Remote Biofeedback Method for Biomedical Data Analysis

In recent years, the introduction of methods supported by technology has positively modified the traditional paradigm of rehabilitation. Interactive systems have been developed to facilitate patient involvement and to help therapist in patient’s management. ReMoVES (REmote MOnitoring Validation Engineering System) platform addresses the problem of continuity of care in a smart and economical way. It can help patients with neurological, post-stroke and orthopedic impairments in recovering physical, psychological and social functions; such system will not only improve the quality of life and accelerate the recovery process for patients, but also aims at rationalizing and help the manpower required monitoring and coaching individual patients at rehabilitation centers. In order to help and support therapist work, the Remote Biofeedback Method is proposed as an instrument to understand how the patient has executed the rehabilitation exercises without seeing him directly. Therefore, the purpose of this method is to demonstrate that through the joint observation of data from simple sensors, it is possible to determine: time and method of execution of the exercises, performance and improvements during the rehabilitation session, pertinence of exercise and plan of care. The system, during the rehabilitation session, automatically transmits patient’s biofeedback through three different channels: movement, physiological signals and a questionnaire. The therapist uses patient’s data to determine whether the plan of care assigned is appropriate for the recovery of lost functionalities. He will then return a remote feedback to the patient who will not see any kind of graphical or verbal output, but you will see lighter rehabilitative session if it was too difficult or more intense if one assigned was too simple. The rehabilitation protocol proposed consists of the performance of different exercises, which begins with a breathing activity, designed to relax the patient before the “effective” rehabilitation session. To make the subject comfortable, and to bring again the heartbeat to a basal value, before the rehabilitation session, the patient, in a sitting position, is leading to breathing with a regular rhythm by following a “breath ball”. From the results obtained in the breathing exercise, it can be concluded that the negative trend of the regression line that approximates the heartbeat signal is an index of relaxation, principal goal for which the exercise was designed. The proposed activities include execution of reaching and grasping, balance and control posture functional exercises, masked through serious games to simulate some of the most common gestures of daily life. In some exercises, a cognitive component will also be involved in achieving the goal required by the activity. For each activity, heart rate, gameplay scores, and different motion parameters were captured and analyzed depending on the type of exercise performed. The heart rate was used as an indicator of motivation and involvement during the execution of several rehabilitative exercises. Others parameters analyzed are the score obtained during the execution of the task, and the time interval between the execution of one exercise and the following one. In addition to the analysis of the individual signals, a preliminary analysis of the correlation between the trend of the heart rate and the performance of the score was also carried out. The results showed that heartbeat in conjunction with score and inter-exercise time could be a high-quality indicator of a patient’s status. The indicators extracted, in fact, in most cases, correspond to the information reported from the therapist who observed the patients during the rehabilitation session. A deep analysis of movement signal was carried on, with the extraction of several indicators for the different body segments involved in rehabilitation, such as the upper limb, the hand, the lower limbs and the posture, included the detection of compensation strategies to reach the targets proposed by the exercise. The results have been extracted by comparing the patient performance to a model extracted by a healthy subjects group. Of particular importance is the spatial map for patients with neglect, an innovative tool that traces the positions where the movement was performed and also provides the therapist with the spatial coordinates where the targets were proposed. Another innovative aspect is the analysis of Center of Pressure (CoP) without the use of a specific footboard, but only through the processing of data from the motion sensor. The results obtained by the application of the Remote Biofeedback Methods to the signals acquired during ReMoVES testing phase show interesting applications of the method to the clinical practice. In fact, the indicators extracted show a realistic correspondence between the disabilities affected the patients and the performance obtained during the execution of the exercises. From the study of the different exercises it can be concluded that the analysis of the signals and the parameters extracted individually, do not provide enough information to outline how the rehabilitation exercise has been executed. By combining the different indicators, it is possible to outline an accurate picture that allows the therapist to make decisions about the assigned plan of care. In conclusion, the Remote Biofeedback Method proposed is now ready to be tested on a wider dataset in order to be consolidated on a larger number of athologies and to associate, if necessary, particular indicators to a particular disease. The future steps will be, a creation of a model starting from patients signals, in order to have a better comparison term, and a testing phase on a larger number of patients, following a clinical protocol, subdividing subject by disease.