Search

Background: Issuing of correct prescriptions is a foundation of patient safety. Medication errors represent one of the most important problems in health care, with 'look-alike and sound-alike' (LASA) being the lead error. Existing solutions to prevent LASA still have their limitations. Deep learning techniques have revolutionized identification classifiers in many fields. In search of better image-based solutions for blister package identification problem, this study using a baseline deep learning drug identification (DLDI) aims to understand how identification confusion of look-alike images by human occurs through the cognitive counterpart of deep learning solutions and thereof to suggest further solutions to approach them.Methods: We collected images of 250 types of blister-packaged drug from the Out-Patient Department (OPD) of a medical center for identification. The deep learning framework of You Only Look Once (YOLO) was adopted for implementation of the proposed deep learning. The commonly-used F1 score, defined by precision and recall for large numbers of identification tests, was used as the performance criterion. This study trained and compared the proposed models based on images of either the front-side or back-side of blister-packaged drugs.Results: Our results showed that the total training time for the front-side model and back-side model was 5 h 34 min and 7 h 42 min, respectively. The F1 score of the back-side model (95.99%) was better than that of the front-side model (93.72%).Conclusions: In conclusion, this study constructed a deep learning-based model for blister-packaged drug identification, with an accuracy greater than 90%. This model outperformed identification using conventional computer vision solutions, and could assist pharmacists in identifying drugs while preventing medication errors caused by look-alike blister packages. By integration into existing prescription systems in hospitals, the results of this study indicated that using this model, drugs dispensed could be verified in order to achieve automated prescription and dispensing. [ABSTRACT FROM AUTHOR]

Deadlocks are an important problem in resource allocation systems such as flexible manufacturing systems. The theory of regions and the siphon-based method are usually used in the most deadlock prevention policies. The theory of regions that can obtain a maximally permissive controller is usually considered to be a natural solution with seasonable computational cost for flexible manufacturing systems. The selective siphon method allows one to use fewer control places than the conventional one. This paper employs both methods above. The former can identify the set of curial marking/transition-separation instance; the latter can reduce the computational cost. We can infer that the novel policy is the most efficient policy than the traditional methods, and also, the maximal permissive behavior of Petri net models can still be obtained. [ABSTRACT FROM AUTHOR]

This paper proposes an image sequence-based navigation method under the teaching-replay framework for robots in piecewise linear routes. Waypoints used by the robot contain either the positions with large heading changes or selected midway positions between junctions. The robot applies local visual homing to move between consecutive waypoints. The arrival at a waypoint is determined by minimizing the average vertical displacements of feature correspondences. The performance of the proposed approach is supported by extensive experiments in hallway and office environments. While the homing speed of robots using other approaches is constrained by the speed in the teaching phase, our robot is not bounded by such limit and can travel much faster without compromising the homing accuracy. [ABSTRACT FROM AUTHOR]

Optimization approach to define geometry (parameters) of the radial matching section in an RFQ accelerator is suggested. This approach gives wider opportunities for the design of the radial matching section because it does not have certain prescribed laws of variation of focusing strength along the section. [ABSTRACT FROM AUTHOR]

Mathematical models of the structural parametric optimization of plasma dynamics are discussed. Optimization approach to plasma dynamic is based on the consideration of trajectory ensemble. This ensemble describes transient process in tokamak subject to the initial data and external disturbances. In the framework of this approach the optimization of dynamics of the trajectory ensemble in ITER tokamak is given. The trajectories of this ensemble are perturbed at the initial point set and the set of external disturbances. [ABSTRACT FROM AUTHOR]