Reduce Processing Time Research Articles

Adaptive random-grid and progressive color secret sharing with CNN super-resolution on a many-core system

This study introduces an innovative method combining Convolutional Neural Networks (CNN) with random grid generation for enhanced visual secret sharing. Our approach supports color images, reduces processing time, and offers non-pixel expansion and flexible share combinations. Utilizing GPU computation, it significantly improves practicality and efficiency by operating in a feed-forward manner, avoiding complex optimization. Experimental results demonstrate superior metrics: maximum PSNR of 32 dB, 99% NCC correlation with benchmarks, 2.9% NAE, and SSIM of 98%. We achieve substantial speedups– 493 × for 256 × 256 and 3145 × for 1024 × 1024 images–compared to sequential models. The scheme exhibits robustness against attacks, evidenced by CMY component and share histogram similarities, and scalability shown in combinatorial explosion visualization. These findings underscore the efficacy and efficiency of our approach, advancing secure image sharing applications significantly.

Integrated multi-objective optimization of rough and finish cutting parameters in plane milling for sustainable machining considering efficiency, energy, and quality

Plane milling is one of the most prevalent methods in metalworking, recognized for its potential to uncover energy-saving opportunities by optimizing cutting parameters. However, existing research on cutting parameters optimization primarily adopts a phased independent optimization approach, with some studies focusing on rough milling parameters and others on finish milling parameters under predetermined allowances. These research methods lack integrated optimization for both rough and finish milling parameters, leading to suboptimal outcomes. To address this issue, this paper proposes an integrated multi-objective optimization method for both rough and finish cutting parameters in plane milling. The aim is to achieve sustainable machining by considering efficiency, energy consumption, and quality. Specifically, the research accomplishes three key objectives: (1) developing detailed models for the efficiency, energy, and quality of plane milling; (2) presenting an integrated multi-objective optimization model and approach for determining rough and finish cutting parameters in plane milling; (3) conducting a specific case study to validate the effectiveness and feasibility of the proposed model and approach. Experimental findings demonstrated that optimizing cutting parameters in the plane milling process through an integrated approach can reduce processing time by 19.37%, decrease energy consumption by 16.88%, and significantly improve the surface roughness of the workpiece by 27.07% compared to traditional empirical methods. Furthermore, compared to independent optimization schemes, processing time is reduced by 2.99%, energy consumption is lowered by 4.72%, and surface roughness is improved by 13.16%.

Experimental performance analysis of methanol adsorption in granular activated carbon packed bed through design of a double pipe heat exchanger with longitudinal fins

Activated carbon (AC) is essential for its exceptional adsorption properties and high surface area in various applications. Adsorption/desorption (A/D) processes with AC can be exothermic or endothermic, necessitating improved heat exchanger (HE) designs due to AC's low thermal conductivity. This study presents an A/D HE design with longitudinal fins in the sorption bed to enhance heat transfer during methanol A/D in AC. Bed heating/cooling uses water flowing through pipes inside the HE. Experimental investigations focus on thermal characteristics and A/D efficiency under diverse conditions, comparing results to a base non-finned HE design. Various experiments measured HE shell temperatures at different water flow rates while varying the sorption bed condition (vacuumed, filled with air, or methanol). The finned HE design showed a 1.8 °C increase in maximum bed surface temperature and a 3-h, 40-min reduction in process time compared to the non-finned HE. Subsequent tests with the finned HE at different water flow rates (5, 10, and 15 l/min) demonstrated superior heat transfer at 15 l/min due to turbulent flow enhancement. Additionally, adsorption ceased at a bed temperature of 54 °C. These findings highlight the efficacy of finned HE for optimizing methanol A/D processes, offering insights into improving heat transfer in AC-based systems.

Optimization Algorithms for Real-Time Image Processing in IoT Networks

This study explores the application of optimization algorithms, with a focus on Particle Swarm Optimization (PSO), to enhance real-time image processing in Internet of Things (IoT) networks. Given the constraints of computational power and energy resources in IoT devices, efficient processing of image data is a critical challenge. The proposed PSO-based framework aims to minimize processing time and energy consumption while maintaining high accuracy in image analysis tasks. Experimental results demonstrate that PSO can reduce processing time by 10% and energy consumption by 15% on average, with a slight improvement in accuracy. Comparisons with traditional and other evolutionary optimization techniques reveal that PSO provides a superior balance between efficiency and performance, making it particularly well-suited for real-time IoT applications. The findings suggest that the integration of advanced optimization algorithms like PSO into IoT systems can significantly enhance their operational efficiency, scalability, and effectiveness, particularly in resource-constrained environments.

AI-Driven Thoracic X-ray Diagnostics: Transformative Transfer Learning for Clinical Validation in Pulmonary Radiography.

Our research evaluates advanced artificial (AI) methodologies to enhance diagnostic accuracy in pulmonary radiography. Utilizing DenseNet121 and ResNet50, we analyzed 108,948 chest X-ray images from 32,717 patients and DenseNet121 achieved an area under the curve (AUC) of 94% in identifying the conditions of pneumothorax and oedema. The model's performance surpassed that of expert radiologists, though further improvements are necessary for diagnosing complex conditions such as emphysema, effusion, and hernia. Clinical validation integrating Latent Dirichlet Allocation (LDA) and Named Entity Recognition (NER) demonstrated the potential of natural language processing (NLP) in clinical workflows. The NER system achieved a precision of 92% and a recall of 88%. Sentiment analysis using DistilBERT provided a nuanced understanding of clinical notes, which is essential for refining diagnostic decisions. XGBoost and SHapley Additive exPlanations (SHAP) enhanced feature extraction and model interpretability. Local Interpretable Model-agnostic Explanations (LIME) and occlusion sensitivity analysis further enriched transparency, enabling healthcare providers to trust AI predictions. These AI techniques reduced processing times by 60% and annotation errors by 75%, setting a new benchmark for efficiency in thoracic diagnostics. The research explored the transformative potential of AI in medical imaging, advancing traditional diagnostics and accelerating medical evaluations in clinical settings.

High Throughput Manufacturing of Highly Ordered Mesoporous Thin Film Oxides for Hybrid-Polymer Nanocomposite Dielectrics

Highly ordered mesoporous metal oxide thin films have potential to become integrated into widely impactful dielectric materials due to the tunable properties of the oxide and its polymer filling; however, current manufacturing techniques to form the mesoporous oxide greatly limit scalable fabrication. Sol-gel mesoporous metal oxides generated via traditional processes require multi-day aging and high temperature, prolonged curing processes to yield controlled porosity at the expense of manufacturability. Moreover, high cost and unstable alkoxide precursors require precision over the solution precursor pH and subsequent thin film aging humidity after deposition.Solution combustion synthesis (SCS) is a next-generation method of generating metal oxides at reduced process temperatures (often <300 ºC) and using green precursors compared to the sol-gel approach. The metal oxide is typically produced from a metal nitrate, which provides both the metal cation and oxidant and an organic chelator fuel. The fuel acts to stabilize the cation against hydrolysis and to participate in the exothermic redox reaction that results in the metal oxide.Here, we present porogen integrated rapid oxidation (PIRO) which utilizes the principles of SCS to fabricate large-area mesoporous thin films of AlOx in less than 1 hour and at a temperature of 230 ºC. The PIRO method uses low-cost precursors to generate micelles with a controllable open-cell porosity ranging from ~30 – 60% and with varying degrees of closed-packed cubic ordering at a nearly 95% reduction in overall processing time. We show the precursor chemistry can be deposited via ultrasonic spray or blade coating at up to 6 m/min to manufacture large area mesoporous oxides and demonstrate filling of the porous AlOx with TOPAS® polymer to generate a hybrid nanocomposite with a dielectric constant of 6.26.

FUTURE TRENDS IN SQL DATABASES AND BIG DATA ANALYTICS: IMPACT OF MACHINE LEARNING AND ARTIFICIAL INTELLIGENCE

This study systematically reviews the integration of machine learning (ML) and artificial intelligence (AI) into SQL databases and big data analytics, highlighting significant advancements and emerging trends. Using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, a comprehensive review of 60 selected articles published between 2010 and 2023 was conducted. The findings reveal substantial improvements in query optimization through ML algorithms, which adapt dynamically to changing data patterns, reducing processing times and enhancing performance. Additionally, embedding ML models within SQL databases facilitates real-time predictive analytics, streamlining workflows, and improving the accuracy and speed of predictions. AI-driven security systems provide proactive and real-time threat detection, significantly enhancing data protection. The development of hybrid systems that combine relational and non-relational databases offers versatile and efficient data management solutions, addressing the limitations of traditional systems. This study confirms the evolving role of AI and ML in transforming data management practices and aligns with and extends previous research findings.

Nanoscale single-vesicle analysis: High-throughput approaches through AI-enhanced super-resolution image analysis

The analysis of membrane vesicles at the nanoscale level is crucial for advancing the understanding of intercellular communication and its implications for health and disease. Despite their significance, the nanoscale analysis of vesicles at the single particle level faces challenges owing to their small size and the complexity of biological fluids. This new vesicle analysis tool leverages the single-molecule sensitivity of super-resolution microscopy (SRM) and the high-throughput analysis capability of deep-learning algorithms. By comparing classical clustering methods (k-means, DBSCAN, and SR-Tesseler) with deep-learning-based approaches (YOLO, DETR, Deformable DETR, and Faster R–CNN) for the analysis of super-resolution fluorescence images of exosomes, we identified the deep-learning algorithm, Deformable DETR, as the most effective. It showed superior accuracy and a reduced processing time for detecting individual vesicles from SRM images. Our findings demonstrate that image-based deep-learning-enhanced methods from SRM images significantly outperform traditional coordinate-based clustering techniques in identifying individual vesicles and resolving the challenges related to misidentification and computational demands. Moreover, the application of the combined Deformable DETR and ConvNeXt-S algorithms to differently labeled exosomes revealed its capability to differentiate between them, indicating its potential to dissect the heterogeneity of vesicle populations. This breakthrough in vesicle analysis suggests a paradigm shift towards the integration of AI into super-resolution imaging, which is promising for unlocking new frontiers in vesicle biology, disease diagnostics, and the development of vesicle-based therapeutics.

Automatic detection of infeasible paths in large-scale program based on program summaries

The existence of infeasible paths in a program reduces the coverage of test cases and causes a waste of valuable testing resources. Detecting infeasible paths allows for focusing testing resources on feasible paths. This paper introduces a method for detecting infeasible paths based on program summaries. Our proposed method partitions the program into sequential statements, conditional statements and loop statements, and automatically generates statement summaries and function summaries. It analyzes the summaries to extract the path constraints and determines the feasibility of paths. We implemented a detection tool named DTSIP based on this method, and conducted experiments using a set of benchmark programs and open source projects. The results confirm the effectiveness of our method in detecting infeasible paths. It can detect both intraprocedural and interprocedural infeasible paths, demonstrating its broad applicability. Our method overcomes challenges associated with analyzing complex paths, achieving efficient feasibility determination while reducing processing time.

Integrated processes (HPSE+scCO2) to prepare sterilized alginate-gelatine-based aerogel

Biopolymers application in biomedical areas has been limited due to the physicochemical degradation that occurs using conventional processing/sterilization methods (e.g., steam heat, γ-radiation, ethylene oxide). Aiming to avoid/minimize degradation and preserve their properties, supercritical carbon dioxide (scCO2) has been proposed as an alternative sterilization method for such materials. ScCO2 can simultaneously be used as a drying method to produce aerogels (i) and sterilize them (ii). However, a solvent exchange is required to prepare the alcogel from hydrogel, achievable through high-pressure solvent exchange (HPSE) (iii). This study integrated three processes: HPSE, scCO2 drying, and sterilization to prepare alginate-gelatine sterilized aerogels. Two scCO2 sterilization methods were tested. Results showed that sterilization did not compromise the aerogels’ chemical, thermal and swelling properties. Conversely, Young’s Modulus increased, and BET surface area decreased, due to the structural changes caused by the fast pressurization/depressurization rates applied during sterilization. Regarding the sterilization efficiency, results showed a reduction in contamination throughout the process, achieving a SAL of 10-4. The sterilized aerogels were non-cytotoxic in vitro and showed improved wound-healing properties. The innovative integrated process produced decontaminated/sterile and ready-to-use aerogels reducing process time by 75 %, from 2 days up to 12 h without compromising the aerogel’s properties.

Artificial Intelligence Function Management in Supporting the Process of Government Implementation and Public Services in Indonesia

This research investigates the impact of Artificial Intelligence (AI) on government administration and public service delivery in Indonesia. The study employed qualitative research methods, including semi-structured interviews and detailed case studies of selected AI projects. The research results indicate that AI has significantly improved administrative efficiency and public service delivery. The “Smart Administration System” has automated routine tasks, resulting in a 60% reduction in manual processing times and enhancing overall productivity. The “Policy Insight Tool” has supported better policy formulation through advanced predictive analytics and scenario modeling. Despite these successes, the study identified challenges such as resistance to change, technical difficulties, and data integration issues. Feedback from users also highlighted limitations in AI language processing and accessibility concerns.

Real-Time Personal Protective Equipment Non-Compliance Recognition on AI Edge Cameras

Despite advancements in technology, safety equipment, and training within the construction industry over recent decades, the prevalence of fatal and nonfatal injuries and accidents remains a significant concern among construction workers. Hard hats and safety vests are crucial safety gear known to mitigate severe head trauma and other injuries. However, adherence to safety protocols, including the use of such gear, is often inadequate, posing potential risks to workers. Moreover, current manual safety monitoring systems are laborious and time-consuming. To address these challenges and enhance workplace safety, there is a pressing need to automate safety monitoring processes economically, with reduced processing times. This research proposes a deep learning-based pipeline for real-time identification of non-compliance with wearing hard hats and safety vests, enabling safety officers to preempt hazards and mitigate risks at construction sites. We evaluate various neural networks for edge deployment and find that the Single Shot Multibox Detector (SSD) MobileNet V2 model excels in computational efficiency, making it particularly suitable for this application-oriented task. The experiments and comparative analyses demonstrate the pipeline’s effectiveness in accurately identifying instances of non-compliance across different scenarios, underscoring its potential for improving safety outcomes.

Adaptation of the protein misfolding cyclic amplification (PMCA) technique for the screening of anti-prion compounds.

Prion diseases result from the misfolding of the physiological prion protein (PrPC) to a pathogenic conformation (PrPSc). Compelling evidence indicates that prevention and/or reduction of PrPSc replication are promising therapeutic strategies against prion diseases. However, the existence of different PrPSc conformations (or strains) associated with disease represents a major problem when identifying anti-prion compounds. Efforts to identify strain-specific anti-prion molecules are limited by the lack of biologically relevant high-throughput screening platforms to interrogate compound libraries. Here, we describe adaptations to the protein misfolding cyclic amplification (PMCA) technology (able to faithfully replicate PrPSc strains) that increase its throughput to facilitate the screening of anti-prion molecules. The optimized PMCA platform includes a reduction in sample and reagents, as well as incubation/sonication cycles required to efficiently replicate and detect rodent-adapted and cervid PrPSc strains. The visualization of PMCA products was performed via dot blots, a method that contributed to reduced processing times. These technical changes allowed us to evaluate small molecules with previously reported anti-prion activity. This proof-of-principle screening was evaluated for six rodent-adapted prion strains. Our data show that these compounds targeted either none, all or some PrPSc strains at variable concentrations, demonstrating that this PMCA system is suitable to test compound libraries for putative anti-prion molecules targeting specific PrPSc strains. Further analyses of a small compound library against deer prions demonstrate the potential of this new PMCA format to identify strain-specific anti-prion molecules. The data presented here demonstrate the use of the PMCA technique in the selection of prion strain-specific anti-prion compounds.

Controlled growth of MPA-capped ZnS quantum dots through concentration-modulated single injection hydrothermal method

This study presents the controlled growth of 3-Mercaptopropionic acid (MPA)-capped ZnS quantum dots (QDs) using a concentration-modulated single injection hydrothermal method. Employing the Concentration optimization by optical spectra (COOS) method, we optimized the MPA:Zn:S ratios to investigate the influence of the capping agent, cation, and anion for exceptional properties suitable for optoelectronic and sensor applications. CMSIH operates as a single-step synthesis process, reducing processing time and complexity. This streamlined approach not only enhances efficiency but also minimizes the risk of contamination and ensures batch-to-batch consistency in QD production. Its moderate operating conditions, compared to other high-energy methods, also contribute to reduced energy consumption and environmental impact, aligning with sustainable manufacturing practices. Further, X-ray diffraction (XRD) confirmed the Zinc blend (cubic) phase of ZnS, and Fourier-transform infrared spectroscopy (FTIR) validated MPA capping. The QDs exhibited strong quantum confinement, causing a blue shift in absorption peaks compared to bulk ZnS. Higher MPA concentrations ranging from 0.02 M to 0.1 M induced a red shift in the absorption edge due to prolonged reaction times and strong cation binding by MPA. Variations in cation Zn and anion S ratios from 0.02 M to 0.1 M caused blue and red shifts in the absorption edge, respectively. For instance, Zn:S = 0.04:0.01 M increased cation concentrations, reducing QD size up to 0.67 nm, while enhanced anion concentrations Zn:S = 0.04:0.04 M enlarged the QDs size up to 2.35 nm. Remarkably, calculated QD sizes using Brus’ equation were smaller than the Bohr radius, even at an elevated temperature of 95 °C, indicating significant quantum confinement. Luminescence studies revealed reduced luminescence with higher MPA concentrations, increased luminescence intensity with higher cation Zn+ concentrations, and a red shift in the luminescence peak with higher anion S concentrations. As the temperature rises, there is an observable decrease in luminescence intensity. Furthermore, the investigation into the relationship between chemical composition and optical properties of MPA-capped ZnS QDs at elevated temperatures expands understanding of quantum confinement effects. The synthesised unique ultra small ZnS QDs can be used in advanced quantum sensors mainly as radiation detectors.

Deep learning-based automatic ASPECTS calculation can improve diagnosis efficiency in patients with acute ischemic stroke: a multicenter study.

The Alberta Stroke Program Early CT Score (ASPECTS), a systematic method for assessing ischemic changes in acute ischemic stroke using non-contrast computed tomography (NCCT), is often interpreted relying on expert experience and can vary between readers. This study aimed to develop a clinically applicable automatic ASPECTS system employing deep learning (DL). This study enrolled 1987 NCCT scans that were retrospectively collected from four centers between January 2017 and October 2021. A DL-based system for automated ASPECTS assessment was trained on a development cohort (N = 1767) and validated on an independent test cohort (N = 220). The consensus of experienced physicians was regarded as a reference standard. The validity and reliability of the proposed system were assessed against physicians' readings. A real-world prospective application study with 13,399 patients was used for system validation in clinical contexts. The DL-based system achieved an area under the receiver operating characteristic curve (AUC) of 84.97% and an intraclass correlation coefficient (ICC) of 0.84 for overall-level analysis on the test cohort. The system's diagnostic sensitivity was 94.61% for patients with dichotomized ASPECTS at a threshold of ≥ 6, with substantial agreement (ICC = 0.65) with expert ratings. Combining the system with physicians improved AUC from 67.43 to 89.76%, reducing diagnosis time from 130.6 ± 66.3 s to 33.3 ± 8.3 s (p < 0.001). During the application in clinical contexts, 94.0% (12,591) of scans successfully processed by the system were utilized by clinicians, and 96% of physicians acknowledged significant improvement in work efficiency. The proposed DL-based system could accurately and rapidly determine ASPECTS, which might facilitate clinical workflow for early intervention. The deep learning-based automated ASPECTS evaluation system can accurately and rapidly determine ASPECTS for early intervention in clinical workflows, reducing processing time for physicians by 74.8%, but still requires validation by physicians when in clinical applications. The deep learning-based system for ASPECTS quantification has been shown to be non-inferior to expert-rated ASPECTS. This system improved the consistency of ASPECTS evaluation and reduced processing time to 33.3 seconds per scan. 94.0% of scans successfully processed by the system were utilized by clinicians during the prospective clinical application.

Flash sintering of tunable dielectric Ba0.6Sr0.4TiO3 electroceramics

Ba(1-x)SrxTiO3 (BST) is a crucial dielectric material in tunable devices for modern wireless communication technologies, owing to high dielectric tunability and low dielectric loss. However, the conventional processing of BST ceramics requires a high sintering temperature of about 1350 ºC. In this work, we demonstrate the feasibility of using Flash sintering (FS) and corresponding maps to process Ba0.6Sr0.4TiO3 ceramics, resulting in a 350 ºC decrease in sintering temperature and 7 h reduction in sintering process time. The Flash-sintered BST exhibits a finer grain microstructure, lower dielectric loss, and significantly higher K-factor (figure of merit for tunability) when compared to conventionally sintered BST, unequivocally meeting the application criteria for tunable materials. Our findings highlight the effectiveness of FS as an alternative method for the rapid sintering of BST, while also opening up possibilities for further research into more efficient sintering processes for electroceramics.

Closed-loop Laser Volume Ablation with Adaptive Scan Paths

This research focuses on closed-loop control in laser volume ablation, also known as laser milling. Such process control enables precise ablation results on workpieces with much wider tolerances regarding the initial surface geometry, internal structure, or its response to the incident laser beam, compared to conventional open-loop processing. However, state of the art closed-loop ablation systems incorporate the process control at the cost of increased processing time. The two main causes are the alternating between processing and measuring, and the use of static scan paths that do not adapt continuously to the evolving geometry of the workpiece during processing. This study addresses this issue by proposing a parallelized work flow of processing, measuring the surface topography and adaptive path planning, eliminating interruptions and achieving faster processing through continuously optimized scan paths. The realized machining system achieved a mean reduction in processing time of 29%, 36%, and 52% on three different test geometries compared to the state of the art.

Optimization of machine learning models through quantization and data bit reduction in healthcare datasets

This study focuses on enhancing complex machine learning models through quantization and data bit reduction. The primary goal is to reduce processing time while maintaining model performance, which is particularly relevant for intricate models with prolonged execution times. The research employs two medical datasets, namely Heart Disease Prediction and Breast Cancer Detection, and applies optimization techniques to K-Nearest Neighbors (KNN) and Support Vector Machine (SVM) machine-learning models. To achieve optimization, the study employs effective quantization and data bit reduction techniques such as QuantileTransformer, Numpy.round, and KBinsDiscretizer functions. These techniques are utilized to convert input data from float64 to float32 and int32, resulting in a streamlined data representation. The trade-off between processing time and model accuracy is explored, acknowledging that some compromise in accuracy might occur after optimization. The experimentation reveals that there is a noticeable reduction in time complexity after optimization, with a marginal impact on model accuracy. Interestingly, the study concludes that the outcome and efficiency of optimization techniques are influenced not only by the specific technique used but also by the nature of the dataset and machine learning model under consideration. This comprehensive research showcases the applicability of optimization techniques, specifically quantization and data bit reduction, in complex machine learning models. By conducting experiments on medical datasets and analyzing KNN and SVM models, the study underscores the delicate balance between processing time and model accuracy. The findings emphasize that the success of optimization strategies is context-dependent, relying not only on the chosen technique but also on the interplay between the technique, model, and dataset.

GAN-based quantitative oblique back-illumination microscopy enables computationally efficient epi-mode refractive index tomography.

Quantitative oblique back-illumination microscopy (qOBM) is a novel imaging technology that enables epi-mode 3D quantitative phase imaging and refractive index (RI) tomography of thick scattering samples. The technology uses four oblique back illumination images captured at the same focal plane and a fast 2D deconvolution reconstruction algorithm to reconstruct 2D phase cross-sections of thick samples. Alternatively, a through-focus z-stack of oblique back illumination images can be used to recover 3D RI tomograms with improved RI quantitative fidelity at the cost of a more computationally expensive reconstruction algorithm. Here, we report on a generative adversarial network (GAN) assisted approach to reconstruct 3D RI tomograms with qOBM that achieves high fidelity and greatly reduces processing time. The proposed approach achieves high-fidelity 3D RI tomography using differential phase contrast images from three adjacent z-planes. A ∼9-fold improvement in volumetric reconstruction time is achieved. We further show that this technique provides high SNR RI tomograms with high quantitative fidelity, reduces motion artifacts, and generalizes to different tissue types. This work can lead to real-time, high-fidelity RI tomographic imaging for in-vivo pre-clinical and clinical applications.

The incremental design of a machine learning framework for medical records processing.

This work presents the development and evaluation of coordn8, a web-based application that streamlines fax processing in outpatient clinics using a "human-in-the-loop" machine learning framework. We demonstrate the effectiveness of the platform at reducing fax processing time and producing accurate machine learning inferences across the tasks of patient identification, document classification, spam classification, and duplicate document detection. We deployed coordn8 in 11 outpatient clinics and conducted a time savings analysis by observing users and measuring fax processing event logs. We used statistical methods to evaluate the machine learning components across different datasets to show generalizability. We conducted a time series analysis to show variations in model performance as new clinics were onboarded and to demonstrate our approach to mitigating model drift. Our observation analysis showed a mean reduction in individual fax processing time by 147.5 s, while our event log analysis of over 7000 faxes reinforced this finding. Document classification produced an accuracy of 81.6%, patient identification produced an accuracy of 83.7%, spam classification produced an accuracy of 98.4%, and duplicate document detection produced a precision of 81.0%. Retraining document classification increased accuracy by 10.2%. coordn8 significantly decreased fax-processing time and produced accurate machine learning inferences. Our human-in-the-loop framework facilitated the collection of high-quality data necessary for model training. Expanding to new clinics correlated with performance decline, which was mitigated through model retraining. Our framework for automating clinical tasks with machine learning offers a template for health systems looking to implement similar technologies.