|Year : 2022 | Volume
| Issue : 1 | Page : 8-13
A shifting paradigm from human intelligence to artificial intelligence in rehabilitation: A descriptive review
Renu B Pattanshetty, Saira Khan
Department of Oncology Physiotherapy, KAHER Institute of Physiotherapy, Belagavi, Karnataka, India
|Date of Submission||29-Jun-2021|
|Date of Decision||07-May-2022|
|Date of Acceptance||07-Jun-2022|
|Date of Web Publication||30-Jul-2022|
Dr. Renu B Pattanshetty
Department of Oncology Physiotherapy, KAHER Institute of Physiotherapy, Belagavi, Karnataka
Source of Support: None, Conflict of Interest: None
Artificial intelligence (AI) is a collection of intelligent processes and behaviors developed by computer models and technology. Health organizations have initiated the partnership with various technology companies to enhance the usage of AI in the health-care sector. It is now widely used in various medical applications such as the disease treatment, screening, diagnosis, prognosis, and rehabilitation. Human intelligence in conjunction with intelligent algorithms has shown to be helpful in better clinical practice. The technology and methods described in this study are some of the features that open up fields for improving rehabilitation services and research in the health-care system. Newer technologies are always trying to connect the human brain and computer, so the “AI brain” may contribute to improve health-care services in a quality manner including to all rehabilitative professionals, caregivers, and clients.
Keywords: Artificial intelligence, Intelligent prosthesis, Rehabilitation, Virtual game
|How to cite this article:|
Pattanshetty RB, Khan S. A shifting paradigm from human intelligence to artificial intelligence in rehabilitation: A descriptive review. Indian J Phys Ther Res 2022;4:8-13
|How to cite this URL:|
Pattanshetty RB, Khan S. A shifting paradigm from human intelligence to artificial intelligence in rehabilitation: A descriptive review. Indian J Phys Ther Res [serial online] 2022 [cited 2022 Nov 27];4:8-13. Available from: https://www.ijptr.org/text.asp?2022/4/1/8/353017
| Introduction|| |
Intelligence is the ability to acquire and apply knowledge and skills. Artificial intelligence (AI) is pointless without human intelligence. The human brain has the greatest ability to relentlessly keep trying to design technologies that could mimic the human brain with all its functions to its optimum. AI is a dynamic part of computer science that aims to create systems that replicate human intelligence and is useful in a wide range of human activities, including medicine. Hence, it is no surprise that technology plays a key role in improving medical and health outcomes. In a viable manner, AI is said to be a computer system that demonstrates specific aspects of the human brain with all cognitive behaviors such as learning, reasoning, and problem-solving. Thus, AI is a collection of intelligent processes and behaviors developed by computers. In humans, intelligent agents such as the upper and lower limbs including the mouth and other body parts that serve as effectors, following instructions. In robotics or camera, robotic agents and infrared range searchers serve as effectors. Enhanced optimizing computers usage and its accessibility to enormous data of a wide range of advanced computations models along with algorithms including machine learning, language processing, and voice technologies have led to provide near precision data with better clinical evidence for various medical diagnoses, treatment options, and decision including medical research and wide spectrum of the entity of health-care developing system.
| History of Artificial Intelligence|| |
The first attempts at AI and its use in medicine were made in the 1950s, with an emphasis on diagnosis and therapy. In 1956, Marvin Minsky and his team hosted an approximately 8-week-long Defence Advanced Research Projects Agency at Dartmouth College in New Hampshire, and the word “AI” was formally coined. Stanford's Ted Shortliffe and his innovative software MYCIN were among the most well-known early studies on AI in medicine. MYCIN was used in the prescription for antibiotics and its selection in various infectious disorders. Stanford, MIT, Rutgers, and Pittsburgh in the United States and a few facilities in Europe were among the early academic centers working on AI in medicine. In 2005, Szolovits conducted one of the first planned educational endeavors on the expanding field of medical AI at MIT, USA. In addition to standard expert systems, fuzzy logic and neural networks were popular AI approaches during this time, with the latter being employed in a number of clinical contexts such as clinical diagnosis and treatment planning.
| Application of Artificial Intelligence in Health Care|| |
With a wide range of AI tools and platforms available in recent times, numerous health organizations have initiated collaboration with technology companies to enhance the usage of AI in the health-care entity.
| Artificial Intelligence and its Applications Including Rehabilitation|| |
Artificial intelligence in various diseases
AI is widely used in various medical applications including the disease management and screening, diagnosis, prognosis estimation, and therapy. It has recently gained a lot of attention due to its benefits in health and chronic illness management including cancer, diabetes mellitus, coronary artery diseases, stroke, and other neurodegenerative diseases. One of the important applications of AI includes patient monitoring, guidance, and status assessment. This has also led to AI applications in areas including mobile computing with the goal of creativity and providing improvement in disease management services. It has also known to have a wide application of using AI in chronic respiratory disease and lifestyle management. ResApp has shown to use mobile phones to monitor patient breathing and provide an assessment for a variety of lung disorders and other chronic respiratory diseases. AI system uses the integration of geographical, clinical data that is combined with sensor-based computational technology in the form of various wearable devices to manage and control various chronic conditions.
AI has shown to be useful in monitoring oxygen saturation levels in critical care patients and enhancing optimal hospital services. It also has a positive effect on the treatment adherence rate in stroke patients who are on anticoagulant therapy. AI has also been found to have a wide range of use in cancer research including detection, invasive to deeper tissues, prognosis, and therapeutic guidance. Breast cancer, head-and-neck cancers, and liver cancers are some of the cancers that have found the wide application of AI.
Ambient intelligence (AMI) is a type of AI that includes surroundings that are aware and respond to human presence. This system are integrated in a patient's home and the priorities of patients. This analysis is the real time of the individual in terms of his/her geographical location within the given environment including the time, his/her physical environment including home/office or rehabilitation setting capturing the individuals' needs and function [Figure 1].
|Figure 1: Interconnected world of AMI health services. AMI: Ambient intelligence|
Click here to view
Speech recognition, natural language processing, eye and facial tracking, and gesture recognition are all examples of naturalistic ways for operators to communicate with intelligent equipment. People with physical or visual impairments can use verbal interaction to communicate with and operate gadgets. Users of respirators can use eye trackers to execute the same. It has shown to be of great use in critical health conditions such as performing resuscitation or fall detection.
Applications of artificial intelligence in wearable gadgets and mobiles
Wearable devices such as smartwatches and intelligent mobiles have become ideal footboard to collect health-related data and contextual data useful for rehabilitation purposes. This has the advantages of being in physical contact with wearers for a longer duration who can be easily engaged using a touch sensor. Wearable gadgets are known to possess fall detection systems with sophisticated sensors that are designed to detect falls and alert users, medical personnel, and caregivers. Health-related information point-of-use tools, clinical decision support tools, medical imaging, and mood and behavior trackers such as pain or incontinence dysfunctions find an increased range of accessibility in rehabilitation services using AI in a wide range of patient population [Figure 2].
|Figure 2: Portable medical and health-care devices are worn on the body parts|
Click here to view
Intelligent virtual agents
These are animated computers with virtual characters that looks like a human or in any form. This is supposed to contain software that enables the virtual imaging to communicate with the individual through text or chat that could also interpret, reason, and occasionally express emotions too. Changes in physical appearance, mannerisms, accents, and traits as per the cultural norms could all aid the system in improving the rapport with patients in better rehabilitation outcomes.
Application of virtual and augmented reality games for artificial intelligence rehabilitation
People can use virtual reality (VR) to immerse themselves and can interact with a three-dimensional simulated environment. It may also be used to build a virtual human being or virtual living forms with which an individual can communicate within the virtual world. VR allows for the creation and management of environmental stimuli in therapeutic applications, allowing for the practice and recording of behavioral responses for clinical evaluation and treatment goals.
In augmented reality, digitally generated VR can be placed on a live “actual world” video picture. This makes information about the user's immediate environment available for involvement and digital editing, enhancing the participants' way of thinking. VR in augmented reality approaches has been implemented to improve motor skills, pain management, and stroke rehabilitation. In traumatic brain injury (TBI) patients, the implication of VR was found to be effective in improving memory training. VR has shown to be a beneficial tool for studying, assessing, and rehabilitating cognitive processes and functional capacities. Individuals with neurocognitive diseases and behavioral difficulties as the consequences of TBI may benefit from these applications.
Virtual games have demonstrated to be beneficial in the management of pain, improvement of cognition and behavioral abilities, rehabilitation services, and the physical training of patients in health-care development. Game-based virtual therapy environments and processes have the ability to be personalized as per the patients' needs. Levels of play, reward structures, NPCs, and the general game storyline are adjusted to offer the participants with the best learning and training experience possible. The VR was found to be effective in improving quality of life and also enhancing the proprioception in the knee joint after total knee replacement surgery.
The Rheo and the power knees were designed by Ossur in 2006, which has a built-in AI system. In 2011, he unveiled the first mechanical leg with automation, whereas Ottoock unveiled the Genium X3 which allowed reverse Walking and provide easy and accessible mobility throughout the gait training. This intelligent prosthesis has shown to have seven sensors and four central processing unit (CPU), distributed throughout the leg's body which senses and evaluates data on the user's movement/activities and surroundings to enable the natural synchronized movement of the lower limb. In 2009, a brain-computer interface-based hand orthosis was created that employed a cursor control interface and a basic linear discriminant analysis classified to divide electroencephalogram signals into three states: right, left, and nil, with related commands of various hand movements.
Applications of artificial intelligence in wheelchair
The artificial neural network is being used in machine learning and in intelligence technology in advanced wheelchair to improve patients' well-being. The advanced wheelchair model is controlled by hand signal which uses the recurrent neural network in the remote control. In stroke patients after discharge, at home, the AI-intend wheelchair was very helpful for the patient as it allows them to do activities independently without burdening the caregiver and also it enhances the quality of life.
Traditional surgery is an invasive procedure and has increased postoperative complications. AI-assisted surgery can cut down on hospital stay and reduces postsurgery infections. The Neuro surgeons selects a goals and direction on a preoperative medical image with the help of medical robots, and the robot guides the instrument into position with precision, as well as guiding needles for biopsy and guiding drills to form burr holes. In an orthopaedic surgeries, robotic systems help in bone resection which has a positive effect on maintaining the proper alignment of the implant with bone and improving the contact area between implant and bone and enhancing the functional outcomes of a patient. In physical rehabilitation training, robotics are used to assist stroke patients, senior care, and transfers of medical essential aids and instruments.
Physically assistive robotics and robotic exoskeletons
Robots have been utilized widely in physical rehabilitation applications across a wide range of clinical fields, including assisting patients in restoring motion after a stroke, enhancing or replacing lost function, and assisting mobility. Researchers and rehabilitation professionals have paid increased attention to rehabilitation robots over the past decade. The use of a rehabilitation robot can relieve clinicians of time-consuming training activities, analyze data collected by the robot throughout the training process, and assess the patient's rehabilitation progress. The use of robotic-assisted machines like continuous passive motion (CPM) was found to be effective in treating shoulder pain in stroke patients. Robotic exoskeletons aid in the functioning of the upper extremities. Four control modes are commonly considered in the rehabilitation of motor or neuromotor function. First, the robotic exoskeleton provides all essential movement to the extremity in the active control mode later in the passive mode it will allow the movement. The assistive mode (15%) partially assists the limb's movement, whereas the resistive mode provides opposition to the limb's movement. This robotic exoskeleton are helpful in treating stroke patients by enhancing neuronal regeneration. All of these monitoring types is integrated into the therapeutic programs and deliver a variety of clinically or verifiable results. Rehabilitative robots are effective in observing the patient at home when physical therapist is unable to assist them physically approach them. It is also helpful in relieving the therapist's burden when giving the exercises to a group of people[Figure 3].
Treadmill-based exoskeleton robots
This type of rehabilitative robotics connects the human body in a wearable method which controls the body movements during rehabilitation. Like Lokomat exoskeletal robotics used for lower limb training are basically used for gait training that helps individuals to complete the walking activity [Figure 4]. The primary goal of rehabilitation training is to restore patients' lower limb motor functions. As a result, a normal gait pattern is required as a reference input to the control system, a training goal, and a rehabilitation evaluation standard during rehabilitation training. Other treadmill robotics have a weight harness with a lower limb exoskeleton which is useful to mobilize the elderly and patients with lower limb dysfunction. Patients' lower limbs are dragged for CPM for passive training in the early phases of rehabilitation, which can effectively keep joints flexible for a long period. Position control, on the other hand, ensures that the robot can precisely follow the required position. These rehabilitation robotics help to decreased the therapist's workload in terms of energy efficiency during training, evaluation, and assessing recovery.
|Figure 4: The Lokomat (left) and gait trainer (right) leg rehabilitation robots|
Click here to view
| Conclusion|| |
All individuals including physical therapists within various health-care industries and health organizations should be ready to collaborate and work together to effectively utilize AI to make it a success in patient care with smarter decision-making skills. More rehabilitation conferences should be held to improve communication and enhance learning among researchers and academics about the AI to understand its impact and utilization in patients' care. Robotic values will govern the standard operating procedure of robots in therapeutic settings, rather than relying on the personal knowledge of physical therapists. Human intelligence in collaboration with an intelligent algorithm can help in better clinical practice and patient care. This also requires humans to evaluate machine recommendations, dispense authority to them, and control them. Therefore, learning to implement AI in rehabilitation is very crucial for quality services since the use of AI-based technology is likely to increase in the near future.
In an intelligence era, the human connection will be the absolute key to success, and we must accept every chance to improve our ability to care for one another, learn successfully over the course of our lives, and develop creative solutions to the challenges that matter to us. After all, “there is no limit of human intelligence, imagination, and wonder.”
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Shu LQ, Sun YK, Tan LH, Shu Q, Chang AC. Application of artificial intelligence in pediatrics: Past, present and future. World J Pediatr 2019;15:105-8.
Nayak S, Das RK. Application of artificial intelligence (AI) in prosthetic and orthotic rehabilitation. Serv Robot 2020;11:10-6.
Haenlein M, Kaplan A. A brief history of artificial intelligence: On the past, present and future of artificial intelligence. Calif Manage Rev 2019;61:5-14.
Zhang Y, Yu H, Dong R, Ji X, Li F. Application Prospect of Artificial Intelligence in Rehabilitation and Management of Myasthenia Gravis. BioMed Research International. 2021;2021:1-6.
Ghazal S, Sauthier M, Brossier D, Bouachir W, Jouvet PA, Noumeir R. Using machine learning models to predict oxygen saturation following ventilator support adjustment in critically ill children: A single center pilot study. PLoS One 2019;14:e0198921.
Acampora G, Cook DJ, Rashidi P, Vasilakos AV. A survey on ambient intelligence in health care. Proc IEEE Inst Electr Electron Eng 2013;101:2470-94.
Labovitz DL, Shafner L, Reyes Gil M, Virmani D, Hanina A. Using artificial intelligence to reduce the risk of nonadherence in patients on anticoagulation therapy. Stroke 2017;48:1416-9.
Aarts E, De Ruyter B. New research perspectives on ambient intelligence. J Ambient Intell Smart Environ 2009;1:5-14.
Hudlicka E. Virtual affective agents and therapeutic games. In: Artificial intelligence in behavioral and mental health care. 1st
ed., Ch. 4. Academic Press; 2016. p. 81-115.
Frank RG, Elliott TR. Virtual Reality and Rehabilitation. In: Handbook of rehabilitation psychology. 3rd
ed., Ch. 31. Washington, DC: American Psychological Association; 2000. p. 521-540.
Harrer S, Shah P, Antony B, Hu J. Artificial intelligence for clinical trial design. Trends Pharmacol Sci 2019;40:577-91.
Man DW, Poon WS, Lam C. The effectiveness of artificial intelligent 3-D virtual reality vocational problem-solving training in enhancing employment opportunities for people with traumatic brain injury. Brain Inj 2013;27:1016-25.
Rizzo A, Buckwalter J, Neumann U. Virtual reality and cognitive rehabilitation: A brief review of the future. J Head Trauma Rehabil 1997;12:1-15.
Hugues O, Fuchs P, Nannipieri O. New augmented reality taxonomy: Technologies and features of augmented environment. In: Handbook of Augmented Reality. New York, NY: Springer; 2011. p. 47-63.
Gianola S, Stucovitz E, Castellini G, Mascali M, Vanni F, Tramacere I, et al
. Effects of early virtual reality-based rehabilitation in patients with total knee arthroplasty: A randomized controlled trial. Medicine (Baltimore) 2020;99:e19136.
Chen CW, Lin CC, Ju MS. Hand orthosis controlled using brain-computer interface. J Med Biol Eng 2009;29:234-341.
Frank RG, Elliott TR. Artificial Intelligence and Robotics in Rehabilitation. In: Handbook of rehabilitation psychology. 3rd
ed., Ch. 30. Washington, DC: American Psychological Association; 2000. p. 507-520.
Rabhi Y, Mrabet M, Fnaiech F. Intelligent Control Wheelchair Using a New Visual Joystick. Journal of Healthcare Engineering. 2018:1-20.
Beasley RA. Medical robots: current systems and research directions. Journal of Robotics. 2012;20:1-14.
Hillman M.2 rehabilitation robotics from past to present–a historical perspective. In :Advances in Rehabilitation Robotics.1st
ed., Ch. 2. Berlin, Heidelberg. Springer, 2004. P. 25-44.
Backus D, Winchester P, Tefertiller C. Translating research into clinical practice: Integrating robotics into neurorehabilitation for stroke survivors. Top Stroke Rehabil 2010;17:362-70.
Kim MS, Kim SH, Noh SE, Bang HJ, Lee KM. Robotic-assisted shoulder rehabilitation therapy effectively improved poststroke hemiplegic shoulder pain: A randomized controlled trial. Arch Phys Med Rehabil 2019;100:1015-22.
Vélez-Guerrero MA, Callejas-Cuervo M, Mazzoleni S. Artificial intelligence-based wearable robotic exoskeletons for upper limb rehabilitation: A review. Sensors (Basel) 2021;21:2146.
Vourganas I, Stankovic V, Stankovic L. Individualised responsible artificial intelligence for home-based rehabilitation. Sensors (Basel) 2020;21:2-6.
[Figure 1], [Figure 2], [Figure 3], [Figure 4]