Complex Protocols Research Articles

Cervical kyphoscoliosis due to substance abuse: a case series on introduction and neurosurgical treatment.

A prospective case series. To investigate the possible associations of multi-directional cervical kyphoscoliosis (CKS) with substance abuse and evaluate treatment outcomes. CKS is a rare medical situation without a well-established pathology. Still there are many obscurities in treatment paradigms and outcomes. To our knowledge, there is no published report on the association between CKS with addiction. Besides the novel report of such a rare etiology for CSK, the complexity of the neurosurgical treatment makes this report the first of its kind. In this series, we presented three patients with a history of crack cocaine addiction who suffered CKS due to prolonged hand-over-neck posture and treated all of them with a complex neurosurgical protocol. All cases were operated on and restored their normal spine alignment. Their long-term outcomes showed independent neurological status with no major surgical complications. Crack cocaine addiction might be associated with CKS in long-term cocaine abuse. A complex neurosurgical approach can achieve a sustainable clinical outcome.

Long-Term Nose Evaluation After Treatment of Bilateral Cleft Lip Using the Spina Technique.

Bilateral cleft lip (BCL) repair is a challenge due to anatomic and functional distortions, worse than unilateral cleft. With regard to the nose, BCL can result in certain characteristics, such as a short columella, a flattened and widened tip, and horizontally oriented nostrils. Lip and nose treatment is advocated by some surgeons to achieve a better outcome. One of the techniques used to treat this condition is the one recommended by Victor Spina, in a complex protocol involving more than one surgical procedure and long-term follow-up. The aim of this study is to evaluate the growth of the nose in patients treated using the Spina technique. This was a retrospective and analytical study of the medical records of patients who were treated using the Spina technique at the Assistance Center for Cleft Lip and Palate (CAIF) at Hospital do Trabalhador in Curitiba, Brazil. Anthropometric measurements of interalar distance and nose tip projection were made from frontal and submental photographs taken at three moments: pre-first surgery, pre-definitive surgery, and late follow-up. Furthermore, other analyzed parameters were: sex, date of birth, age at time of surgery, and late follow-up. With these data, a comparison of nose growth was made between patients with BCL and a population without craniofacial anomalies observed in a study by FARKAS et al (1992). The comparison also took sex and age variables into account. Regarding the interalar distance measurements, both sexes showed a widening of the nose before starting treatment. This widening was maintained throughout follow-up. The growth curve, in turn, remained similar among males, with an average growth rate of only 3.2% higher in patients with BCL. Among females, the difference was one of 16.92%. In terms of nasal tip projection, there was an average increase of 1.12cm (86.15%) in males and 0.69cm (57.02%) in females. It was possible to notice a nasal growth rate in those with BCL similar to the normal pattern and a significant improvement in nasal tip projection in male patients. However, with regard to the data of both sexes, but especially females, the low number of patients jeopardized the significance of the results.

The Effectiveness of High Intensity Laser in Improving Motor Deficits in Patients with Lumbar Disc Herniation.

High-Intensity Laser (HIL) therapy, known for its biostimulatory effects on nerve cell growth and repair, shows promise for improving motor deficits caused by morphopathological changes. This research study aimed to comprehensively assess muscle strength changes through muscle testing, complemented by functional tests evaluating factors contributing to disability in patients with Lumbar Disc Herniation (LDH) and associated motor impairment, following a complex rehabilitation protocol incorporating HIL therapy. A total of 133 individuals with LDH and motor deficits were divided into two groups. Group 1 (n = 66) received HIL therapy followed by standard rehabilitation, while Group 2 (n = 67) underwent only the standard rehabilitation program. Functional parameters, including muscle strength, the ability to walk on tiptoes or heels, and self-assessed fall risk, were monitored. Both groups showed statistically significant improvements in all monitored parameters. A comparative analysis revealed a significant result for the HIL therapy regimen across all indicators. The group undergoing a rehabilitation program with integrated HIL therapy displayed significantly greater improvement in motor deficits, affirming the positive impact of HIL therapy on functional parameters among LDH patients.

Location-aware job scheduling for IoT systems using cloud and fog

The rapid growth of Internet of Things (IoT) applications generates vast volumes of high-speed data streams from numerous devices. Cloud computing solutions often handle and manage this data; however, for certain applications, the latency introduced by transmitting data from edge devices to the cloud may be unacceptable. This issue is exacerbated by the bandwidth constraints of public networks, which become significant barriers in large-scale IoT implementations. Consequently, effective resource management, service management, data storage, and power management require more robust infrastructure and complex protocols. An "intelligent gateway" based on fog computing can enhance the efficient utilization of cloud and network resources. Resource planning and management in a fog-cloud environment significantly impact system performance, particularly latency. This problem is known to be NP-hard. This paper addresses the challenge of lifetime optimization for scheduling data-intensive tasks in fog and cloud-based IoT systems. Initially, we propose an Integer Linear Programming (ILP) optimization model to frame the problem. Subsequently, we introduce a heuristic method named Data-Locality Aware Job Scheduling in Fog-Cloud (DLSFC), which is derived from the proposed formulation. The effectiveness of DLSFC is evaluated across various system characteristics, with results demonstrating that DLSFC achieves solutions within 85 % of the ideal outcome provided by the LP solver.

Alternative Method for Obtaining Human-Induced Pluripotent Stem Cell Lines and Three-Dimensional Growth: A Simplified, Passage-Free Approach that Minimizes Labor.

Induced pluripotent stem cells (iPSCs) hold significant promise for numerous applications in regenerative medicine, disease modeling, and drug discovery. However, the conventional workflow for iPSC generation, with cells grown under two-dimensional conditions, presents several challenges, including the need for specialized scientific skills such as morphologically assessing and picking colonies and removing differentiated cells during the establishment phase. Furthermore, maintaining established iPSCs in three-dimensional culture systems, while offering scalability, necessitates an enzymatic dissociation step for their further growth in a complex and time-consuming protocol. In this study, we introduce a novel approach to address these challenges by reprogramming somatic cells grown under three-dimensional conditions as spheres using a bioreactor, thereby eliminating the need for two-dimensional culture and colony picking. The iPSCs generated in this study were maintained under three-dimensional conditions simply by transferring spheres to the next bioreactor, without the need for an enzymatic dissociation step. This streamlined method simplifies the workflow, reduces technical variability and labor, and paves the way for future advancements in iPSC research and its wider applications. Key features • Establishment of induced pluripotent stem cells in a three-dimensional environment. • Maintenance and cryopreservation of iPSCs without the need for a dissociation step.

A Credibility Monitoring Approach and Software Monitoring System for VHF Data Exchange System Data Link Based on a Combined Detection Method

Due to VDES’s higher data transmission speed and complex communication protocols, vulnerabilities within its data link infrastructure are more pronounced. To ensure the reliability of VDES data transmission, this manuscript proposes a credibility monitoring approach based on the combined detection method of radio interference detection and spoofing source identification and localization, focusing on key data link vulnerabilities outlined in the IALA G1181 VDES VDL Integrity Guide. Automated monitoring is achieved through VDES data link monitoring software (VDES(AIS 2.0)), which is based on a three-tier architecture and a Client/Server (C/S) model. The software validates monitoring techniques and software against various interference scenarios. Visualization of monitoring results, alarm notifications, and relevant data through the front-end interface enhances understanding of VDES data link credibility. This framework supports effective surveillance and detection of vulnerabilities, such as radio interference and spoofing sources.

The use of digital twin for mobile robot swarm task allocation

Cyber-physical systems, with the advent of the Internet of Things (IoT), have paved the way for the rise of the Industrial Internet of Things (IIoT) in industrial environments. A significant manifestation of this synergy is the Digital Twin—a virtual representation of industrial entities, including robots. This paper examines the fusion of digital twins in swarm robotics, emphasizing communication, and task allocation among robots. This study employs several mobile robots interconnected via a wireless network, each equipped with a Raspberry Pi and specific sensors. A centralized server, acting as both a Docker host and a Docker registry, facilitates task allocation. The results indicate effective communication among robots, with the digital twin playing an instrumental role in safety assessment and task allocation, as well as deploying Docker containers using Portainer and GitLab APIs. Unity3D aids in visualizing IMU motion and orientation. The motivation behind this study stems from the need to enhance efficiency and safety in industrial settings through advanced automation and coordination in swarm robotics. This research not only demonstrates the feasibility of integrating digital twins with IoT-enabled robots but also sets the stage for future investigations into more complex task allocation and safety protocols in industrial environments. By leveraging digital twins in swarm robotics, we open new pathways for research into more adaptive, responsive, and efficient industrial systems.

Administration sequences in single-day chemotherapy regimens for breast cancer: a comprehensive review from a practical perspective.

Breast cancer is one of the most prevalent malignant tumors globally, posing a severe threat to human life and health. Chemotherapy, a cornerstone in the treatment of breast cancer, often overlooks the sequence of drug administration within single-day regimens. This study aims to explore the impact of drug administration order on the efficacy and toxicity of combination chemotherapy protocols for breast cancer. Through a comprehensive review and analysis based on current evidence from evidence-based medicine, we delved into how the order of drug administration affects both efficacy and toxicity. We systematically classified and analyzed commonly used combination drug regimens, providing graded recommendations and a reasoned analysis to offer valuable references for clinical decision-making. Our findings indicate that the sequence of drug administration in complex combination chemotherapy protocols is not arbitrary but necessitates multifaceted considerations. Rational drug sequencing can maximize synergistic effects between drugs, thereby augmenting therapeutic efficacy while effectively mitigating drug-related adverse effects. Additionally, some drug labels and clinical trials have explicitly highlighted the therapeutic benefits of specific drug sequences. This study underscores the importance of considering the sequence of drug administration in clinical practice. It is recommended to prioritize the sequential drug administration according to official drug product labeling, while also considering factors such as the administration sequence from large randomized controlled trials, cell proliferation kinetics specific to cancer types, drug interactions, chronopharmacology, drug irritability, clinical experiences, and patient preferences. By taking these factors into account, the goal is to maximize treatment efficacy and minimize the occurrence of adverse reactions.

Catalytic Generation of Pyridyl Radicals via Electron Donor-Acceptor Complex Photoexcitation: Synthesis of 2-Pyridylindole-Based Heterobiaryls.

We report the catalytic generation of pyridyl radicals through photoexcitation of the electron donor-acceptor (EDA) complex, which enables the C2-selective heteroarylation of indole under ambient conditions. In this manifold, catalytic triarylamine and chloropyridine aggregate into an EDA complex in the presence of an inorganic base, making readily available chloropyridines good precursors for the generation of diverse pyridyl radicals. Given the broad reaction scope, this catalytic EDA complex protocol provides robust access to heterobiaryl scaffolds that are widely present in biologically important molecules.

Designing a Website to Support a Multisite, International Clinical Trial.

Safe and reliable implementation of complex study protocols in multisite clinical trials requires that all study personnel have 24/7 access to up-to-date study materials. Study websites can serve as an electronic manual of operations (eMOO) to support trial conduct. We describe the development, organization, and maintenance of a study website and eMOO to facilitate quick and efficient communication during conduct of a complex, multisite, international clinical trial. We worked closely with our information technology (IT) department to develop and maintain our study website, which includes a public home page, a section for parents and families, and three password-protected portals that serve as an eMOO for (a) study sites, (b) study site investigators, and (c) the operations team (e.g., clinical coordination center, data coordination center). The public home page is helpful for families contemplating study participation and for nonparticipating sites considering joining our trial. The patient and family education section supports family participation in the study. The study site portal contains all information needed for local study teams to safely manage a study patient. The investigator portal provides access to research-specific materials needed to lead the study at each site. The operations team portal supports overall study management. For other scientists considering use of a study website for their multisite research, we recommend close collaboration with IT for development and maintenance, limited and clearly defined roles for version control, and use of unmodifiable file formats to prevent unapproved alterations of study materials. While investment in development and maintenance has been significant, we have appreciated marked value to our operations team and study sites. Our study website development process is relevant to other scientists conducting multisite clinical research.

'SNOMEDizing' Questionnaires for Standardizing Stroke Registry Data.

International interoperability of healthcare and research data requires a commitment to standards. To this end, SNOMED CT was evaluated for representing questionnaire items of the European Registry of Stroke Care Quality using a complex annotation protocol. The agreement between validators and annotators was 72.4%. At least 64% of the information could be represented by using SNOMED CT only, including complex post-coordinations. 9% of the information would require an information model, and 14% the addition of new content to SNOMED CT. Next steps will be the creation of an annotation guideline for questionnaires, a specific reference set, and the combination of both with an information model such as HL7 FHIR.

Extraction of Bacterial Membrane Vesicle and Phage Complex by Density Gradient Ultracentrifugation.

The bacterial membrane vesicles (MVs) are non-replicative, nanoscale structures that carry specific cargos and play multiple roles in microbe-host interactions. An appropriate MV isolation method that mimics complex pathogen infections in vivo is needed. After bacterial MVs extraction, flagella or pili can be frequently observed along with MVs by transmission electron microscope (TEM). Recently, MVs from Pseudomonas aeruginosa were found to coexist with Pf4 phages, and this MV-phages complex exhibited a different impact on host cell innate immunity compared with MVs or phages solely. The presence of this MVs-phages complex simulates the real condition of complex pathogen infections within the host. This protocol outlines the extraction of the MVs and Pf4 phages complex of P. aeruginosa PAO1, including the respective isolation and qualification approaches. Our step-by-step bacterial MVs-phages complex extraction protocol provides valuable insights for further studying microbe-host cell interactions and the development of novel phage therapies. Key features • Detailed density gradient extraction procedures of MVs-phages complex • TEM, plaque assay, and PCR to verify the coexistence of MVs and phages • The obtained MVs-phages complex can be used for exploring phage-microbe-host cell interactions Graphical overview.

Detecting DDoS Attacks using Machine Learning: Survey

Phishing attacks have increased dramatically in recent years affecting many areas of society. Phishing attempts often use DDoS attacks to flood a server with too many requests, overwhelming it. DDoS attacks represent a major threat to cybersecurity and pose a significant risk to computer networks. Creating a solid defense system against these attacks is essential but complex due to the wide range of attack methods and complex networks and communication protocols. Ransom demands, revenge, rivalry, or other motives may trigger attacks. This survey discusses DDoS attacks, the advantages and disadvantages of detecting DDoS using machine and deep learning, and a framework for detection using machine learning and deep learning. And use their classifiers to detect DDoS attacks. Furthermore, we explore datasets used in related works. This research is necessary because DDoS attacks are diverse and pose a significant threat to computer networks.

Continuous Insulin Therapy to Prevent Post-Transplant Diabetes Mellitus: A Randomized Controlled Trial

Rationale & ObjectivesHyperglycemia is frequently observed early after transplantation and associated with development of post-transplant diabetes mellitus (PTDM). Here, we assessed continuous subcutaneous insulin infusion (CSII) targeting afternoon hyperglycemia. Study DesignOpen label randomized parallel three-arm design. Settings & ParticipantsN=85 kidney transplant recipients (KTRs) without previous diabetes diagnosis were randomized to postoperative CSII therapy, basal insulin or control. InterventionsInsulin was to be initiated at afternoon capillary blood glucose ≥140 mg/dL (7.8 mmol/L; CSII and basal insulin) or fasting plasma glucose ≥200 mg/dL (11.1 mmol/L; control). OutcomesHbA1c at 3 months post transplant (primary endpoint). PTDM assessed by OGTT at 12 and 24 months. ResultsCSII therapy lasted until median day 18 and maximum day 88. Median HbA1c at month 3 was 5.6% (38 mmol/mol) in the CSII group versus 5.7% (39 mmol/mol) in the control group (p=0.70) and 5.4% (36 mmol/mol) in the basal insulin group (p=0.02). At months 12 and 24, the odds for PTDM were similar compared to the control group (odds ratios [95% confidence intervals], 0.80 [0.18-3.49] and 0.71 [0.15-3.16], respectively) and the basal insulin group 0.96 [0.18-5.68] and 1.51 [0.24-12.84], respectively). Mild hypoglycemia events occurred in the CSII and the basal insulin groups. LimitationsThis study is limited by outdated insulin pump technology, frequent discontinuations of CSII, a complex protocol and concerns regarding reliability of HbA1c measurements. ConclusionsCSII therapy was not superior at reducing HbA1c at month 3 or PTDM prevalence at months 12 and 24 compared to the control or basal insulin group.

Planning Service Protocols for Extra-Long Trains with Transfers

This paper presents a modeling framework for optimizing operational protocols of extra-long trains (XLTs) in metro systems (i.e., trains longer than station platforms). With the rising travel demand in megacities, metro systems face challenges such as overcrowded stations, delays, and passenger anxieties. XLTs have been proposed as a promising solution to increase metro line capacity without additional infrastructure construction. The study explores the trade-offs between the additional capacity gained through complex protocols, the potential benefits of protocols with inline transfers, and the importance of effective passenger information systems from both passengers’ and operators’ perspectives. Mathematical programs are proposed to optimize protocols for a given demand distribution and to estimate the maximum line capacity of an XLT system. The benefits of implementing XLTs are evaluated in hypothetical and real-world cases with varying demand distributions and network sizes. The results demonstrate significant capacity increases ranging from 24% to 126% as compared with regular train operations, depending on system parameters and demand distribution. These findings demonstrate promise for using such systems to improve metro line capacity in the real world. History: This paper has been accepted for the Transportation Science Special Issue on ISTTT 25 Conference. Funding: The work was supported in part by the University of Illinois, Urbana Champaign [Grant Grainger STII Seed Fund] and the Zhejiang University-University of Illinois Urbana-Champaign Institute [Grant DREMES-202001]. Supplemental Material: The online appendix is available at https://doi.org/10.1287/trsc.2024.0527 .

Multiplex, high-throughput method to study cancer and immune cell mechanotransduction

Studying cellular mechanoresponses during cancer metastasis is limited by sample variation or complex protocols that current techniques require. Metastasis is governed by mechanotransduction, whereby cells translate external stimuli, such as circulatory fluid shear stress (FSS), into biochemical cues. We present high-throughput, semi-automated methods to expose cells to FSS using the VIAFLO96 multichannel pipetting device custom-fitted with 22 G needles, increasing the maximum FSS 94-fold from the unmodified tips. Specifically, we develop protocols to semi-automatically stain live samples and to fix, permeabilize, and intracellularly process cells for flow cytometry analysis. Our first model system confirmed that the pro-apoptotic effects of TRAIL therapeutics in prostate cancer cells can be enhanced via FSS-induced Piezo1 activation. Our second system implements this multiplex methodology to show that FSS exposure (290 dyn cm−2) increases activation of murine bone marrow-derived dendritic cells. These methodologies greatly improve the mechanobiology workflow, offering a high-throughput, multiplex approach.

Scalable Distributed Inverted List Indexes in Disaggregated Memory

Memory disaggregation separates compute (CPU) and main memory resources into disjoint physical units to enable elastic and independent scaling. Connected via high-speed RDMA-enabled networks, compute nodes can directly access remote memory. This setting often requires complex protocols with many network roundtrips as memory nodes have near-zero compute power. In this paper, we design a scalable distributed inverted list index for disaggregated memory architectures. An inverted list index maps a set of terms to lists of documents that contain this term. Current solutions either partition the index horizontally or vertically with severe limitations in the disaggregated memory setting due to data and access skew, high network latency, or out-of-memory errors. Our method partitions lists into fixed-size blocks and spreads them across the memory nodes to balance skewed accesses. Block-based list processing keeps the memory footprint of compute nodes low and masks latency by interleaving remote accesses with expensive list operations. In addition, we propose efficient updates with optimistic concurrency control and read-write conflict detection. Our experiments confirm the efficiency and scalability of our method.

#1805 Renal progenitor cells isolated from urine of IgA nephropathy patients formed renal spheroids that can recapitulate the typical IgA1 deposition

Abstract Background and Aims Recent developments in stem cell biology have made it possible to create spheroids and kidney organoids from human and mouse pluripotent stem cells and renal stem cells. Spheroids are a suitable in vitro model for studies of drug studies and can transform into organoids whose rapidly developing field has proven useful in simulating nephrotoxicity, kidney disorders, and kidney development. However, to date in all studies the generation of organoids and tubuloids required the use of sophisticated and complex differentiation protocols involving external administration of several growth factors and needs compliance with a precise and specific timescale. Here we showed for the first time that human adult renal progenitor cells (ARPCs) can be isolated from urines of healthy subjects (HS) or IgA Nephropathy (IgAN) patients and can form spheroids generating very long tubule-like structures naturally, without the external use of chemokines or growth factors to artificially induce the process and recapitulating the IgA1 deposition that is the typical characteristic of the disease. Method Spheroids and tubule-like structures were characterized by whole-mounting and frozen section immunofluorescence (IF) assays and by Flow Cytometry Analysis using CD133, NanoG, Oct3/4, GATA-3, SSEA4, CD249, Aminopepidase N, ZO-1, Uromodulin, and Lotus Tetragonolobus Lectin (LTL) antibodies. ELISA was used to determine the renin secretion and CAM assays to determine ARPCs and ARPC-derived Spheroids angiogenic properties. PKH-26 labeling was used for their in vitro and in vivo tracking. Serum of HS and IgAN patients in culture with spheroids was used to recapitulate the disease and IgA1 deposits were measured by IF. Results ARPC-derived spheroids reflected the renal progenitor properties and recapitulated organoids generating very long tubule-like structures that expressed structural and functional renal tubule markers such as CD249, Aminopepidase N, ZO-1, Uromodulin, and LTL and in some cases shared many structural similarities with some regions of nephrons, such as the distal convoluted tubule, the loop of Henle and proximal convoluted tubules. Moreover, ARPC-derived spheroids secreted elevated levels of renin and, as the cellular counterpart, showed angiogenic properties. To study the proof of concept that ARPC-derived spheroids can be used to establish in vitro models recapitulating renal diseases, we generated IgAN patient-specific renal spheroids using the ARPC mixed cell population method. ARPC-derived spheroids of IgAN patients showed deposits of IgA1 already after four days of culture with patient serum (p = 0.004) and, with a different pattern, after 8 and 15 days (p = 0.0029 and p = 0.0012, respectively). The IgA1 deposition resembled that in the glomeruli of patients diagnosticated with IgAN. In contrast, no positive signal was detected in either IgAN spheroids in culture with FBS, or with inactivated IgAN serum, as well as in spheroids derived from urine of HS in culture with IgAN serum. Conclusion The ability of ARPCs to form spheroids and differentiate into tubular-like structures without the need for external chemokines or growth factors is a significant advancement in the field, even more considering that these cells can be isolated from the urine of patients. These findings open up new possibilities for the regenerative medicine in studying kidney development, disease mechanisms, and regenerative therapy development for renal disorders.

ForCEPSS—A framework for cardiac electrophysiology simulations standardization

Background and Objective:Simulation of cardiac electrophysiology (CEP) is an important research tool that is increasingly being adopted in industrial and clinical applications. Typical workflows for CEP simulation consist of a sequence of processing stages starting with building an anatomical model and then calibrating its electrophysiological properties to match observable data. While the calibration stages are common and generalizable, most CEP studies re-implement these steps in complex and highly variable workflows. This lack of standardization renders the execution of computational CEP studies in an efficient, robust, and reproducible manner a significant challenge. Here, we propose ForCEPSS as an efficient and robust, yet flexible, software framework for standardizing CEP simulation studies. Methods and Results:Key processing stages of CEP simulation studies are identified and implemented in a standardized workflow that builds on openCARP11https://opencarp.org/. Plank et al. (2021) and the Python-based carputils22https://git.opencarp.org/openCARP/carputils. framework. Stages include (i) the definition and initialization of action potential phenotypes, (ii) the tissue scale calibration of conduction properties, (iii) the functional initialization to approximate a limit cycle corresponding to the dynamic reference state according to an experimental protocol, and, (iv) the execution of the CEP study where the electrophysiological response to a perturbation of the limit cycle is probed. As an exemplar application, we employ ForCEPSS to prepare a CEP study according to the Virtual Arrhythmia Risk Prediction protocol used for investigating the arrhythmogenic risk of developing infarct-related ventricular tachycardia (VT) in ischemic cardiomyopathy patients. We demonstrate that ForCEPSS enables a fully automated execution of all stages of this complex protocol. Conclusion:ForCEPSS offers a novel comprehensive, standardized, and automated CEP simulation workflow. The high degree of automation accelerates the execution of CEP simulation studies, reduces errors, improves robustness, and makes CEP studies reproducible. Verification of simulation studies within the CEP modeling community is thus possible. As such, ForCEPSS makes an important contribution towards increasing transparency, standardization, and reproducibility of in silico CEP experiments.

Automatic Evaluation Method for Functional Movement Screening Based on a Dual-Stream Network and Feature Fusion

Functional Movement Screening (FMS) is a movement pattern quality assessment system used to assess basic movement capabilities such as flexibility, stability, and pliability. Movement impairments and abnormal postures can be identified through peculiar movements and postures of the body. The reliability, validity, and accuracy of functional movement screening are difficult to test due to the subjective nature of the assessment. In this sense, this paper presents an automatic evaluation method for functional movement screening based on a dual-stream network and feature fusion. First, the RAFT algorithm is used to estimate the optical flow of a video, generating a set of optical flow images to represent the motion between consecutive frames. By inputting optical flow images and original video frames separately into the I3D model, it can better capture spatiotemporal features compared to the single-stream method. Meanwhile, this paper introduces a simple but effective attention fusion method that combines features extracted from optical flow with the original frames, enabling the network to focus on the most relevant parts of the input data, thereby improving prediction accuracy. The prediction of the four categories of FMS results was performed. It produced better correlation results compared to other more complex fusion protocols, with an accuracy improvement of 3% over the best-performing fusion method. Tests on public datasets showed that the evaluation metrics of the method proposed in this paper were the most advanced, with an accuracy improvement of approximately 4% compared to the currently superior methods. The use of deep learning methods makes it more objective and reliable to identify human movement impairments and abnormal postures.