Читать книгу Intelligent Systems for Rehabilitation Engineering - Группа авторов - Страница 18

A 2D vision-based localization system was illustrated, which could identify the light-emitting markers. It was equipped with a web camera and human–machine interaction interface [49]. An attempt was made for bridging the gap between assistive robotic systems and smart homes. Robotic system assisted the patient and smart home adjusted as per the requirements of the patient [50]. An algorithm was proposed for the therapy of impaired patients, which adopted with the patients as they recovered [51]. The feed-forward neural network (FFNN) was used for the determination of EMG signals from eight shoulder muscles [52]. A diagram representing the architecture of the feed-forward neural network (FFNN) to obtain the EMG signal from the different shoulder muscles is shown in Figure 1.1.

Figure 1.1 The architecture of the feed-forward neural network (FFNN) to obtain EMG signals from shoulder muscles.

Table 1.2 Summary of certain studies which explored smart robotics for rehabilitation.

Ref. number	Area of rehabilitation robotics explored	Remarks
[49]	2D vision-based localization system	The illustrated system could identify the light-emitting markers. The system was equipped with a web camera along with a human–machine interaction interface.
[50]	Assistive rehabilitation robotics and smart homes	The proposed methodology bridged the gap between these two aspects. The robotic systems extended the movement of patients and smart homes accessed their requirements and made necessary changes in itself.
[51]	A novel algorithm for impaired patients’ therapy	The proposed algorithm exploited the similarities between motor recovery and motor learning, which adopted with the patients as they recovered.
[52]	Feed-forward neural network (FFNN)	FFNN was used for predicting the EMG signals from eight shoulder muscles of patients.
[53]	VR strategies	VR was integrated with multimodal displays, which enhanced the performance and also provided feedback information to the patient and motivated the patients using additional audiovisual features.
[54]	Neural networkingbased facial emotion interpreter	Thermal images of persons who suffered from speech disorder were prepared, and then using a confusion matrix, its performance was evaluated.
[55]	Decision-making ability	Task-oriented robots were studied and were tested on BAXTER to check whether it was able to assist the person for training or not.

VR technology was integrated with multimodal displays, which provided feedback information to the patient and motivated him [53]. The facial emotions were determined using the thermal images for speech disorder patients using a confusion matrix [54]. The decision-making ability of task-oriented rehabilitation robot was tested on BAXTER robot, and its feasibility was determined [55]. A synopsis of the above discussion on the smart robotics being employed for rehabilitation purposes is deliberated in Table 1.2.

Table 1.2 summarizes some of the work that presented smart robotics for rehabilitation purposes. The work done in VR and NN technologies will instigate future research in this field.