Search

The development of body sensor networks (BSNs) with rich multimodal signals has enabled highly accurate fine-grained action detection, which is the cornerstone of many human–computer interaction applications. However, in the case of consecutive fine-grained actions, most existing wearable sensor-based detection methods are constrained by sliding windows because of their limited temporal receptive fields, and existing sequence-to-sequence detection methods cannot effectively leverage the potential of multimodal information of wearable sensors. Herein, to give multimodal signals full play in fine-grained action detection, we propose a novel temporal convolutional network by designing a channel attention-based multistream structure. We apply it to a promising application for correct and incorrect patient transfer nursing action detection. A data set is collected from a BSN on a patient when nurses perform patient transfer. Extensive experiments on our data set and public data set (C-MHAD) demonstrate that the proposed method is superior to the state-of-the-art methods, because it can strengthen the utilization of prediction features from the more convincing modal stream at each time frame. The source code is available at https://github.com/zzh-tech/Continuous-Action-Detection .

To more efficiently enhance the patient transfer skills of nursing students, this study aims to integrate a transfer skills evaluation system and a robot patient. The evaluation parameters, namely, the translational acceleration of the waist, rotational speed of the chest, and joint angles of the shoulder, hip, and knee, were selected on the basis of the pre-experimental results obtained with a simulated patient acted by the human individuals. To measure these parameters, inertial measurement unit (IMU) and angular position sensors were installed on the robot patient. An experiment was conducted with four nursing teachers to verify whether the robot patient could distinguish the incorrect methods of the transfer skills, determined to be a common mistake made by the nurses. According to the results, most transfer steps had the same effect on the simulated patient and the robot patient, which demonstrates that the robot patient is a suitable substitute for an actual patient. However, in certain steps, the robot patient was not able to distinguish between the correct and incorrect methods using the chosen parameters owing to the differences being insignificant. These insignificant differences were mostly attributed to the passive joint design of the robot patient.

With the number of hospital stays increasing, nurses require more training to handle a variety of patients. However, time for training in nursing schools is limited, and students lack the opportuni...

Recently, human patient simulators have been widely developed as substitutes for real patients with the objective of applying them as training tools in nursing education. Such simulated training is perceived as beneficial for imparting the required practical skills to students. Considering the aging world population, this study aimed to develop a robot patient for training nursing students in the sit-to-stand (STS) transfer skill, which is indispensable in caring for elderly people. To assess a student’s skill, the robot patient should be able to access the skill correctness and behave according to whether the skill is correctly or incorrectly implemented. Accordingly, an STS control method was proposed to reproduce the different STS movements during correct and incorrect applications of the skill by the nurses. The lower limbs of a prototype robot were redesigned to provide an active joint with a compliant unit, which enables the measurement of external torque and flexibility of the human joint to be reproduced. An experiment was conducted with four nurse teachers, each of whom was asked to demonstrate both correct and incorrect STS transfer skills. The results of the external torque and joint torque measured in robot’s lower limbs revealed that a significant difference (p &lt, 0.05) between correct and incorrect skills. It also indicates the introduction of the proposed control method for the robot can satisfy the requirement of the assessment of STS skill. Among the various measurements conducted, the external torque of the hip joint exhibited the most significant difference and therefore represented the most robust measure for assessing whether the STS transfer skill was correctly applied.

A nursing action evaluation system that can assess the performance of students practicing patient handling related nursing skills becomes an urgent need for solving the nursing educator shortage problem. Such an evaluation system should be designed with less hand-crafted procedures for its scalability. Additionally, realizing high accuracy of nursing action recognition, especially fine-grained action recognition remains a problem. This reflects in the recognition of the correct and incorrect methods when students perform a nursing action, and low accuracy of that would mislead the nursing students. We propose a multi-attention deep recurrent neural network (MA-DRNN) model for nursing action recognition by directly processing the raw acceleration and rotational speed signals from wearable sensors. Data samples of target nursing actions in a nursing skill called patient transfer were collected to train and compare the models. The experiment results show that the proposed model can reach approximately 96% recognition accuracy for four target fine-grained nursing action classes helped by the attention mechanism on time and layer domains, which outperforms the state-of-the-art models of wearable sensor-based HAR.

To improve the patient transfer skill of nursing education students, we developed a robot patient that can simulate three categories of patients: 1) patients whose movements are affected by paralysis; 2) patients whose movements are sensitive to pain with painful expression; and 3) patients whose movements are constrained by medical devices. By practicing with the robot patient, nursing students can learn the skills required for interacting with various patients. To simulate trunk movements of these different patients, novel waist and hip joints with hardware-inherent compliance and force-sensing capability were proposed. In addition, control methods were developed and the parameters were tuned based on actual patient videos. To evaluate the developed robot, nursing teachers performed trials of transferring the robot patient as they would transfer an actual patient. The nursing teachers scored the robot patients based on a checklist. Moreover, subjective evaluations of a questionnaire were performed by the nursing teachers. The results showed that the nursing teachers performed most of the required skills of the checklist and agreed regarding the learning effectiveness of the robot. They recommended training nursing students using the robot patient in the questionnaire. Finally, hugging speed comparison showed that the nurses slow down the speed when dealing with a robot patient with painful expression.

In the past few decades, simulation training has been used to help nurses improve their patient-transfer skills. However, the effectiveness of such training remains limited because it lacks effective ways of simulating patients’ actions realistically. It is difficult for nurses to use the skills learned from simulation training to transfer an actual patient. Therefore, we developed a robot patient that could simulate the behavior of patients’ limbs for patient-transfer training. This study examined the performance of the robot used in training and evaluated its training effectiveness. Four nursing teachers individually transferred the robot patient and then scored the robot patient's ability to simulate patients’ actions and its suitability for skill training. An experiment using pre-post control group design was carried out to examine the robot patient's training effectiveness compared with the human simulated patient. The participants were 20 nursing students and one nursing teacher who was responsible for scoring the students’ skills in the pre-test and post-test. All of the students were assigned to train with either the proposed robot patient or a healthy person simulating the patient. The results show that all four nursing teachers regarded the robot patient's actions as realistic. In addition, all four teachers agreed that the robot patient was suitable for skill training. The results also show that the proposed robot patient is more challenging than the current method, which employs a healthy person to simulate the patient. Significant skill improvement ( p < 0.01) was observed in the experimental group when transferring the robot patient.

Currently, due to shortages in the nursing faculty and low access to actual patients, it is difficult for students to receive feedback from teachers and practice with actual patients to obtain clinic experience. Thus, both evaluation systems and simulated patients have become urgent requirements. Accordingly, this study proposes a method to evaluate the nurse&rsquo, s transfer skill through observation from the patient. After verifying the proposed method, it will be integrated with a robotic patient as a future work. To verify if such an evaluation is practical, a checklist comprising 16 steps with correct and incorrect methods was proposed by the nursing teachers. Further, the evaluation parameters were determined as translational acceleration, rotational speed, and joint angle of patient. Inertial sensors and motion capture were employed to measure the translational acceleration, rotational speed, and joint angle. An experiment was conducted with two nursing teachers, who were asked to carry out both correct and incorrect methods. According to the results, three parameters reveal the difference for a patient under correct/incorrect methods and can further be used to evaluate the nurse&rsquo, s skill once the thresholds are determined. In addition, the applicability of inertial sensors is confirmed for the use of robot development.

Nowadays with the advancement of technology, many evaluation systems have been developed for skill learning. However, those systems are inevitable to set the sensors on trainees or environment (e.g. camera). Therefore this study aims to propose a method to indirectly evaluate the patient transfer skill through only measuring the patient, instead of the nurse. Based on the observation of the patient’s movements during transfer, joint angle and acceleration of patient are determined as two parameters to evaluate the nurse’s skill. Incorrect ways of transfer skill were added into a checklist, proposed in our pre-work, based on clinical experience of nursing teachers. An experiment was conducted by two nursing teachers, and they were asked to carry out the patient transfer skill through both correct and incorrect ways as proposed in the checklist. The results exhibit those two parameters enable to show the difference while the nurses carried out the incorrect way. The angles of hip and knee joints enable to reveal the difference in the steps related to standing up and sitting down; while the acceleration of patient can exhibit the difference of patient’s dynamic movement. According to such results, the indirect evaluation became practical for the future works.

This research aims to develop a robot to help nursing students learn how to physically transfer paralyzed patients. Our prior robotic prototype had only been designed to imitate a patient with weak lower limbs; it was unable to imitate the imbalance and instability of the trunk. Therefore, we developed and waist joint on the prototype robot and also the control system in order to emulate quadriplegic and hemiplegic patients’ tendency to fall over. The waist was designed with 2 DOF compliant joints. Evaluation of the robot showed that the robot was able to properly imitate the unstable waist movements of paralyzed patients.