary conditions and strategically integrating virtual edges, thereby improving node connectivity and ensuring stable and efficient convergence. The method supports gradient-free attribute reconstruction with low computational overhead, and we provide a rigorous convergence analysis. Extensive experiments on several real-world benchmark datasets demonstrate the effectiveness of ARB, achieving an average accuracy improvement of 5.11% over state-of-the-art methods. In addition, ARB exhibits remarkable computational efficiency, processing a large-scale graph with 2.44 million nodes in just 16 s on a single GPU. Our code is available at https://github.com/limengran98/ARB.

Single-cell RNA sequencing (scRNA-seq) has emerged as a transformative omics technology for cell type identification in cancer diagnostics, enabling high-throughput parallel generation of cellular-resolution data that revolutionizes precision medicine. However, scRNA-seq data are frequently compromised by prevalent drop-out events due to limitations in sample quality and technical bottlenecks, which cause a large number of the results lost. To address this challenge, we propose a nonlinear imputation via state transition process (NISP) method for the diffusion and imputation of missing values in single-cell sequencing data. Our results demonstrate that the NISP framework effectively preserves nonlinear characteristics inherent to biological state transitions, enabling to recover more than 50% of missing values and remove more than 98% of noise. Therefore, NISP exhibits superior sensitivity in missing value imputation and significantly enhances the structural clarity of post-imputation datasets. Finally, validation using spatiotemporal transcriptomic arrays derived from colorectal cancer organoids further corroborates the capability of NISP to accurately capture the intrinsic manifold structure of cellular states. The result shows the significant potential of NISP in biological applications, notably its pivotal role in elucidating the mechanisms driving tumorigenesis and cancer progression.

Engineered assembloids fabricated from tissue spheroids hold immense promise for developmental biology, disease modeling, and regenerative medicine. However, fabricating heterogeneous assembloids with precise spatial patterning remains a critical bottleneck, often reliant on manual pipetting that lacks scalability and reproducibility. While magnetic microrobotics, which transforms spheroids into controllable robots, offers a non-invasive alternative, it faces a fundamental challenge: the global input of magnetic actuation. A single command moves all robots simultaneously, leading to coupled motion and frequent assembly failures. Here, we present a collaborative control framework that overcomes this limitation by leveraging local constraints and a novel motion decoupling strategy. We reformulate the high-dimensional, coupled multi-robot planning problem into a low-dimensional aggregate space, effectively transforming the assembly task into a dynamic sequential decision problem. This framework, coupled with a graph-based dynamic path optimization algorithm, enables deterministic, collision-free assembly. Experimental validation demonstrates a 78.57% higher success rate and a 33.20% higher assembly efficiency compared to conventional strategies. This work establishes a foundational engineering principle for assembloid fabrication, transitioning the process from a qualitative experiments to a controllable and programmable engineering discipline, thereby unlocking the potential for deterministic construction of complex biological structures.

Quantitative Phase Imaging (QPI) has been shown to complement established fluorescence microscopy as well as objective measurements of morphology and dynamics for cellular tissue studies. However, due to its inherent weak-signal measurements, the coherence of the laser light source, the roughness of the object under test or the complex scattering environment, QPI exhibits various types of complex noise, with Poisson-Gaussian noise and scattering noise being the main noise sources. In recent years, significant advancements have been made in the field of deep learning-based denoising algorithms, which have shown considerable efficacy in the denoising of individual noisy data. However, these algorithms have been observed to be less effective when confronted with other types of noisy data, and lack a unified model that can simultaneously remove complex noise from QPI. The present study proposes a QPI denoising approach based on the denoising diffusion probability model (DDPM). The denoising process of DDPM is comprised of two constituent parts: the forward process, which gradually adds standard Gaussian a priori noise to the original image until the image is completely random; and the reverse diffusion chain, which gradually recovers an undisturbed ‘clean’ image by inference from a given a priori noise, thus eliminating various types of complex noise in QPI. A comparative analysis was conducted between the conventional denoising approach based on BM3D and the deep network denoising algorithm with U-Net as the backbone, and the proposed method was evaluated through experimental validation using simulated Gaussian noise, scattering noise and fluorescence microscopy dataset (FMD). The experimental results demonstrate the superior denoising, detail restoration and generalization performance of the proposed method, signifying its significant potential for practical applications.

Organoids are cell clusters cultured in vitro that maintain the structure and function of the donor organs. They have found important applications in biomedicine, such as drug screening and personalized therapy. However, conventional organoid culture methods lack control of physical properties like size and distribution, leading to increased heterogeneity and very low batch-to-batch reproducibility, which significantly limits their widespread use. Controlling these properties at the microscale is challenging, particularly for fragile fragments, which are the main source for culturing organoids. To address this issue, we present a robotic micromanipulation system that allows operators to select fragments of particular sizes and automatically transfer them into a customized in-situ organoid chip (IOC) for culture. The chip was designed with microwell arrays to uniform the culture environment and facilitate imaging analysis. The transfer of fragments is modeled based on computational fluid dynamics (CFD) and is enabled by designing a robust model predictive control (RMPC) framework. Simulation and experiment results demonstrated the effectiveness of the model and controller. In colorectal cancer organoid culture experiments, our system significantly improved the morphological homogeneity of organoids. Note to Practitioners—Organoids have been demonstrated to be one of the most promising in vitro models. Lacking control of its size and distribution results in significant heterogeneity and low batch-to-batch reproducibility, which limits its wide uses. Here, we report a robotic micromanipulation system that allows operators to select fragments of particular sizes and morphologies and automatically transfer them into a customized organoid chip for culture. The results of colorectal cancer organoids culture experiments verified the effectiveness of our system in reducing the morphological heterogeneity among organoids.

This article studies nonconservative stability conditions of continuous-time switched linear systems under mode-dependent dwell time (MDT). To establish a unified analysis approach for switched systems with stable and/or unstable subsystems, a concept called “dictionary” is introduced to characterize admissible MDT switching sequences. Subsequently, two equivalent nonconservative conditions of the global uniform asymptotic stability (GUAS) are obtained based on quadratic Lyapunov functions (LFs). Moreover, the stability results are transformed into convex conditions for facilitating the controller design. In addition, the developed stability results are applied toL2-gain analysis andH∞controller design for the continuous-time switched linear system subject to external disturbances. Simulations are provided to validate the effectiveness and the superiority over existing results.

Sperm immobilization is a crucial procedure in clinical cell surgery for infertility treatment. Current immobilization is implemented by tapping the sperm tail with a glass micropipette, but its effectiveness is restricted by sperm orientation and ineffective membrane ablation. Ineffective ablation leads to limited release of oocyte activating factors and lowers fertilization rate; and sperm swim in small angles relative to the micropipette tip cannot be tapped due to the risk of damaging the sperm’s genetic materials contained in the sperm head. This paper reports automated piezo-assisted sperm immobilization with enhanced efficacy of cell membrane ablation and sperm orientation control. The designed piezo drill consists of two orthogonal vibration modules to generate controlled micropipette vibration along axial and lateral axes. Through stiffness modeling, the flexure joints guide the motion of the central beam of each vibration module. To achieve sperm orientation control, whirl flow is induced by both axial and lateral vibration of the micropipette tip. To immobilize sperm, only micropipette’s axial vibration is generated to prevent lateral vibration from damaging sperm head. A visual servoing scheme is developed by decoupling sperm wiggling from positioning error for immobilization. Experimental results showed that sperm orientation control by the piezo drill achieved an error of 1.4$^{\circ}$and a time cost of 2.5 s. Visual servoing with sperm wiggling decoupling achieved a positioning error of 1.7$\mu$m. Furthermore, the piezo-assisted sperm immobilization technique led to effective membrane ablation. With membrane-impermeable stains, it took 5.6 s for the immobilized sperm to be stained after piezo-assisted immobilization, significantly less than 49.2 s by conventional micropipette tapping.Note to Practitioners—This work tackled the challenge of ineffective membrane ablation and orientation limit in clinical cell surgeries. Conventional manual immobilization suffers from low membrane ablation efficacy, which leads to limited release of oocyte activating factors and lowers fertilization rate. Moreover, sperm swim in small angles relative to the micropipette tip cannot be tapped due to the risk of damaging the sperm’s genetic materials contained in the sperm head. In this paper, we propose automation techniques for effective membrane ablation and orientation control of sperm. A clinically compatible piezo drill is developed to generate controllable micropipette motion along both axial and lateral directions. The whirl flow generated by micropipette vibration is employed to rotate sperm, which greatly increased the number of available sperm for immobilization. A visual servoing controller is developed to keep the sperm at the center of field of view for immobilization by decoupling sperm wiggling from positioning error. The developed methods can be generalized to the manipulation of other types of cells. The piezo drill can be used for effective membrane ablation of oocyte, embryo, yeast cell and so on. The orientation control strategy leveraging piezo-induced whirl flow is applicable to non-contact rotation of a variety of microorganism.

This paper reports, for the first time, automated grading of human blastocysts (day-5 embryos) from multi-focus images. Based on a novel attention module, a convolutional neural network (CNN) was developed to predict the morphological grade of a blastocyst. The attention module integrates high-level features extracted from the blastocyst’s multi-focus images. Experimental results revealed that multi-focus blastocyst images help improve the grading accuracies than a single blastocyst image. Comparisons of the accuracy achieved by the model and the average accuracy of five embryologists demonstrated that the proposed model can outperform embryologists in the morphological grading of blastocysts (88% versus 86% for development stage prediction, 83% versus 79% for inner cell mass grade prediction, 89% versus 82% for trophectoderm grade prediction).Note to Practitioners—This work was motivated by the subjectivity and significant intra-and inter-evaluator variations in manual morphological grading of blastocysts. Existing approaches to automate the grading process mainly use a single blastocyst image although multi-focus images captured at different focal planes reveal more morphological features of a blastocyst than a single blastocyst image. This paper describes a new CNN-based method using multi-focus images to improve the grading accuracy. The accuracy of the proposed method was verified on multi-focus images of human blastocysts captured by a standard time-lapse incubator at fixed focal depths ($-$45$\mu$m,$-$30$\mu$m,$-$15$\mu$m, 0$\mu$m, 15$\mu$m, 30$\mu$m, 45$\mu$m).

The capability to obtain detailed motility information of model organisms is fundamental to reveal their functional and social behavior characteristics. Zebrafish is a powerful vertebrate model organism. Despite recent success in the automatic quantification of adult zebrafish movement, it remains a laborious task for group zebrafish larval tracking due to their similar appearance, frequent occlusions, and highly discontinuous kinematics. This article presents DanioSense (DS), an automatic tracker for group larval zebrafish, to overcome these tracking challenges. The integration of a light convolutional neural network and a centerline extraction algorithm enables the tracker to localize individuals even in occlusion cases where objects’ identities are prone to switch. With reliable detections, an adaptive Kalman filter is designed to optimally estimate locomotive parameters, which is also used for object reidentification accomplished by a two-stage data association protocol. Experimental results demonstrated a tracking accuracy of over 97%, median errors of$102~{\mathrm{\mu m}}$, and 8.8° for the position and orientation measurement, and a processing speed of over 30 frames/s with a normal computer configuration. DS provides detailed quantitative data for a large-scale larvae group in nearly real time, highly boosting the efficiency of characterizing individual phenotypes and analyzing social interactions.Note to Practitioners—This article aimed to tackle the problem of automated tracking groups of zebrafish larvae, an ideal vertebrate model organism for large-scale chemical and genetic screens. The task of group tracking is to record each individual’s movement and calculate their position, velocity, direction, and other parameters for further analysis, where the correct identity of each individual must be maintained. Existing algorithms either switch larvae’ identities easily or are unable to achieve online tracking due to the limitations of their methods to address individuals’ intersections. DanioSense (DS) adopts a convolutional neural network to identify larval heads whenever they intersect and uses an adaptive Kalman filter to calculate the movement parameters optimally. Besides, a range of visualization options is designed to bring insight into underlying patterns through massive amounts of data. Theoretically, this algorithm’s approach to solving intersections and calculating movement statistics can also apply to other fish-like animals. Its visualization options are applicable to other tracking systems. The key advantage of Daniosense over existing trackers is the capability to track each larva within a group and output detailed quantitative data in nearly real time. The tracking performance of DS is based on the quality of image segmentation and the success rate of classifying samples. Many state-of-the-art image segmentation and classification neural networks can be adopted to extend this system’s applications to more complex environments but at a higher computation and time cost, which is a tradeoff between efficiency and capability. Some applications require a higher video sampling rate, so the system’s processing speed needs to be further improved with better hardware and software framework optimization. The next steps include improving the processing efficiency, providing more tracking modules and visualization options, and extending its application fields.

Predators in nature grip their prey in different ways, which give innovational ideas of gripping approaches in industrial applications. Octopus performs flexible gripping with the help of vacuum grippers, suction cups, which inspired a new type of microgripper for biological sample micromanipulation. The proposed gripper consists of a glass pipette and a pump driven by a step-motor. The step-motor is controlled with adaptive robust control to adjust the gripping pressure applied on the biological sample. A dynamic model is developed for the biological sample aiming for better deformation control performance. A visual detection algorithm is developed for data processing to identify the parameters in the dynamic model and the detection result of visual algorithm is also used as feedback of adaptive robust control, which diminishes the negative influence of parameter and model uncertainties. Zebrafish larva was used as the testing sample for experiment and the corresponding parameters were identified experimentally. The experimental results correlated well with the model predicted deformation curve and visual detection algorithm provided promising accuracy, which is less than [Formula: see text]. Adaptive robust control provides fast and accuracy response in point-to-point deformation testing, and the average responding time is less than 30 s and the average error is no larger than 1 pixel.

Precise rotational positioning of end-effectors under microscopy is crucial for robotic micromanipulation. However, the end-effector is presently limited to a fixed orientation, which is manually set before a given micromanipulation task, lacking accuracy and versatility ofin situreorientation. In this article, we present a unified framework for rotationally positioning the end-effector in three dimensions by establishing a general rotational model, developing a detection method within the limited field of view under microscopy, and designing a three-loop control strategy that adapts to different experimental requirements and model parameters. In experiments, a standard angled micropipette was used as the end-effector to verify the validity of the proposed methods. The performance was evaluated experimentally where the micropipette was robotically rotated to an arbitrarily desired orientation with an average orientation error less than$2^\circ$. In the experiments of sperm manipulation, thein situmicropipette orientation control capability improved the success rate of sperm immobilization and achieved dexterous robotic sperm orientation for facile aspiration.

Cell manipulation is a key technology in many biomedical and clinical applications, in which end-effector alignment is a critical procedure. Presently, end-effector alignment is performed manually and suffers from large misalignment error and inconsistency. Manual alignment often undesirably moves the end-effector (e.g., a glass micropipette) out of the limited field of view under microscopy and risks breaking the fragile end-effector. This paper presents automated end-effector alignment for robotic cell manipulation. A rotational degree of freedom was added to a micromanipulator with translational degrees of freedom. The kinematic model of end-effector’s rotation was established, and the unknown model parameters were calibrated and updated via quadratic optimization. A controller was designed based on the kinematics modeling and parameter optimization to compensate for rotation-induced translation and achieve end-effector alignment. Experimental results demonstrate that the robotic alignment technique achieved an accuracy of 0.6±0.3° and a time cost of 18.5 ± 10.2 s, both significantly less than manual alignment. The developed controller cost significantly less time for micropipette alignment than the PID controller. A glass micropipette was used as the end-effector for human sperm immobilization, a critical procedure in clinical cell surgery. The success rate of sperm immobilization was 97% by robotic micropipette alignment, higher than the success rate of 90% by manual alignment due to the higher accuracy of robotic alignment.

This article presents an automated strategy to touch the injection site on zebrafish larva skin with the injection pipette tip accurately in the presence of water-depth variation, which is a crucial problem to automate zebrafish larva microinjection. The presented method consists of two parts: adaptive coordinate transformation and curve evolution for edge detection. In the first part, the impact of refraction is taken into consideration. An adaptive calibration method is developed, which enables the coordinate transformation matrix to adapt to the changing water depth. In the second part, the abovementioned calibration result is used to keep the injection pipette tip descending along the desired route. A curve-evolution-based edge detection algorithm is introduced to detect the deformation of larva skin caused by contact with the injection pipette tip. Experimental results demonstrate that high accuracy and success rates are achieved. The effect of uncertainties caused by water-depth variation and the skill requirement in manual manipulation are eliminated. The proposed contact detection strategy can be extended to microinjection for other organisms.Note to Practitioners—As a typical multicellular model organism, the zebrafish has been increasingly used in biological research. For studying drug toxicity and disease models, exogenous substances need to be injected into zebrafish larvae. However, for both manual and automated injection, a fatal problem is that the camera on the microscope only provides 2-D positional information. It is laborious to align the pipette tip with the injection site along the$z$-axis. Moreover, due to the characteristic of stereomicroscopes, the impact of refraction at the water surface cannot be ignored. In order to address these issues, in this article, we present an adaptive calibration method and an edge detection algorithm for zebrafish larva heart injection to avoid contact failure in practical implementations.

We give a brief review of the developments in terahertz time-domain spectroscopy (THz-TDS) systems and micro-cavity components for probing samples in the University of Shanghai for Science and Technology. The broadband terahertz (THz) radiation sources based on GaAs m-i-n diodes have been investigated by applying high electric fields. Then, the free space THz-TDS and fiber-coupled THz-TDS systems produced in our lab and their applications in drug/cancer detection are introduced in detail. To further improve the signal-to-noise ratio (SNR) and enhance sensitivity, we introduce three general micro-cavity approaches to achieve tiny-volume sample detection, summarizing our previous results about their characteristics, performance, and potential applications.

Three-dimensional (3D) morphological information of Zebrafish larvae is important for investigating the development of the vertebrate model. Some existing automated handling systems have already been developed to reconstruct 3D models of micro-objects, but many commercial devices are generally costly and complicated to assemble, which limits their wide usage. In this paper, we present an automated visual servo platform to carve 3D model of Zebrafish larva in a simple and controllable manner. The proposed 3D carving strategy only involves a 4-DOF manipulator, a glass capillary and a micropump. The Zebrafish larva is first captured by the capillary mounted at the end of the manipulator. Then, the manipulator rotates larva body to desired orientations in order to obtain 2D images from different views. A structure-from-motion algorithm finally carves the 3D model of the larva body. Experimental results verify the validity of proposed methods, and a guideline of selecting the number of views is also given. As a high-cost-performance system, it has a considerable reference for reconstructing other microobjects.

Zebrafish has been demonstrated to be an important model organism in the study of genetics, diseases and drugs. For investigating drug toxicity, we need to inject foreign substances into specific organs within zebrafish larvae. Traditionally, zebrafish larva microinjection is conducted manually and requires operators to control larva's orientation with flexible ends, which is time-consuming, labor-intensive, and inaccurate. In this paper, we present an integrated microfluidic system to facilitate zebrafish larva organs microinjection, which is capable of adjusting larvae to appropriate orientation conveniently. The head's direction of zebrafish is adjusted through a microfluidic chip and a series of pumps so that the larva moves tail-first at the exit of the microfluidic channel, where it is rotated around its body axis to a desired orientation. Finally, the binary image of the larva is analysed to locate the organs and injection is executed. Experimental results are presented to verify the efficiency of the proposed method.