A total of 28 of the included studies [16,17,28-53] explored the integration of cloud computing technologies in the context of rehabilitation and rehabilomics. Cloud platforms are used for data storage, analysis, and sharing to facilitate collaborative research efforts.

Out of 28 studies, 16 (57%) studies [16,17,28,29,33-35,38,39,42,46,48,49,52,53] focused on the concept of personalized rehabilitation by leveraging omics data to tailor rehabilitation programs for individual patients. This approach aims to improve treatment efficacy and patient outcomes.

Out of 28 studies, 12 (43%) studies [16,17,28,34-36,41,44,47,49,52,53] addressed the critical issues of data security and patient privacy in cloud-based rehabilitation. Ensuring the confidentiality and integrity of patient data has become a key concern.

Collaboration between researchers from diverse backgrounds, including rehabilitation specialists, geneticists, and data scientists, was a recurring theme [16,17,28-53]. Such interdisciplinary collaboration is essential for advancing this field.

This refers to a cloud-based framework for rehabilitation management [28].

The authors introduced “CRehab,” a cloud-based framework featuring web-based and augmented reality support aimed at managing the rehabilitation processes. This framework targets the administrative, communication, and patient-motivation aspects by leveraging cloud capabilities. The primary objective was to devise an architecture adaptable to diverse rehabilitation types, irrespective of the diagnosis, environment, or patient demographics.

This refers to a cloud-based motor rehabilitation digital assistant [29]. The integration of body sensor networks with cloud computing for motor rehabilitation enables remote monitoring, reduces costs, and provides comfort and safety to patients. Rehab-aaService, the newly introduced motor rehabilitation digital assistant, is based on wearable motion sensor nodes, a personal mobile device, and a cloud-based backend, offering enhanced support and a more comprehensive rehabilitation experience.

The authors designed an open-source eHealth monitoring and fall detection system that integrated medical sensors, a microcontroller, and cloud communication [35]. This system stores data in the cloud to aid diagnosis, real-time monitoring, and rehabilitation facilitation. In addition, it features a fall detection system that promptly alerts medical personnel in the event of unexpected falls.

A low-cost rehabilitation gaming system using Microsoft Kinect was developed for individuals with chronic arm pain [30]. This system encompasses goal-oriented exercises, enabling accurate performance measures for quantitative evaluation of the treatment plan. It notably boosts user motivation, hastens muscle recovery, and provides therapists with a means to assess patient progress.

The authors designed a cloud-based VR platform for neurorehabilitation to analyze its impact on the sense of agency and ownership in healthy individuals [40]. The experimental platform featured a virtual avatar that replicated the individual’s movements using motion capture technology and a head-mounted display. The findings revealed that perceived arm length changed according to the displayed arm length of the avatar, demonstrating the considerable potential of the VR system.

The authors introduced a cloud-based VR platform to explore human-robot interaction and obtain social intelligence for robots [50]. This successfully showcased the system’s practicality in real human-robot interaction scenarios, indicating its potential for fostering social intelligence development.

This study referred to technology-enhanced rehabilitative care for patients with stroke through the use of edge cloud and IoT technology [53]. This involved the development of an edge-cloud model to track the vital signs and rehabilitation progress of patients in real time. The system exhibited remarkable operational efficiency and significantly improved rehabilitation treatment outcomes, paving the way for more advanced and intelligent medical service systems.

A cloud-based rehabilitation system focused on improving motor function in patients with stroke was developed. This study used data from healthy participants to accurately assess hand function [33]. Using the autoregressive integrated moving average model with the dynamic time-warping algorithm, this study made predictions regarding the rehabilitation and recovery of patients with stroke. Over a 10-week period involving 3 patients, the system showed significant improvements in upper limb control.

The Limbs Alive project was introduced at the Newcastle University and is dedicated to using video games for therapeutic movements to assist patients with stroke [34]. The authors developed a cloud-based therapy platform that gathers data from therapeutic games played by patients with stroke, estimates performance metrics, and delivers results to patients and clinicians via web applications.

A robotic part-task trainer was developed to evaluate upper limb spasticity using cloud-stored clinical data [39]. It enables the remote coaching of trainees and continuous updates by rehabilitation physicians, thereby fostering a dynamic learning environment.

This study referred to focusing on aiding facial paralysis rehabilitation via cloud-based information and communication technologies [46]. The proposed system aims to support patients and physicians by delivering rehabilitation training, automating progress reviews, and evaluating the results. It involves training a client for patient rehabilitation and data collection, leveraging machine learning on a cloud platform to analyze outcomes.

The authors developed a conceptualized cloud-based telerehabilitation framework using cost-effective, noninvasive Microsoft Kinect for rehabilitation exercises [31]. This framework enables patients to engage in rehabilitation exercises at home while maintaining high-quality care standards. The initial experiments validated the potential of the system for training healthy participants to mimic physical rehabilitation exercise motions.

The authors designed “PD Dr,” a mobile health app aimed at gathering PD-related motion data and evaluating symptoms [16]. The system was examined using information from 40 patients with PD to validate its efficiency in capturing significant symptoms and estimate their severity. It demonstrated robust connections between the collected motion features and the severity of PD, hand-resting tremors, and gait difficulty.

The authors suggested a cloud-based framework designed to detect and monitor patients with PD, particularly in low-resource settings [17]. The goal was to provide remote diagnosis for patients in areas with limited health care access, achieving high accuracy in the process.

These studies showcase a broad spectrum of cloud-based rehabilitation applications, ranging from pediatric developmental language disorder rehabilitation to stroke rehabilitation assessments using wearable technology [45]. These efforts aim to enhance rehabilitation outcomes through innovative technological approaches and data-driven analyses by incorporating cloud computation and sensor-based data for accurate assessment.

This study was centered on telerehabilitation as a service by leveraging cloud computing, IoT, and big data analytics [49]. It emphasizes the management of significant health care data generated from remote rehabilitation devices within the hospital cloud using NoSQL databases.

The authors proposed a system that used wearable wireless sensors and handheld devices to track body motion, focusing on patients with stroke [36]. This setup allowed patients to conduct rehabilitation exercises at home while health care providers remotely monitored their progress. This system integrates cloud services for remote monitoring, eliminating the need for expensive hardware and software licenses for data storage. Initial user tests demonstrate the feasibility of the system.

The authors described an IoT-based wearable system designed for monitoring physical rehabilitation using accelerometer and gyroscope sensors [37]. This system can record movement data to characterize patient movement and estimate their condition, thereby enabling remote monitoring through a cloud-based service with accurate capabilities for monitoring elbow rehabilitation characteristics.

The authors developed a cloud-based platform that facilitates independent at-home rehabilitation through customizable exercise games [43]. This system uses a flexible database and structured rehabilitation data to allow clinicians to review and fine-tune exercise parameters efficiently. The platform incorporates parameterized exercise games designed to offer personalized goal-oriented rehabilitation.

The authors developed “CoSHE,” a cloud-based smart home environment for monitoring the health of older populations within their homes [44]. The system incorporates wearable sensors and environmental data to provide real-time information to remote caregivers. The integration of contextual information successfully enhances understanding of an individual’s health status.

A cloud-based physical therapy monitoring and guidance system was developed to address the limitations of traditional therapy owing to the restricted interaction time [48]. The system leverages the actions of a physical therapist recorded as an avatar to guide patients in real time for self-training. By aligning the physical therapy and user motion sequences with a gesture-based dynamic time-warping algorithm, it evaluates user performance based on the physical therapy criteria and provides real-time guidance, optimizing patient learning.

A prototype speech-controlled cloud-based wheelchair platform using WebKit Speech Application Programming Interface in the cloud was developed, which introduced a graphical user interface compatible with the web and mobile devices that enabled live video streaming [32]. It achieved speech recognition accuracy between 60% and 97% and recorded latencies spanning from 0.4 to 1.3 seconds. Through clinical testing, it exhibited promising outcomes, signaling potential for future implementation.

The authors designed a cloud-based architecture specifically tailored for aphasia rehabilitation in Romanian-speaking patients [38]. The system comprises an application database, logic, and interfaces for patients and therapists. Using a virtual assistant, the system guided patients through exercises crafted by therapists. Treatment adjustments were made based on the obtained scores and the data were stored in a statistical module for further assessment.

The authors investigated the performance and stability of cloud-based somatosensory games in rehabilitation systems, while handling various user loads [41]. In this study, Python and Java programming languages were compared, and the findings indicated that Python consumed fewer resources in both the simulation and monitoring systems.

The authors suggested a cloud-based rehabilitation framework that combined augmented reality and sensor gloves for patients after an episode of stroke [42]. Notably, considerable enhancements in finger strength were observed after 6 weeks of therapeutic exercise. Sensor gloves were used to recognize the real-time conditions of the patients engaging in rehabilitative exercises.

A unified cloud-based resource was developed to manage psychometric tests used for attitude evaluation, personal selection, educational purposes, rehabilitation, and diagnosis of cognitive disorders [47]. This resource aims to simplify the process of validation, standardization, and reorganization of tests traditionally conducted using complex statistical methods.

This study referred to the application of cloud-based rehabilitation for children with developmental language disorders during the COVID-19 pandemic [51]. The authors investigated the effectiveness of the JingYun Rehab Cloud Platform on language and cognitive outcomes in 162 children with developmental language disorders through a prospective cohort study. Patients who underwent remote cloud-based rehabilitation were compared with a control group that received home-based rehabilitation. The results indicated significantly better language abilities among the children in the cloud-based rehabilitation group.

The authors proposed a deep learning-assisted system incorporating IoT sensors in a body area network to diagnose sports injuries using cloud computing resources. IoT sensors gather data for diagnosis of injury and cloud computing to provide flexible computing resources [52]. They developed a brain injury monitoring framework and optimized neural networks to forecast sports injuries and evaluated the model performance using accuracy, precision, recall, and F₁-score metrics.

Our study established a strategy for personalized rehabilomics tailored toward individual needs by integrating clinical information from hospitals and extending it further to genomics, epigenomics, proteomics, and metabolomics for intervention, measurement, assessment, and prognosis (Figure 3). We emphasized the active use of clouds and platforms. In addition, we derived a strategic direction emphasizing the integration of hospitals and homes, promoting active community-based rehabilitation and linking it with individualized care providers. The first aspect involves a cloud that encompasses rehabilomics big data along with genomics, epigenomics, proteomics, metabolomics, and personalized medicine. The second aspect includes game-based platforms; robot–human-based platforms; and the use of VR, augmented reality, mixed reality, and extended reality. The third aspect involves hospitals and rehabilitation centers conducting diagnosis, measurement, evaluation, and prognosis, and using and linking such rehabilitation data. The fourth aspect is telehealth, which actively uses IoT-based systems, wearable sensors, and monitoring and communication. The fifth aspect involves home and community, using individual rehabilitation, and actively using information and rehabilitation between caregivers and care receivers. The sixth aspect focuses on disease and disability, actively using rehabilitation and data from children to the older adults, and targeting neurological and musculoskeletal disorders. All these processes, interconnected and used, ultimately lead to the development of a strategy for individualized rehabilitation.