MitoGraph is a widely used tool for the automated segmentation of mitochondrial networks in three-dimensional (3D) fluorescence microscopy. However, the emergence of advanced live-cell microscopes such as lattice light-sheet microscopy (LLSM) has produced massive four-dimensional (4D, 3D+time) datasets that highlight a critical bottleneck: current CPU-based implementations are computationally prohibitive, often requiring days or weeks to process. To address this limitation, we developed MitoGraph-GPU, a Python-based GPU implementation that accelerates the dominant filtering steps by vectorizing Hessian/eigenvalue and vesselness computations using CuPy, and streamlines network processing with faster skeletonization and topology analysis. Tested across budding yeast and human lung organoid datasets, MitoGraph-GPU achieves up to 11× speedup in yeast cells and 30× speedup in per-frame segmentation of lung cells. Segmentation fidelity is preserved, with ~99.9% agreement in maximum intensity projections of segmented images, and minimal differences in downstream measurements. Critically, this throughput enables practical analysis of large 4D datasets : in an LLSM organoid use case (10 movies, 60 frames, ~ 50 cells per movie), total processing time decreases from ~ 500h on CPU to ~ 20h on GPU (25× faster). By producing accurate mitochondrial surfaces and skeletons, MitoGraph-GPU can serve as an efficient segmentation module for downstream mitochondrial tracking and analyses, enabling scalable high-throughput 4D mitochondrial phenotyping.

Routine culling of day-old male chicks in the global poultry industry triggers severe resource waste and critical animal welfare crises, while non-destructive early sex identification techniques for embryonated quail eggs remain a prominent research gap. This study developed a novel Raman spectroscopy-based method for quail embryo sexing on incubation day 5, using an innovative "shell perforation without inner membrane damage" sampling strategy. The optimized GA-CARS-ELM model achieved 80.95% accuracy on the independent test set with 0.39 ms single-sample model inference time and 5.3 ± 0.5 min total per-egg processing time under manual operation, outperforming mainstream machine learning and deep learning algorithms. This work fills the relevant research gap and provides vital technical support for the development of automated pre-hatching sex sorting systems in poultry farming.

This study proposes a hybrid deep learning framework for multi-class brain tumor classification using a combined MRI dataset constructed from Figshare, Br35H, and SARTAJ sources. The dataset includes 7023 MRI images belonging to four clinically important classes: glioma, meningioma, no-tumor, and pituitary tumor. In the proposed approach, a dedicated preprocessing pipeline is first applied to enhance tumor-related image regions and reduce irrelevant background information. Then, the Multi-scale GHOST Residual Attention Autoencoder (MS-GHOST-RAAE) is used for deep feature extraction. This architecture integrates multi-scale GHOST modules, residual connections, and attention mechanisms to obtain compact, stable, and discriminative feature representations. The extracted features are subsequently classified using conventional machine learning classifiers. Experimental results show that the proposed hybrid framework achieves 98.75% accuracy on the combined dataset, with a total processing time of 543 + 269.4571 s. These findings indicate that MS-GHOST-RAAE provides strong classification performance on heterogeneous MRI images and offers an effective computer-aided decision-support approach for brain tumor classification.

To determine diagnostic agreement between frozen sectioning (FS) and formalin-fixed, paraffin-embedded (FFPE) tissue-processing methods for veterinary oral tumors and to chronicle the method and time required to perform FS sample processing in a private practice setting. 27 canine and feline oral tumor samples were collected between October 2024 and February 2025 and processed in 2 ways: (1) in-house using FS and (2) conventional FFPE. All biopsies were interpreted by 2 veterinary pathologists, one who read the FFPE samples for normal diagnostic turnaround time, who then blinded and digitized the FS slides for submission to the second pathologist. The final diagnoses for FS and FFPE tissue samples were compared for each tumor. Mitotic count and FS slide quality were also recorded for each sample. Time elapsed between surgical biopsy and completion of FS was recorded. Excellent diagnostic agreement (0.915 [95% CI, 0.775 to 1.0]) was documented in veterinary oral tumors processed with FS and FFPE methods. Poor to moderate agreement was documented between FS and FFPE mitotic counts (0.42 [95% CI, 0.12 to 0.65]). The mean total FS slide processing time was 19.80 minutes (SD, 8.58). High diagnostic concordance was documented in veterinary oral tumors processed with FS and FFPE methods, with results similar to human studies. Slides were processed in a reasonable amount of time. Future study evaluating veterinary oral tumor FS samples using telepathology is warranted. FS should be considered a feasible and accurate tool for intraoperative veterinary oral tumor diagnosis.

Abstract This study proposes a machine vision-based automatic tool setting system for ultrasonic thread root rolling (UTRR), which addresses challenges in tool setting accuracy and the inefficiency of traditional manual or contact-based methods. The system integrates ROI extraction based on YOLOv5, an improved Canny edge detection algorithm, and the Iterative Least-Squares Piecewise Fitting Algorithm (ILS-PFA) to build a robust tool setting framework suitable for complex metallic threads under challenging conditions such as strong reflections and shadows. YOLOv5 is employed for reliable detection of the thread and rolling tool, even under varying lighting conditions. The improved Canny algorithm, augmented with adaptive bilateral filtering, Sobel gradients, and Otsu thresholding, effectively suppresses noise from metallic reflections and enhances edge clarity. The ILS-PFA method reconstructs thread profiles by fitting arc and line segments, while iterative optimization restores data lost due to shadows and reflections. A localization method based on curvature transition is used to precisely locate the thread root feature point, and the system calculates the spatial distance between the tool feature point and this feature point using camera calibration parameters. Experimental results demonstrate that the proposed method achieves micron-level measurement accuracy, with a mean absolute error of approximately 3.2 μm and a maximum relative error below 0.4% across different thread specifications. The tool-setting repeatability exhibits sub-micron stability, with a standard deviation below 0.6 μm. The total execution time of the automatic tool-setting process is approximately 1.78 s, satisfying practical industrial requirements. Error analysis indicates that the overall system error is controlled within 4 μm. These results confirm that the proposed method provides a robust, accurate, and cost-effective solution for automatic tool setting in UTRR applications.

Large-scale tissue proteomics requires workflows that are efficient, rapid, and repeatable across diverse samples. Herein, we present a Simple Workflow for Integrated and Fast Tissue-preparation (SWIFT), which enables complete processing of fresh-frozen (FF) and formalin-fixed, paraffin-embedded (FFPE) tissues in either one- or two-step formats while maintaining deep proteome coverage with high repeatability from low- to microgram-level tissues. For FF tissues, an incubation process integrating lysis, reduction, alkylation, and digestion generates peptide samples directly from tissues in ≤1.5 h. For FFPE tissues, concurrent deparaffinization, rehydration, and de-cross-linking are achieved within 0.5 h, followed by one-step peptide preparation. Furthermore, our workflows eliminate desalting and offline cleanup steps, thereby reducing variability and total processing time. Using our methods, we identified up to ∼10,000 protein groups and ∼150,000 peptides across multiple mouse organs on the timsTOF Pro. Repeatability was high (pairwise Pearson's r > 0.96 across six experimental replicates), with dynamic ranges spanning 6-7 orders of magnitude. Organ-enriched protein analysis identified functionally distinct proteins unique to each tissue. Paired FF and FFPE analyses revealed preservation-induced shifts, with FFPE tissues showing reduced detection of membrane-associated and respiratory proteins, including mitochondrial Complex I. Together, our fast and simplified workflows enable deep tissue proteomics for large-scale clinical and translational studies in a cost-effective and widely accessible manner.

Effective management and amalgamation of the multi-source financial information so as to provide individualized investment and wealth management services require optimization of ETL (extract, transform, load) pipeline in customer-centric investment solution will be provided as the general methodology in this paper. The project contains data conversion of the former on-premises Oracle systems into the snowflake based data warehouse with the type of data they held was the contributions to the portfolios, product relations, customer relations and information about the participants. There is systematization of the process and includes data modelling, pipeline design and performance optimization. To make sure that the data models are aligned to the business objectives as well as the data governance sets, logical, conceptual and physical data models were developed. Informatica IICS and Powercenter were adopted as effective in extracting, processing and loading of different sources like SQL, Oracle and Azure Databrinks. AWS Lambda control-M and cloud-native services, S3, KMS, SQS and SNS were introduced to automate the processes to reduce the number of processes completed manually and provide scale execution. According to the quantitative statistics, the effectiveness of processing increased significantly: the total ETL processing time was decreased by 42 percent and the data rate increased by 38 percent, which makes the reporting and analysis almost real-time. By applying DAX as KPIs of power BI dashboards, the action could be taken on the customer portfolios based on the specific investment campaigning and tailored financial recommendation. The paper sheds light on the effectiveness, quality of data and customer value of business in customer-centric financial platforms with optimized ETL pipelines.

Considering optical Braille patterns has been investigated in several studies. There is a huge number of studies which analyzed Braille patterns in different natural languages. However, the Arabic patterns have not been examined as same as the other languages. This is due to the lack of the datasets of the Arabic patterns and the shortage of researches in this area. This study utilizes YOLOv11 model as a detection tool because of its relative effectiveness and the level of accuracy as well as the staged training approach with the AdamW optimizer and Automatic Mixed Precision. For the translation of the Arabic patterns into text, post-processing steps are performed including: detecting cells vertically clustered, horizontally sorted within a line, adaptively defined word boundaries, and corrected reading order of right-to-left. In the analysis of experiments, the best findings achieved is 0.99 of all the precision, recall, and F1 scores. Moreover, the framework-level runtime indicates that the total processing time (inference + post-processing for text extraction) ranges between 29 and 82 ms per image. The proposed framework has been examined with a primary dataset of 5924 pages of images of Braille patterns of 45 classes of Arabic letters and diacritics. The yielded results show that the proposed framework is a robust approach toward effective, scalable, responsive Arabic Braille recognition (OBR) for assistive technologies to be mobile and wearable. By building the first dedicated corpus in Arabic Braille and providing an end-to-end recognition suite, this study laid the groundwork for future research and applications in the field. This research bridges the accessibility gap, so of allow sighted individuals to access content encoded in Braille.

This study introduces and validates dynamic multiplex PCR, a newly developed multiplex PCR method designed to overcome the inherent limitations of conventional multiplexing, particularly the challenges associated with heterogeneous primer melting temperatures (Tm) among multiple primer pairs in a single reaction. While traditional protocols rely on a single, fixed annealing temperature (Ta) per cycle, the dynamic multiplex PCR approach applies four sequential annealing temperatures (52°C, 54°C, 57°C, and 60°C) within each individual PCR cycle. This novel intra-cycle multi-temperature strategy was evaluated using an 8-plex microsatellite panel in European sea bass (Dicentrarchus labrax) samples and compared against three established protocols: multiplex touchdown PCR and two conventional PCR setups at fixed temperatures of 54°C and 57°C. The performance of the methods was assessed through a dual-platform quantitative analysis. Amplification yields were first quantified via densitometric analysis of agarose gel electrophoresis, while capillary fragment analysis was utilized to evaluate peak height, total peak area, quality scores, and fragment sizing precision. Sizing accuracy was rigorously determined as the absolute deviation between observed fragment sizes and assigned allele values. Statistical evaluations were conducted using a General Linear Model (GLM) and ANOVA framework to examine the effects of the PCR method, operator, locus, and biological sample on amplification performance and genotyping reliability. The results demonstrate that dynamic multiplex PCR significantly outperforms both conventional and multiplex touchdown PCR protocols by providing more balanced and robust amplification across all eight loci. By accommodating diverse primer requirements within a single cycle, this method eliminates the need for extensive trial-and-error optimization, thereby substantially reducing the consumption of laboratory consumables and total processing time. These findings establish dynamic multiplex PCR as a highly reliable, cost-effective, and reproducible multiplex PCR approach, offering a significant methodological advancement for high-quality multi-locus genotyping and fragment analysis workflows.

Abstract Background/Aims Biologic therapies have revolutionised the treatment of inflammatory and autoimmune rheumatic diseases. However, their administration - often through hospital-based infusion suites - creates logistical challenges for service providers and patients. As a result, subcutaneous biologics are increasingly being delivered to patients by private ‘Homecare’ companies. As highlighted in a recent Government inquiry, Homecare services are often subject to delays in prescribing, pharmacy processing and drug delivery. Patients receiving treatment via these pathways typically have significant disease burden, and treatment delays increase the risk of relapse or progression. This quality improvement project aimed to improve Homecare prescribing in the Rheumatology department at Sandwell and West Birmingham NHS Trust. Methods Three audit loops were conducted: baseline, early and late post-intervention; each included all patients with a Homecare prescription in a 2-week window. For each patient, we calculated the time taken to process their prescription by a) the Rheumatology department and b) the hospital pharmacy. From this, the total processing time and delay in prescription compared to the estimated due date were calculated. Data were non-normally distributed and analysed using the Mann-Whitney U test. The project was registered locally and patient data were held securely. Results Pre-intervention, prescriptions took on average 56 days to process, largely due to departmental delays (Table 1). As a result, a Biologics Coordinator was appointed to organise monitoring blood tests, triage requests from Homecare companies and chase prescriptions. While no benefit in patient delay was seen in loop two (p = 0.20), there was an improvement in delay of 11 days from the first to third loop (p < 0.00001). This came despite slower pharmacy processing of prescriptions. Conclusion Significant delays in Homecare biologic prescribing create the potential for under-treatment of patients. Through a targeted intervention, the Rheumatology department has effectively reduced prescribing delays in the space of a year. While this has meaningfully improved service quality, the potential benefits have been negated by slower pharmacy processing. This highlights the need for a system-wide approach, and future work alongside pharmacy colleagues is needed to understand the challenges they face. Disclosure I.G. Raza: None. D. Stanley: None. A. Tabassum: None. J. Reynolds: None.

Most industries in developing countries are highly dependent on tasks that involve human effort. Allocation of manual tasks to workers is thus a very important factor for reducing lead time in labor-dependent industries. The goal of this research lies in improving worker-task allocation in labor-dependent industries using a two-stage Predict-then-Optimize framework. In this study, machine learning was used to predict task completion time for each worker-task pair, and the results gained from the prediction model were then optimized through mixed-integer programming. Task-related data were collected multiple times at different hours of the day from 54 worker samples across 17 distinct tasks, making the dataset size 400. Random Forest, SVM, XGBoost, and ANN were used for prediction, and the optimization problem was formulated using Mixed Integer Programming and solved with the Gurobi optimizer. Random Forest Regressor (RFR) demonstrated the best performance with the lowest MAPE (6.63%) and the highest R2 score (0.98) for the prediction of task completion times based on demographic features and experience. Using these predictions, task allocation was optimized, cutting total process time by 67.5%-from over 50 minutes to just 16.3 minutes. So, this integrated Predict-then-Optimize framework approach can improve the industrial environment in labor-intensive industries by assisting top-level management in decision-making about reducing the total task completion time.

Next-generation sequencing (NGS) library preparation is a core component of precision genomics, but it is commonly constrained by inefficiency, variability, and low throughput of manual protocols. To address these limitations, we developed and systematically evaluated a fully automated NGS workstations and further validated its performance across representative application scenarios. The automated system reduced total processing time from 8 to 10 to 4–6 h. At the same time, it maintained similar performance in pre-library metric, including DNA yield and fragment size, as well as post-capture sequencing metrics (Q30 > 90%, mapping rates > 95%, on-target rates 85–90%). The duplication rate was reduced to 5–8%, compared with 10–15% for manual methods, indicating increased library complexity. Bioinformatic evaluation of inter-species read mapping showed minimal cross-contamination, with a maximum contamination ratio of 0.0003%, indicating effective sample isolation in the automated workflow. High concordance in variant detection was observed between automated and manual workflows. Overall, this automated workstation provides a standardized and reproducible workflow that supports scalable precision genomics applications.

Printed Circuit Board manufacturing is a key sector of modern electronics industry where improving the throughput of microvia drilling operations is of paramount importance. One effective solution is maskless laser technology which provides high accuracy and processing flexibility. However, its bottleneck remains the beam positioning speed, limited by the inertia of galvanometer-based scanners. This work proposes a hybrid control method for microvia laser drilling that combines a galvanometer-based scanner and an acousto-optic deflector. The set of vias is pre-partitioned into clusters so that all vias within each cluster can be processed by the acousto-optic deflector inside its deflection field without involving the galvanometer. Cluster centers are then connected by a minimal trajectory computed using a combination of a greedy algorithm and the pairwise exchange method (2-opt) which minimizes the total travel length of the galvanometer and the overall drilling cycle time. This approach enables coordinated use of the high-speed acousto-optic deflector for local processing and the long-range galvanometer for movements between clusters. Implementation of the proposed method reduced the galvanometer travel length from 3,097.05 mm to 1,674.19 mm and decreased the total processing time by more than a factor of 3.3 compared with traditional approaches. The effect is achieved by minimizing the number of large inertial moves and shifting a portion of the motion tasks to the high-speed acousto-optic deflector. Unlike known approaches that optimize a single traveling salesman problem route over all vias, the proposed method realizes a hierarchical routing scheme. Classical methods minimize route length but do not account for the dynamic limitations of the galvanometer which leads to excessive inertial moves. Pure acousto-optic deflector based systems provide very high speed but are limited by a small deflection field. The hybrid approach combines advantages of both technologies: the acousto-optic deflector delivers high-speed processing within clusters, while the galvanometer performs efficient transitions between them. The method requires no substantial hardware modifications, can be integrated into existing control systems, and is adaptable to microprocessing of glass substrates (for through glass vias) for 2.5D and 3D packaging architectures.

High hydrostatic pressure (HHP) can extend the shelf life and ensure safety of meat products such as lamb burgers. Lupinus albus variety Orden Dorado (a low alkaloid content variety) flour, rich in protein and phenolic compounds, offers the potential to enhance the preservation of meat products during storage. Lamb burgers were formulated with Lupinus albus flours (1%, w/w; weight/weight), either conventional or obtained by compression milling, and processed by HHP treatments (untreated, two consecutive cycles at 600 MPa for 1 s; or a single cycle at 600 MPa for 4 min), with the total processing time using the HHP unit being the same for both. Then, they were subsequently stored for 14 days under refrigerated conditions. Proximate composition, microbiological changes, color, and oxidation of burgers during storage were evaluated. Flour obtained by compression milling presented higher phenolic compound content, while its antioxidant activity is similar to that obtained by conventional methods. In lamb burgers, the incorporation of both lupine flours maintained the proximate composition and fatty acids profile. Lipid oxidation after 14 days was significantly lower in burgers with lupine flour obtained by compression milling, whereas protein oxidation responses depended on treatment-formulation interactions. HHP drastically reduced microbial counts, with sustained inactivation of coliforms and Escherichia coli (E. coli) although instrumental color was significantly altered in fresh burgers after processing. However, sensory scores of grilled burgers remained unaffected by either flour type or HHP treatment. Incorporation of Lupinus albus flour into lamb burgers processed by HHP preserved sensory quality while enhancing the microbial and lipid oxidation stability of burgers. Finally, the application of two short (1 s) cycles at 600 MPa was more beneficial than a single 4 min cycle, offering similar microbial inactivation with less impact on the quality changes in burgers. Finally, applying two short (1 s) HHP cycles at 600 MPa was more beneficial than a single 4 min cycle, as it achieved similar microbial inactivation while better preserving the color and oxidative stability of the burgers.

ABSTRACT Conventional manufacturing processes for thermoplastic composites typically rely on external heating sources, resulting in high energy consumption and extended processing cycles. To overcome these limitations, this study applied self‐resistance electric (SRE) heating technology to fabricate carbon fiber‐reinforced polymethyl methacrylate (CF/PMMA) composites via compression molding. Initially, the electrical conductivity of the prepreg was measured to establish a temperature‐current correlation model based on experimental data, validating the feasibility of the SRE heating process. The physical properties, interfacial characteristics, porosity, and mechanical performance of the resulting composites were systematically characterized. The results demonstrate that SRE heating significantly shortens the consolidation cycle, reducing the total processing time to approximately 22% of that required for conventional hot pressing. Although SRE processing leads to a higher porosity (3.72%) and a corresponding reduction in tensile strength (−9.2%), the flexural strength is enhanced by 11.8%, attributed to improved interfacial load transfer in surface plies. These findings indicate that internal Joule heating provides an efficient manufacturing route for PMMA‐based CFRTPs where reduced cycle time and flexural performance are critical.

The demand for effective downstream processing of adeno-associated virus (AAV) is increasing as gene therapies advance toward broader clinical applications. Robust, efficient, and scalable ultrafiltration and diafiltration (UF|DF) operations are essential for generating high-quality AAV preparations, with tangential flow filtration (TFF) serving as a critical unit operation for vector concentration, impurity reduction, and buffer exchange while maintaining viral functionality. Development of TFF processes requires careful consideration of membrane characteristics-including chemistry, pore size or channel architecture-as these parameters directly influence vector retention, fouling behavior, and overall process efficiency. Equally important is the optimization of critical process parameters such as recirculation rate, transmembrane pressure (TMP), and total processing time, all of which govern hydrodynamic performance and product quality. This study assessed two Sartocon® Hydrosart® TFF cassette architectures-ECO-Screen and E-Screen-for the ultrafiltration and diafiltration of AAV8 clarified lysate. Through flux characterization and controlled small-scale evaluations, cassette-specific operating regions were defined. Both configurations supported high viral genome retention; however, the E-Screen geometry achieved faster processing and superior removal of host-cell protein and DNA contaminants, whereas the ECO-Screen format allowed for efficient operation under reduced pump rates and, therefore, lower shear conditions. Reproducibility assessments demonstrated minimal run-to-run variability, confirming the robustness of the optimized operating parameters. A 10-fold scale-up further validated the linearity and predictability of the UF|DF process, with consistent impurity-reduction profiles and only modest deviations in viral recovery. Collectively, these findings provide a quantitative basis for rational cassette selection in AAV purification workflows and establish a scalable, scientifically grounded UF|DF framework applicable across development and manufacturing scales.

Background and Objectives: Epilepsy is a chronic neurological disorder characterized by recurrent seizures caused by abnormal brain activity. Reliable near-real-time seizure detection is essential for preventing injuries, enabling early interventions, and improving the quality of life for patients with drug-resistant epilepsy. This study presents a near-real-time epileptic seizure detection framework designed for low-latency operation, focusing on improving both clinical reliability and patient comfort through electrode reduction. Method: The framework integrates bidirectional long short-term memory (BiLSTM) networks with wavelet-based feature extraction using Electroencephalogram (EEG) recordings from the EPILEPSIAE dataset. EEG signals from 161 patients comprising 1032 seizures were analyzed. Wavelet features were combined with raw EEG data to enhance temporal and spectral representation. Furthermore, electrode reduction experiments were conducted to determine the minimum number of strategically positioned electrodes required to maintain performance. Results: The optimized BiLSTM model achieved 86.9% accuracy, 86.1% recall, and an average detection delay of 1.05 s, with a total processing time of 0.065 s per 0.5 s EEG window. Results demonstrated that reliable detection is achievable with as few as six electrodes, maintaining comparable performance to the full configuration. Conclusions: These findings demonstrate that the proposed BiLSTM-wavelet approach provides a clinically viable, computationally efficient, and wearable-friendly solution for near-real-time epileptic seizure detection using reduced EEG channels.

While the impact of task complexity on translation performance has received considerable attention, relatively little research has explored whether affective factors such as translation anxiety moderate this effect. This study investigated the joint influence of task complexity and translation anxiety on English as a foreign language (EFL) learners' written translation performance using a 2 × 2 mixed design. A total of 106 EFL learners from diverse disciplines in China participated in the study, with 100 included in the final analyses; their translation anxiety was assessed using a Translation Anxiety Scale adapted for written translation. Participants completed two written translation tasks at different complexity levels. Their translation performance was assessed in terms of process efficiency (i.e., total processing time, subjective cognitive effort, and the number of effective revisions) and product quality (i.e., accuracy, fluency, and analytical quality measures). Linear mixed-effects models showed consistent effects of task complexity on translation performance, whereas translation anxiety played a selective and context-dependent role, mainly by moderating the effects of task complexity on some efficiency outcomes. The study develops a conceptual framework that integrates task-related factors and learner factors as predictors of translation performance and offers pedagogical insights for translation teaching.

This paper presents a vision-based real-time 3D position sensor for spherical actuators using Iterative Closest Point (ICP) pose estimation. Unlike conventional rotary motors with fixed axes, spherical actuators enable omnidirectional rotation, which makes traditional one-dimensional sensing methods, such as encoders, Hall sensors, and resolvers, unsuitable for accurate orientation measurement. To address this challenge, we propose a non-contact sensing approach that reconstructs three-dimensional surface point clouds from monocular camera images and estimates the full 3D orientation of the spherical rotor. Arbitrary and non-uniform visual patterns are applied to the rotor surface, and both intentional patterns and natural surface imperfections (e.g., scratches, dust, and wear) are exploited as geometric features for pose estimation. Edge features are extracted using Canny edge detection, reconstructed into 3D point clouds through a geometrically consistent mapping model, and aligned via ICP to estimate the spherical angles θ and ϕ. Experimental validation is conducted using a spherical rotor with a diameter of 15 cm. Under single-axis rotations, a minimum mean absolute error of 0.45° in θ is achieved after pixel-per-revolution calibration. Under dynamic multi-axis rotational motions, the estimated tangential azimuth direction ϕ exhibits compact zero-centered error distributions within approximately 0.2°– 0.9°, confirming consistent 3D directional estimation. Long-term stability analysis further shows that cumulative angular error accumulation remains approximately 0.15° per degree under appropriate sensing conditions. A computational architecture comparison demonstrates the feasibility of real-time operation. The complete processing pipeline, including image preprocessing, 3D reconstruction, and ICP-based pose estimation, achieves a total processing time below 32 ms per frame, satisfying real-time requirements at 30 fps. These results confirm that the proposed vision-based sensor provides a robust, non-contact, and practical solution for real-time 3D orientation sensing in spherical actuator systems.

This research addresses the growing challenges of patient scheduling with limited medical resources, such as consultation and examination services. Based on the concept of a pipeline, multiple intelligent heuristic algorithms are considered to optimize outpatient scheduling while reducing total time spent, including in consultation, the laboratory, or examination and idle time caused by scheduling conflicts. Multiple cross-department medical services with various patient conditions are considered for scheduling. To handle the complexity of limited examination resources, intelligent heuristic algorithms, like the genetic algorithm, simulated annealing, and ant colony optimization, are deployed. The modularized artificial intelligence service prioritizes minimizing not only patients’ average waiting time but also the total process time for all patients completing the medical service requests, while ensuring effective allocation of shared medical resources. According to the verification results, the genetic algorithm can adapt quickly to diverse hospital or patient requirements. Both average waiting time and total process time can be reduced and saved.