Manual Method Research Articles

A Condition Assessment Tool for Steel Bridge Deck Pavement Systems Based on Data Balancing Methods and Machine Learning Algorithms

The primary challenge in the operation of steel deck pavement systems lies in the inspection and assessment of their condition. Traditionally, manual inspection methods are employed. However, these approaches are not only time-consuming and labor-intensive but also prone to human error. As a result, integrating data-driven machine learning technologies into the evaluation of pavement systems presents a significant advantage in addressing these issues. This study proposes a decision-making tool for estimating the condition levels of steel bridge deck pavement systems by employing classification techniques. To address the issue of class imbalance in the dataset, the SMOTE algorithm is utilized. Additionally, seven different machine learning methods—Light Gradient Boosting Machine, Extreme Gradient Boosting, Random Forest, Adaptive Boosting, K-Nearest Neighbor, Multilayer Perceptron, and Logistic Regression—are applied for training. Comparative analysis reveals that the Light Gradient Boosting performs optimally, achieving classification accuracies of 0.841 and 0.929 on the original and synthetic datasets, respectively.

Comparative Study on Cost Analysis of Mechanical and Manual Transplanting of Rice

In Telangana, rice cultivation predominantly relies on traditional manual transplanting, which requires a significant labor force. However, the migration of agricultural laborers to urban areas in search of better wages has led to a shortage of workers, particularly during the critical transplanting season. To address this labor shortage and reduce cultivation costs, an efficient and cost-effective method of rice transplanting is essential, without compromising grain yield. To explore such a solution, Krishi Vigyan Kendra in Jammikunta, Karimnagar district, conducted 30 frontline demonstrations between 2018-19 and 2020-21, showcasing rice cultivation using a self-propelled, walk-behind six-row mechanical transplanter. The performance of this mechanical method was evaluated by collecting data on various growth parameters, including plant height, number of productive tillers per hill, panicle length, grains per panicle, and grain yield, and comparing it with traditional manual transplanting. Economic analysis revealed that the gross returns from mechanical transplanting were Rs. 135,956, compared to Rs. 128,629 for manual transplanting. Similarly, the net returns for mechanical and manual methods were Rs. 95,106 and Rs. 82,204, respectively. The benefit-cost ratio was found to be 3.32 for mechanical transplanting and 2.77 for manual transplanting. These findings indicate that mechanical rice transplanting offers a viable, cost-effective alternative, reducing labor requirements and drudgery while delivering higher yields.

Validation of the Sysmex XN-V hematology analyzer for feline specimens.

The Sysmex XN-V is derived from the new Sysmex XN series of human hematology analyzers. The main changes from the previously validated XT-2000iV analyzer include an optic-fluorescent analysis for platelets and a nucleated red blood cell (NRBC) count. We aimed to validate the Sysmex XN-V for feline blood following the American College for Veterinary Clinical Pathology and International Council for Standardization in Hematology recommendations. Feline EDTA blood specimens were analyzed on the Sysmex XN-V to evaluate repeatability, linearity, comparison with the XT-2000iV analyzer and manual methods, stability, and to verify the previously established Sysmex XT-2000iV RIs. Repeatability was excellent for most variables. Visually determined linearity was excellent or good for most variables except eosinophils and platelet variables. The correlation between the XN-V and XT-2000iV analyzers was good (≥0.82) for all variables except reticulocyte indices. Correlations between the Sysmex XN-V and manual differential counts were good to excellent for most variables, acceptable for neutrophils, and fair for monocytes and NRBC. The previously established Sysmex XT-2000iV RIs can be used to interpret results from the Sysmex XN-V analyzer for most variables except red cell distribution width and reticulocyte variables. The RI for platelet variables could not be evaluated because of platelet clumps. Changes in the Sysmex XN-V measurements after storage at 4 and 24°C were similar to those described for the Sysmex XT-2000iV analyzer. The performance of the Sysmex XN-V analyzer was good and compared favorably with the Sysmex XT-2000iV analyzer.

Research on Applying Deep Learning to Visual–Motor Integration Assessment Systems in Pediatric Rehabilitation Medicine

In pediatric rehabilitation medicine, manual assessment methods for visual–motor integration result in inconsistent scoring standards. To address these issues, incorporating artificial intelligence (AI) technology is a feasible approach that can reduce time and improve accuracy. Existing research on visual–motor integration scoring has proposed a framework based on convolutional neural networks (CNNs) for the Beery–Buktenica developmental test of visual–motor integration. However, as the number of training questions increases, the accuracy of this framework significantly decreases. This paper proposes a new architecture to reduce the number of features, channels, and overall model complexity. The architectureoptimizes input features by concatenating question numbers with answer features and selecting appropriate channel ratios and optimizes the output vector by designing the task as a multi-class classification. This paper also proposes a model named improved DenseNet. After experimentation, DenseNet201 was identified as the most suitable pre-trained model for this task and was used as the backbone architecture for improved DenseNet. Additionally, new fully connected layers were added for feature extraction and classification, allowing for specialized feature learning. The architecture can provide reasons for unscored results based on prediction results and decoding rules, offering directions for children’s training. The final experimental results show that the proposed new architecture improves the accuracy of scoring 6 question graphics by 12.8% and 12 question graphics by 20.14% compared to the most relevant literature. The accuracy of the proposed new architecture surpasses the model frameworks of the most relevant literature, demonstrating the effectiveness of this approach in improving scoring accuracy and stability.

Cognitive State Classification Using Convolutional Neural Networks on Gamma-Band EEG Signals

This study introduces a novel methodology for classifying cognitive states using convolutional neural networks (CNNs) on electroencephalography (EEG) data of 41 students, aimed at streamlining the traditionally labor-intensive analysis procedures utilized in EEGLAB. Concentrating on the 30–40 Hz frequency range within the gamma band, we developed a CNN model to analyze EEG signals recorded from the inferior parietal lobule during various cognitive tasks. The model demonstrated substantial efficacy, achieving an accuracy of 91.42%, precision of 71.41%, and recall of 72.51%, effectively distinguishing between high and low gamma activity states. This performance surpasses traditional machine learning methods for EEG analysis, such as support vector machines and random forests, which typically achieve accuracies between 70–85% for similar tasks. Our approach offers significant time savings over manual EEGLAB methods. The integration of event-related spectral perturbation (ERSP) analysis with a novel CNN architecture enables capture of both fine-grained and broad spectral EEG features, advancing the field of computational neuroscience. This research has implications for brain-computer interfaces, clinical diagnostics, and cognitive monitoring, offering a more efficient and accurate alternative to current EEG analysis methods.

Comparison of manual titrations to automated microplate and gel titration assays used in the screening of blood donors for production of Low-Titer O Whole Blood.

The popularity of Low-Titer O Whole Blood (LTOWB) for treating trauma patients requires that donor centers and transfusion services make decisions on what titer testing capabilities to institute and an appropriate titer level threshold. This study compared the titer results determined by four methods to find a rate of agreement. Isohemagglutinin titers were tested on 300 plasma samples utilizing various methods, each determining IgM antibody levels by direct hemagglutination with A1 and B reference cells. The methods used were the Beckman Coulter's PK7300, Immucor's NEO Iris microplate technology, Ortho Clinical Diagnostics (OCD) Vision, and manual titrations. Only 42.7% of the samples tested showed agreement across all methods on ABO isohemaglutinin titer levels and only 32.5% demonstrated "High titer" agreement. Sample agreement was close to 90% if the Immucor method was excluded. At a <1:256 titer level threshold, the pass rate was 94.3% for Immucor, 89.7% for the PK7300, 87.3% for manual testing, and 75.7% for OCD's gel method. Sensitivity and specificity rates at a ± 1 titer level were respectively 100% and 95.4% for OCD's gel, 73.7% and 100% for Immucor, and 100% and 99.6% for the PK7300. Overall method accuracy was 91.7% for Immucor, 90.3% for the PK7300, and 86.7% for OCD's gel method as compared to manual titration. All three automated methods perform comparably to the manual method at a ± 1 titer tolerance level. Based on these comparisons, a titer level of <1:256 would maximize LTOWB production regardless of the method used.

Visual Target Interaction Modeling with Multichannel Data Fusion

Facing the demand for accurate, fast, and natural human–computer interaction in the multi-screen control environment, the traditional exchange method of frequent switching control of multi-screen through mouse, keyboard, and other manual methods has problems of low efficiency, weak flexibility, and poor experience. In this paper, we research multi-screen precision control technology based on eye movement to break through the difficulties of precise recognition of human vision under complex environments and poor stability of vision estimation under frequent switching control of multi-screen. This study explores the technology for controlling multiple screens with high precision by tracking eye movements. It overcomes the challenges associated with reliably discerning human visual focus in intricate settings and the instability of visual assessments during rapid transitions between multiple screens. The experimental results of 1143 eye-movement capture tasks are studied, including the accurate recognition technology of human head posture and pupil gaze direction under complex backgrounds, illumination, and different visual field environments, and the theoretical modeling method of multi-screen precise target interaction based on eye-movement is explored. The expression data will be processed using convolutional neural networks, and the dataset will be trained through Python programming and tested on test samples collected by simulated flight crews in the laboratory with an accuracy rate of more than 85%.

Image Analysis Artificial Intelligence Technologies for Plant Phenotyping: Current State of the Art

Modern agriculture is characterized by the use of smart technology and precision agriculture to monitor crops in real time. The technologies enhance total yields by identifying requirements based on environmental conditions. Plant phenotyping is used in solving problems of basic science and allows scientists to characterize crops and select the best genotypes for breeding, hence eliminating manual and laborious methods. Additionally, plant phenotyping is useful in solving problems such as identifying subtle differences or complex quantitative trait locus (QTL) mapping which are impossible to solve using conventional methods. This review article examines the latest developments in image analysis for plant phenotyping using AI, 2D, and 3D image reconstruction techniques by limiting literature from 2020. The article collects data from 84 current studies and showcases novel applications of plant phenotyping in image analysis using various technologies. AI algorithms are showcased in predicting issues expected during the growth cycles of lettuce plants, predicting yields of soybeans in different climates and growth conditions, and identifying high-yielding genotypes to improve yields. The use of high throughput analysis techniques also facilitates monitoring crop canopies for different genotypes, root phenotyping, and late-time harvesting of crops and weeds. The high throughput image analysis methods are also combined with AI to guide phenotyping applications, leading to higher accuracy than cases that consider either method. Finally, 3D reconstruction and a combination with AI are showcased to undertake different operations in applications involving automated robotic harvesting. Future research directions are showcased where the uptake of smartphone-based AI phenotyping and the use of time series and ML methods are recommended.

From Fax to Secure File Transfer Protocol: The 25-Year Evolution of Real-Time Syndromic Surveillance in England.

The purpose of syndromic surveillance is to provide early warning of public health incidents, real-time situational awareness during incidents and emergencies, and reassurance of the lack of impact on the population, particularly during mass gatherings. The United Kingdom Health Security Agency (UKHSA) currently coordinates a real-time syndromic surveillance service that encompasses 6 national syndromic surveillance systems reporting on daily health care usage across England. Each working day, UKHSA analyzes syndromic data from over 200,000 daily patient encounters with the National Health Service, monitoring over 140 unique syndromic indicators, risk assessing over 50 daily statistical exceedances, and taking and recommending public health action on these daily. This English syndromic surveillance service had its origins as a small exploratory pilot in a single region of England in 1999 involving a new pilot telehealth service, initially reporting only on "cold or flu" calls. This pilot showed the value of syndromic surveillance in England, providing advanced warning of the start of seasonal influenza activity over existing laboratory-based surveillance systems. Since this initial pilot, a program of real-time syndromic surveillance has evolved from the single-system, -region, -indicator pilot (using manual data transfer methods) to an all-hazard, multisystem, automated national service. The suite of systems now monitors a wide range of syndromes, from acute respiratory illness to diarrhea to cardiac conditions, and is widely used in routine public health surveillance and for monitoring seasonal respiratory disease and incidents such as the COVID-19 pandemic. Here, we describe the 25-year evolution of the English syndromic surveillance system, focusing on the expansion and improvements in data sources and data management, the technological and digital enablers, and novel methods of data analytics and visualization.

Ventilation and Oxygenation During and After Adult Cardiopulmonary Resuscitation: Changing Paradigms.

Cardiac arrest (CA) remains a major cause of death despite advancements in cardiopulmonary resuscitation (CPR), post-resuscitation care, and international efforts to develop evidence-based guidelines. Effectively managing ventilation and oxygenation during and after CPR is vital for patient survival and neurological outcomes, yet it remains a challenging task. This review examines current strategies for ventilation and oxygenation during and after CPR, focusing on evidence-based guidelines, the balance between ventilation effectiveness and risks, and proposed methods for monitoring ventilation quality. It emphasizes the need to provide adequate ventilation and oxygenation during and after CPR while avoiding hyperventilation and hypoventilation, which can negatively impact resuscitation and post-CA outcomes. The review also explores mechanical ventilation as an alternative to manual methods and the use of feedback devices. The impact of post-CA ventilation and oxygenation on patient outcomes and recommended management strategies are discussed. Finally, the review highlights current gaps in the literature and the need for more well-designed large clinical studies, such as the impact of different ventilation variables (tidal volume and respiratory rate) on the return of spontaneous circulation (ROSC) and long-term outcomes.

Adaptive denoising graph contrastive learning with memory graph attention for recommendation

Graph contrastive learning has emerged as a powerful technique for dealing with graph noise and mining latent information in networks, that has been widely applied in GNN-based collaborative filtering. Traditional graph contrastive learning methods commonly generate multiple augmented views, and then learn node representations by maximizing the consistency between these views. However, on one hand, manual view construction methods necessitate expert knowledge and a trial-and-error process. On the other hand, adaptive view construction methods require decoders which results in increased training costs. To address the aforementioned limitations, in this paper, we propose the Adaptive Denoising Graph Contrastive Learning with Memory Graph Attention for Recommendation (ADGA) framework. Firstly, we introduce the memory graph attention mechanism to capture node attention during multi-hop information aggregation. Then, unlike previous methods that required additional node representations to generate views, ADGA proposes, for the first time, directly using attention to adaptively generate structure-aware contrastive learning views. It reduces the training cost of the model and improves the cross-view consistency of node representations, that offers a new paradigm for adaptive graph contrastive learning. Experimental results on three real-world datasets demonstrate that ADGA achieves state-of-the-art performance in recommendation tasks. The code is available at https://github.com/Andrewsama/ADGA.

Automated recognition and rebar dimensional assessment of prefabricated bridge components from low-cost 3D laser scanner

Dimension quality inspections for rebar spacing and length of prefabricated bridge components are critical for subsequent efficient assembly on-site. Currently, the traditional manual inspection method is time-consuming, labor-intensive, and error-prone. The existing commercial 3D LiDAR-based methods are difficult in making balances between inspection cost, efficiency, and accuracy. Thus, the automated recognition and segmentation method of 3D point clouds acquired by a self-developed low-cost LiDAR is proposed to evaluate the rebar spacing and length of bridge prefabricated components. Due to the high-cost commercial 3D laser scanner, a spinning 2D LiDAR-based low-cost 3D laser scanner device and geometric calibration model were developed to acquire the 3D spatial point cloud. Then, two-stage automated point cloud segmentation framework is proposed with coarse extraction of the point cloud in the interest region and fine recognition of point cloud using machine learning methods, including plane segmentation and circle fitting employing random sample consensus (RANSAC) algorithm, and rebars segmentation utilizing Gaussian mixture model (GMM) and density-based spatial clustering of applications with noise (DBSCAN) algorithm. The proposed system was evaluated in two prefabricated concrete columns and shows the potential for improving the quality inspection efficiency and accuracy.

Enhancing clinical data retrieval with Smart Watchers: a NiFi-based ETL pipeline for Elasticsearch queries

BackgroundThe aim is to develop and deploy an automated clinical alert system to enhance patient care and streamline healthcare operations. Structured and unstructured data from multiple sources are used to generate near real-time alerts for specific clinical scenarios, with an additional goal to improve clinical decision-making through accuracy and reliability.MethodsThe automated clinical alert system, named Smart Watchers, was developed using Apache NiFi and Python scripts to create flexible data processing pipelines and customisable clinical alerts. A comparative analysis between Smart Watchers and the legacy Elastic Watchers was conducted to evaluate performance metrics such as accuracy, reliability, and scalability. The evaluation involved measuring the time taken for manual data extraction through the electronic patient record (EPR) front-end and comparing it with the automated data extraction process using Smart Watchers.ResultsDeployment of Smart Watchers showcased a consistent time savings between 90% to 98.67% compared to manual data extraction through the EPR front-end. The results demonstrate the efficiency of Smart Watchers in automating data extraction and alert generation, significantly reducing the time required for these tasks when compared to manual methods in a scalable manner.ConclusionsThe research underscores the utility of employing an automated clinical alert system, and its portability facilitated its use across multiple clinical settings. The successful implementation and positive impact of the system lay a foundation for future technological innovations in this rapidly evolving field.

Intercomparison of Different Agrochemicals (Herbicides) and Manual Methods of Weed Control on Weed Population and Weed Control Efficiency in Wheat (Triticum aestivum L.) Crop in Himachal Pradesh Dist

Aims: A research study entitled Intercomparison of different Agrochemicals (Herbicides) and manual methods of weed control on weed population and weed control efficiency in wheat crop (Triticum aestivum L.) in Himachal Pradesh district Mandi. Study Design: The experiment was designed using a Randomized Block design method. Place and Duration of Study: Department of agronomy, School of Agriculture, Abhilashi university, Chail Chowk, Mandi, (H.P.) during the Rabi season of 2022-23. Methodology: The field trail was conducted with eight treatments and replicated thrice. The investigation included eight different weed control techniques i.e. T1 -weedy check, T2 - weed free, T3 - hand weeding at 30 DAS, T4 - 2, 4-D @ 0.5 kg a.i. ha-1 + Sulfosulfuron @ 0.02 kg a.i. ha-1 at 30 DAS,T5 - Pendimethalin @ 1.5 kg a.i. ha-1(P.E) + Quizalofop-ethyl @ 50 gm a.i ha-1 (PoE at 25 DAS) + one hand weeding, T6 - 2, 4-D @ 0.5 kg a.i. ha-1, T7 - 2, 4-D @ 0.5 kg a.i. ha-1 + one hand weeding at 30 DAS, T8 - 2, 4-D @ 0.5 kg a.i. ha-1 + two hand weeding at 30 and 45 DAS. Results: The investigation found that the most effective ways to control weeds are by keeping the area completely weed free by using pendimethalin @ 1.5 kg a.i. ha-1(P.E) + Quizalofop-ethyl @ 50 gm a.i ha-1(PoE at 25 DAS) + one Hand weeding (Treatment T2) was found the most successful in terms of achieving the lowest weed density and highest weed control efficiency. Application of pendimethalin @ 1.5 kg a.i. ha-1(P.E) + Quizalofop-ethyl @ 50 gm a.i ha-1 (PoE at 25 DAS) + one hand weeding was found to be better than using herbicides to reduce weed infestation in wheat crops. Conclusion: Based on the results, it can be concluded that integrated weed management practices weed-free treatment was the most effective option for managing weeds in wheat.

Modeling of Applying Road Pricing to Airport Highway Using VISUM Software in Jordan

Road congestion in Amman City has been increasing yearly, due to the increase in private car ownership and traffic volumes. This study aims to (a) evaluate the toll road’s effects on society and the economy in Amman, Jordan, through a survey questionnaire using statistical software (SPSS), (b) assess the impact of the toll road on reducing congestion and delays using micro-simulation (VISUM), (c) identify the optimal toll price for a selected road using VISUM and (d) validate the simulated models with the optimal revenue. Traffic, geometric, and cost data about the toll technique of two sections on the Airport Highway (from the Ministry of Foreign Affairs to the Madaba Interchange; and from the Madaba Interchange to the Queen Alia International Airport (QAIA) Interchange) were used for simulation purposes. The toll road (across seven different scenarios at different prices) was evaluated for optimal revenue. The survey questionnaire was made based on all scenarios, including the AM peak hour. The operation cost for the toll road was determined based on the Greater Amman Municipality (GAM). The best scenario was determined based on the value of revenue (JOD). The results indicate that higher acceptance is achieved when applying road pricing during the AM peak hour and that users prefer the charging method based on travelled distance (54.02%). Additionally, the total cost of the manual toll collection (MTC) method is 126,935 JOD. Road pricing can reduce traffic delay (or speed up traffic flow) by 4.61 min in the southbound direction and by 9.52 min in the northbound direction. The optimal toll value is 0.25 JOD (34.08%), with revenues of 1089.6 JOD for 2024 and 1122.6 JOD for 2025. Eventually, applying road pricing on the airport road is shown to be effective and economically feasible only when using the manual method.

Da_Tracker: Automated workflow for high throughput single cell and single phagosome tracking in infected cells.

Time-lapse microscopy has emerged as a crucial tool in cell biology, facilitating a deeper understanding of dynamic cellular processes. While existing tracking tools have proven effective in detecting and monitoring objects over time, the quantification of signals within these tracked objects often faces implementation constraints. In the context of infectious diseases, the quantification of signals at localized compartments within the cell and around intracellular pathogens can provide even deeper insight into the interactions between the pathogen and host cell organelles. Existing quantitative analysis at a single-phagosome level remains limited and dependent on manual tracking methods. We developed a near-fully automated workflow that performs with limited bias, high-throughput cell segmentation and quantitative tracking of both single cell and single bacterium/phagosome within multi-channel, z-stack, time-lapse confocal microscopy videos. We took advantage of the PyImageJ library to bring Fiji functionality into a Python environment and combined deep-learning-based segmentation from Cellpose with tracking algorithms from Trackmate. The 'da_tracker' workflow provides a versatile toolkit of functions for measuring relevant signal parameters at the single-cell level (such as velocity or bacterial burden) and at the single-phagosome level (i.e. assessment of phagosome maturation over time). Its capabilities in both single-cell and single-phagosome quantification, its flexibility and open-source nature should assist studies that aim to decipher for example the pathogenicity of bacteria and the mechanism of virulence factors that could pave the way for the development of innovative therapeutic approaches.

GeGra: Approaching a generic model for quantitative grain size analysis from materials microscopy data using deep learning

Grain size has a significant impact on the properties of materials, and is crucial for predicting material properties. Traditional grain size measurement relies heavily on human operators, leading to subjective results, and existing machine learning methods are typically material-specific, requiring significant labeling and training efforts for each new material. This paper provides insight into developing a deep learning-based generic grain boundary detection model (GeGra) from different material micrographs. The model is trained on 1006 images from various microscopy techniques such as light optical, Kerr, and scanning electron microscopy, acquired at different magnifications for different materials such as copper, austenite, brass, sintered hard magnet, hard metal, bronze, nickel silver, and aluminum. The developed GeGra model effectively handles visual artifacts and substructures such as twin grains, which often pose challenges for material-specific, state-of-the-art grain boundary segmentation models. The developed model achieved an IoU score of 69 % on a diverse test set and enables accurate grain size analysis using external image analysis software in less than one minute, according to ASTM standards, which is more than 5 times faster than the manual method. The developed model prioritizes generality with objective that it can have broader applicability for various materials instead of high-precision grain boundary detection. Additionally, the model has the potential to be a foundational tool for generalized grain size analysis in material microscopy, reducing the effort required for such analysis and assisting both material science experts and machine learning users.

Lightweight deep learning model for underwater waste segmentation based on sonar images

In recent years, the rapid accumulation of marine waste not only endangers the ecological environment but also causes seawater pollution. Traditional manual salvage methods often have low efficiency and pose safety risks to human operators, making automatic underwater waste recycling a mainstream approach. In this paper, we propose a lightweight multi-scale cross-level network for underwater waste segmentation based on sonar images that provides pixel-level location information and waste categories for autonomous underwater robots. In particular, we introduce hybrid perception and multi-scale attention modules to capture multi-scale contextual features and enhance high-level critical information, respectively. At the same time, we use sampling attention modules and cross-level interaction modules to achieve feature down-sampling and fuse detailed features and semantic features, respectively. Relevant experimental results indicate that our method outperforms other semantic segmentation models and achieves 74.66 % mIoU with only 0.68 M parameters. In particular, compared with the representative PIDNet Small model based on the convolutional neural network architecture, our method can improve the mIoU metric by 1.15 percentage points and can reduce model parameters by approximately 91 %. Compared with the representative SeaFormer T model based on the transformer architecture, our approach can improve the mIoU metric by 2.07 percentage points and can reduce model parameters by approximately 59 %. Our approach maintains a satisfactory balance between model parameters and segmentation performance. Our solution provides new insights into intelligent underwater waste recycling, which helps in promoting sustainable marine development.

Implementation and Evaluation of APOTECAchemo in a Community Cancer Center: A Comparative Study of Robotic Versus Manual Antineoplastic Preparation.

Background: The ever-increasing complexity and demand for antineoplastic therapy necessitates innovative solutions to improve the accuracy and safety of drug preparation. Objective: To evaluate the utilization of an advanced robotic chemotherapy drug compounding system (APOTECAchemo) at a Community Cancer Center (CCC), examining accuracy, efficiency, and staff perceptions. Methods: This single-center, retrospective study evaluated the preparation of 7 intravenous (IV) antineoplastics at a CCC over a 1-year period. We compared manual methods with the APOTECAchemo system. The primary measure of accuracy was the absolute drug error percentage, with a comparison of pass and fail rates. Secondary endpoints included the overall use of APOTECAchemo for all IV antineoplastic preparations and average preparation times. An end-user satisfaction survey gathered feedback from pharmacists and pharmacy technicians. Results: A total of 8210 doses were prepared at the CCC, with 52.1% compounded by APOTECAchemo and 47.9% manually. Of these, the CCC prepared 5526 doses of the 7 routinely compounded antineoplastics. APOTECAchemo prepared 3851 (69.7%) doses, while manual compounding accounted for 1675 (30.3%) doses. The average absolute drug error was 1.44% (95% CI, 1.35-1.53) with robot compounding versus 1.17% (95% CI, 1.03-1.32) with manual (P < 0.001). The overall failure rate was 0.72%. There were 25 failed doses (0.45%), with 8 (0.2%) failures attributed to APOTECAchemo and 17 (1%) to manual compounding (P < 0.001). The average dose preparation time was longer with APOTECAchemo compared with manual methods. The end-user satisfaction survey indicated a positive reception toward APOTECAchemo. Conclusions: Our study demonstrates the successful implementation, extensive utilization, and high accuracy of both APOTECAchemo and manual compounding methods in the preparation of routinely administered antineoplastics at a CCC.

Lite-YOLOv8: a more lightweight algorithm for Tubercle Bacilli detection.

Deep learning is a transformative force in the medical field and it has made significant progress as a pivotal alternative to conventional manual testing methods. Detection of Tubercle Bacilli in sputum samples is faced with the problems of complex backgrounds, tiny and numerous objects, and human observation over a long time not only causes eye fatigue, but also greatly increases the error rate of subjective judgement. To solve these problems, we optimize YOLOv8s model and propose a new detection algorithm, Lite-YOLOv8. Firstly, the Lite-C2f module is used to ensure accuracy by significantly reducing the number of parameters. Secondly, a lightweight down-sampling module is introduced to reduce the common feature information loss. Finally, the NWD loss is utilized to mitigate the impact of small object positional bias on the IoU. On the public Tubercle Bacilli datasets, the mean average precision of 86.3% was achieved, with an improvement of 2.2%, 1.5%, and 2.8% over the baseline model (YOLOv8s) in terms of mAP0.5, precision, and recall, respectively. In addition, the parameters reduced from 11.2 to 5.1 M, and the number of GFLOPs from 28.8 to 13.8. Our model is not only more lightweight, but also more accurate, thus it can be easily deployed on computing-poor medical devices to provide greater convenience to doctors.