期刊,Robotics & Machine Learning Daily News 2024年卷Apr.2期_国家学术搜索

期刊信息/Journal information

Robotics & Machine Learning Daily News

NewsRx

主办单位：NewsRx

Robotics & Machine Learning Daily News/Journal Robotics & Machine Learning Daily News

正式出版

收录年代

New Artificial Intelligence Findings from Weill Cornell Medicine Reported (Early Detection of Optic Nerve Changes On Optical Coherence Tomography Using Deep Lea rning for Risk-stratification of Papilledema and Glaucoma)

68-69页

查看更多>>摘要：By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-A new study on Artificial Intelligence is now available. According to news reporting from New York City, New York, by NewsRx journalists, research stated, "The use of artificial intelligence is beco ming more prevalence in medicine with numerous successful examples in ophthalmol ogy. However, much of the work has been focused on replicating the works of opht halmologists." The news correspondents obtained a quote from the research from Weill Cornell Me dicine, "Given the analytical potentials of artificial intelligence, it is plaus ible that artificial intelligence can detect microfeatures not readily distingui shed by humans. In this study, we tested the potential for artificial intelligen ce to detect early optic coherence tomography changes to predict progression tow ard papilledema or glaucoma when no significant changes are detected on optical coherence tomography by clinicians. Prediagnostic optical coherence tomography o f patients who developed papilledema (n = 93, eyes = 166) and glaucoma (n = 187, eyes = 327) were collected. Given discrepancy in average cup-to-disc ratios of the experimental groups, control groups for papilledema (n = 254, eyes = 379) an d glaucoma (n = 441, eyes = 739) are matched by cup-to-disc ratio. Publicly avai lable Visual Geometry Group-19 model is retrained using each experimental group and its respective control group to predict progression to papilledema or glauco ma. Images used for training include retinal nerve fiber layer thickness map, ex tracted vertical tomogram, ganglion cell thickness map, and ILM-RPE thickness ma p. Trained model was able to predict progression to papilledema with a precision of 0.714 and a recall of 0.769 when trained with retinal nerve fiber layer thic kness map, but not other image types. However, trained model was able to predict progression to glaucoma with a precision of 0.682 and recall of 0.857 when trai ned with extracted vertical tomogram, but not other image types. Area under prec ision-recall curve of 0.826 and 0.785 were achieved for papilledema and glaucoma models, respectively. Computational and analytical power of computers have beco me an invaluable part of our lives and research endeavors. Our proof-of-concept study showed that artificial intelligence (AI) algorithms have the potential to detect early changes on optical coherence tomography for prediction of progressi on that is not readily observed by clinicians."

原文链接:

NETL
NSTL

Reports from Stanford University Add New Data to Findings in Robotics (Elephant Trunk Inspired Multimodal Deformations and Movements of Soft Robotic Arms)

69-70页

查看更多>>摘要：By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-Current study results on Robotics have been published. According to news reporting from Stanford, California, by NewsR x journalists, research stated, "Elephant trunks are capable of complex, multimo dal deformations, allowing them to perform task-oriented high-degree-of-freedom (DOF) movements pertinent to the field of soft actuators. Despite recent advance s, most soft actuators can only achieve one or two deformation modes, limiting t heir motion range and applications." Financial supporters for this research include National Science Foundation (NSF) , NSF-Office of the Director (OD), ARO ECP Award. The news correspondents obtained a quote from the research from Stanford Univers ity, "Inspired by the elephant trunk musculature, a liquid crystal elastomer (LC E)-based multi-fiber design strategy is proposed for soft robotic arms in which a discrete number of artificial muscle fibers can be selectively actuated, achie ving multimodal deformations and transitions between modes for continuous moveme nts. Through experiments, finite element analysis (FEA), and a theoretical model , the influence of LCE fiber design on the achievable deformations, movements, a nd reachability of trunk-inspired robotic arms is studied. Fiber geometry is par ametrically investigated for 2-fiber robotic arms and the tilting and bending of these arms is characterized. A 3-fiber robotic arm is additionally studied with a simplified fiber arrangement analogous to that of an actual elephant trunk. T he remarkably broad range of deformations and the reachability of the arm are di scussed, alongside transitions between deformation modes for functional movement s. It is anticipated that this design and actuation strategy will serve as a rob ust method to realize high-DOF soft actuators for various engineering applicatio ns. These elephant trunk-inspired soft robotic arms, through the design of their selectively addressable liquid crystal elastomer (LCE) artificial muscle fibers , can achieve multimodal deformations and transitions between deformation modes to generate functional movements much like those of an elephant."

原文链接:

NETL
NSTL

Royal Orthopedic Hospital Reports Findings in Arthroplasty (Comparative analysis of radiation exposure in robot-assisted total knee arthroplasty using popular r obotic systems)

70-71页

查看更多>>摘要：By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-New research on Surgery-Arthroplasty is the subject of a report. According to news reporting originating from Birmin gham, United Kingdom, by NewsRx correspondents, research stated, "Robotic-assist ed TKA (RATKA) is a rapidly emerging technique that has been shown to improve pr ecision and accuracy in implant alignment in TKA. Robotic-assisted TKA (RATKA) u ses computer software to create a three-dimensional model of the patient's knee. " Our news editors obtained a quote from the research from Royal Orthopedic Hospit al, "Different types of preoperative imaging, including radiographs and CT scans , are used to create these models, each with varying levels of radiation exposur e. This study aims to determine the radiation dose associated with each type of imaging used in RATKA, to inform patients of the potential risks. A retrospectiv e search of our clinical radiology and arthroplasty database was conducted to id entify 140 knees. The patients were divided into three groups based on the type of preoperative imaging they received: (1) CT image-based MAKO Protocol, (2) Ant ero-posterior long leg alignment films (LLAF), (3) standard AP, lateral, and sky line knee radiographs. The dose of CT imaging technique for each knee was measur ed using the dose-length product (DLP) with units of mGycm, whereas the measurem ent for XRAY images was with the dose area product (DAP) with units of Gycm. The mean radiation dose for patients in the CT (MAKO protocol) image-based group wa s 1135 mGy.cm2. The mean radiation dose for patients in the LLAF group was 3081 Gycm2. The mean radiation dose for patients undergoing knee AP/lateral and skyli ne radiographs was the lowest of the groups, averaging 4.43 Gycm2. Through an AN OVA and post hoc analysis, the results between groups was statistically signific ant. In this study, we found a significant difference in radiation exposure betw een standard knee radiographs, LLAF and CT imaging."

原文链接:

NETL
NSTL

Findings from Tampere University Provides New Data about Machine Learning (Evalu ating the Performance of Machine Learning Cfd-based and Hybrid Analytical Models for Transient Flow Prediction In Temperature-compensated Digital Flow Units)

71-71页

查看更多>>摘要：By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-Fresh data on Machine Learning are pre sented in a new report. According to news originating from Tampere, Finland, by NewsRx correspondents, research stated, "This investigation utilized binary-code d, parallel-connected on-off valves that can achieve high flow rates with fewer valves while addressing flow peak challenges." Funders for this research include Ministry of Education in Finland, Egyptian Cul tural Affairs and Missions. Our news journalists obtained a quote from the research from Tampere University, "By considering temperature and refining modeling techniques, the study rectifi es certain limitations observed in previous research, such as neglecting tempera ture, imprecise valve orifice flow coefficients, and absent flow pattern visuali zation, thereby enhancing flow prediction accuracy. The results for the ML_ CFD-based model suggest that although extrapolation challenges exist in rarely d atadriven systems, the proposed approach exhibits errors under 5 % across diverse metrics, attributable to the effectiveness of well-constrained ov erparameterized models and the segmented structure of digital flow control units ."

原文链接:

NETL
NSTL

Study Results from Huazhong University of Science and Technology in the Area of Robotics Reported (Robotic Compliant Grinding of Curved Parts Based On a Designe d Active Force-controlled Endeffector With Optimized Series Elastic Component)

72-73页

查看更多>>摘要：By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-A new study on Robotics is now availab le. According to news originating from Wuhan, People's Republic of China, by New sRx correspondents, research stated, "High-accuracy, fast-response, and low-over shoot force control are important to guarantee the material removal accuracy and surface quality of robotic compliant grinding. In order to achieve the above ta rget, this study develops an active compliant force-controlled end-effector base d on a series elastic actuator for the robotic grinding of curved parts." Financial supporters for this research include National Natural Science Foundati on of China (NSFC), Natural Science Foundation of Hubei Province, National Natur al Science Foundation of China (NSFC), Natural Science Foundation of Hubei Provi nce. Our news journalists obtained a quote from the research from the Huazhong Univer sity of Science and Technology, "Firstly, a decoupled robotic grinding system co mposed of an industrial robot and an endeffector is developed, and a novel forc e-controlled end-effector is designed by adding an elastic component between the servo motor and the load to improve compliance. Secondly, the influences of the elastic component and grinding tool stiffness on the stability of the force-con trolled end-effector system are analyzed by establishing a contact model of the compliant end-effector and the transfer function of the entire force control sys tem. Then, the stiffness of the grinding tool and the elastic component are opti mized with full consideration of the system cutoff frequency and the end-effecto r compliance. To improve the force tracking accuracies of the developed complian t end-effector, a proportional-integral (PI) controller with first-order differe ntial force feedforward control is designed. Finally, grinding experiments are c onducted for verification. The effects of spring stiffness and grinding tool sti ffness on the force control system are tested, which match well with the theoret ical analysis. Grinding results show that the maximum force control error of the compliant force-controlled end-effector is decreased by 70% compa red to that of the rigid force-controlled end-effector, and the overshoot of the force control is reduced from 30 % to almost 0 under the same resp onse speed. The maximum and average absolute grinding depth errors using the des igned end-effector are reduced by 57.2% and 58.6%, re spectively, compared with those of the traditional rigid end-effector, and the a verage surface roughness of the finished part is reduced by 19.2 %."

原文链接:

NETL
NSTL

New Data from Nanjing University of Aeronautics and Astronautics Illuminate Find ings in Robotics (Enabling Collaborative Assembly Between Humans and Robots Usin g a Digital Twin System)

73-74页

查看更多>>摘要：By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-Investigators discuss new findings in Robotics. According to news reporting originating from Nanjing, People's Republi c of China, by NewsRx correspondents, research stated, "Human-robot collaboratio n (HRC) systems are intelligent systems that guide robots to collaborate with hu mans based on a cognitive understanding of human intention, ensuring safe, flexi ble, and efficient collaboration between humans and robots in shared workspaces. In industrial settings, the current methods for constructing a human digital tw in model rely on motion capture devices that require personnel to wear cumbersom e equipment, which goes against the principle of flexible interaction advocated for HRC." Financial supporters for this research include National Natural Science Foundati on of China (NSFC), China Postdoctoral Science Foundation, Jiangsu Provincial Po st-doctoral Excellence Program. Our news editors obtained a quote from the research from the Nanjing University of Aeronautics and Astronautics, "Furthermore, the current methods do not model humans and robots in a unified space, which is both unintuitive and inconvenient for perceiving and understanding the overall environment. To address these limi tations, this paper proposes a digital twin system for HRC. This system facilita tes the construction of a digital twin scene, the mapping from the real space to the virtual space, and the planning and execution of collaborative strategies f rom the virtual to the real space. Designed explicitly for common workstation se ttings, a robust human mesh recovery algorithm is introduced to address the chal lenge of reconstructing occluded human bodies. Additionally, uncertainty estimat ion is employed to enhance the action recognition algorithm, ensuring a controll able level of risk in the recognition process. Experimental results demonstrate the superiority of the proposed methods over baseline methods."

原文链接:

NETL
NSTL

Federal State Budgetary Educational Institution of Higher Education Reports Find ings in COVID-19 (Evaluation of serum and urine biomarkers for severe COVID-19)

74-75页

查看更多>>摘要：By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-New research on Coronavirus-COVID-19 is the subject of a report. According to news reporting originating in Moscow, Russia, by NewsRx journalists, research stated, "The new coronavirus disease, CO VID-19, poses complex challenges exacerbated by several factors, with respirator y tissue lesions being notably significant among them. Consequently, there is a pressing need to identify informative biological markers that can indicate the s everity of the disease." The news reporters obtained a quote from the research from the Federal State Bud getary Educational Institution of Higher Education, "Several studies have highli ghted the involvement of proteins such as APOA1, XPNPEP2, ORP150, CUBN, HCII, an d CREB3L3 in these respiratory tissue lesions. However, there is a lack of infor mation regarding antibodies to these proteins in the human body, which could pot entially serve as valuable diagnostic markers for COVID-19. Simultaneously, it i s relevant to select biological fluids that can be obtained without invasive pro cedures. Urine is one such fluid, but its effect on clinical laboratory analysis is not yet fully understood due to lack of study on its composition. Methods us ed in this study are as follows: total serum protein analysis; ELISA on moderate and severe COVID-19 patients' serum and urine; bioinformatic methods: ROC analy sis, PCA, SVM. The levels of antiAPOA1, antiXPNPEP2, antiORP150, antiCUBN, antiH CII, and antiCREB3L3 exhibit gradual fluctuations ranging from moderate to sever e in both the serum and urine of COVID-19 patients. However, the diagnostic valu e of individual anti-protein antibodies is low, in both blood serum and urine. O n the contrary, joint detection of these antibodies in patients' serum significa ntly increases the diagnostic value as demonstrated by the results of principal component analysis (PCA) and support vector machine (SVM). The non-linear regres sion model achieved an accuracy of 0.833. Furthermore, PCA aided in identifying serum protein markers that have the greatest impact on patient group discriminat ion."

原文链接:

NETL
NSTL

Study Results from Guangdong Academy of Science Update Understanding of Machine Learning (Accurate and Rapid Image Segmentation Method for Bayberry Automatic Pi cking Via Machine Learning)

75-76页

查看更多>>摘要：By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-A new study on Machine Learning is now available. According to news reporting originating from Guangzhou, People's Rep ublic of China, by NewsRx correspondents, research stated, "Due to the short rip ening period and complex picking environment, bayberry generally relies on mecha nical equipment for picking, especially the automatic picking system guided by v ision. Thus, it is crucial to locate the bayberry in the view accurately and rap idly." Funders for this research include Guangdong Provincial Rural Revitalization Stra tegy Special Fund Project, GDAS'Project of Science and Technology Development. Our news editors obtained a quote from the research from the Guangdong Academy o f Science, "Although efforts have been made, the existing methods are difficult to implement due to the limited amount of data and the processing speed. In this study, an accurate and rapid segmentation method based on machine learning was proposed to address this problem. First, the images collected by the visual guid ance system were pre-processed by contrast-limited adaptive histogram equalizati on (CLAHE) based on the Y component of the YUV color space. Taking advantage of the color difference map of RB and RG for the segmentation of different colors, an adaptive color difference map foreground segmentation method was then adopted for bayberry region foreground segmentation. Finally, distance transforms and m arking control watershed methods were exploited to achieve single bayberry fruit segmentation. Furthermore, with the help of the convex hull theory and fruit sh ape characteristics, the irregular background interference areas were filtered o ut, which improved the accuracy of bayberry segmentation performance."

原文链接:

NETL
NSTL

Investigators at Nanyang Technological University Describe Findings in Robotics (Fine Robotic Manipulation Without Force/torque Sensor)

76-77页

查看更多>>摘要：By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-Investigators discuss new findings in Robotics. According to news reporting originating from Singapore, Singapore, by NewsRx correspondents, research stated, "Force Sensing and Force Control are ess ential to many industrial applications. Typically, a 6-axis force/torque (F/T) s ensor is installed between the robot's wrist and the end effector to measure the forces and torques exerted by the environment on the robot (the external wrench )." Financial support for this research came from National Research Foundation. Our news editors obtained a quote from the research from Nanyang Technological U niversity, "While a typical 6-axis F/T sensor can offer highly accurate measurem ents, it is expensive and vulnerable to drift and external impacts. Existing met hods aiming at estimating the external wrench using only the robot's internal si gnals are limited in scope. For instance, the estimation accuracy has mainly bee n validated in free-space motions and simple contacts, rather than tasks like as sembly that require high-precision force control. In this letter, we present a N eural-Network-based solution to overcome these challenges. We offer a detailed d iscussion on model structure, training data categorization and collection, as we ll as fine-tuning strategies. These steps enable precise and reliable wrench est imations across a variety of scenarios."

原文链接:

NETL
NSTL

Inner Mongolia Agricultural University Reports Findings in Machine Learning (Dam age detection of road domain waveform guardrail structure based on machine learn ing multi-module fusion)

77-77页

查看更多>>摘要：By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-New research on Machine Learning is th e subject of a report. According to news reporting originating from Inner Mongol ia, People's Republic of China, by NewsRx correspondents, research stated, "The current highway waveform guardrail recognition technology has encountered proble ms with low segmentation accuracy and strong noise interference. Therefore, an i mproved U-net semantic segmentation model is proposed to improve the efficiency of road maintenance detection." Financial support for this research came from Department of Science and Technolo gy of Inner Mongolia Autonomous Region. Our news editors obtained a quote from the research from Inner Mongolia Agricult ural University, "The model training is guided by mixed expansion convolution an d mixed loss function, while the presence of guardrail shedding is investigated by using partial mean values of gray values in ROI region based on segmentation results, while the first-order detail coefficients of wavelet transform are appl ied to detect guardrail defects and deformation. It has been determined that the Miou and Dice of the improved model are improved by 8.63% and 17. 67%, respectively, over the traditional model, and that the method of detecting defects in the data is more accurate than 85%."

原文链接:

NETL
NSTL