Artificial intelligence (AI) may not be ubiquitous but its presence in society is growing rapidly, and it reaches into many aspects of our daily lives. Banks use AI to monitor credit card use to prevent fraud. Motor vehicle companies are using AI to develop autonomous vehicles that can drive in a variety of environments. Facial recognition software relies on AI to provide secure access to our phones and computers. AI filters out spam emails so our inbox doesn't overflow with emails. E-commerce, streaming services, and social media use AI to provide customized recommendations and to increase user engagement. GPS technology uses AI to identify efficient routes for travel. These are just a few of the AI applications that are designed to make our lives easier.
IBM defines AI as the use of computer science, engineering, and data to enable problem-solving.1 There are 2 general types of AI: narrow AI, which is AI trained for a specific task, and general AI, which is theoretical at this time, in which a machine would have the same intelligence as a human. Machine learning (ML) is a subset of AI that uses algorithms to identify patterns in data.2 ML techniques can continuously learn from the data to improve their accuracy.
AI has the potential to revolutionize health care. The large amount of available health care–related data and advances in computing that can analyze and act on these data provide unique possibilities to enhance the care and the health of our patients. In medicine, AI is being used to interpret images in radiology, dermatology, and diabetes care. An AI prediction model can detect the development of breast cancer up to 5 years in advance and is equally accurate for White and Black women.3 AI is used to screen fundus images to detect diabetic retinopathy.4 AI is being used in robotic-assisted minimally invasive surgery, which may allow for quicker recovery with less adverse events.2 AI can process digital images of tissue samples to enable the selection and stratification of treatment in oncology.5
The potential for AI in neurologic physical therapy is also great. Narrow AI is being developed that can be used with robotic-assisted therapy,6,7 and to assess motor function, gait, functional status, upper extremity recovery, and movement.8 Puyuelo-Quintana and colleagues6 developed a knee exoskeleton with AI that uses pressure, movement, and force data from sensors in the exoskeleton to provide smart assistance as needed during gait. Similarly, Lyu and colleagues7 developed an AI human-robot cooperative control robotic knee system that utilizes electromyography signals to aid with movement as needed.
Using data from wearable sensors, AI is able to detect gait events, abnormal gait patterns, and gait asymmetry.9–11 AI is also able to detect abnormal movement patterns during functional movements.12 In this context, AI could be valuable during functional task movement analysis13 and prescription of personalized interventions. AI can also be used to predict outcomes and develop a personalized plan of care for people with neurologic disorders.10,14 When combined with telerehabilitation, virtual reality, and body-worn sensors, AI may be able to provide an effective method to monitor the type and dosage of home exercises.15 A remote therapist could be alerted when a patient is not able to perform their home exercise program or when there is a need to modify exercises.
As neurologic physical therapy researchers and clinicians we need to be aware of these developments and work with engineers, data scientists, and computer scientists to develop AI applications so that they can be used to complement our clinical decision-making skills to deliver patient-centered care that meets the desires and goals of the patients we serve. I am excited to see what the future brings.
1. IBM Cloud Education. Artificial Intelligence (AI). https://www.ibm.com/cloud/learn/what-is-artificial-intelligence
. Accessed October 11, 2022, 2022.
2. Yu KH, Beam AL, Kohane IS. Artificial intelligence in healthcare. Nat Biomed Eng. 2018;2(10):719–731. doi:10.1038/s41551-018-0305-z.
3. Yala A, Lehman C, Schuster T, Portnoi T, Barzilay R. A deep learning mammography-based model for improved breast cancer risk prediction. Radiology. 2019;292(1):60–66. doi:10.1148/radiol.2019182716.
4. Gulshan V, Peng L, Coram M, et al. Development and validation of a deep learning algorithm for detection of diabetic retinopathy in retinal fundus photographs. JAMA. 2016;316(22):2402–2410. doi:10.1001/jama.2016.17216.
5. Bera K, Schalper KA, Rimm DL, Velcheti V, Madabhushi A. Artificial intelligence in digital pathology—new tools for diagnosis and precision oncology. Nat Rev Clin Oncol. 2019;16(11):703–715. doi:10.1038/s41571-019-0252-y.
6. Puyuelo-Quintana G, Cano-de-la-Cuerda R, Plaza-Flores A, et al. A new lower limb portable exoskeleton for gait assistance in neurological patients: a proof of concept study. J Neuroeng Rehabil. 2020;17(1):60. doi:10.1186/s12984-020-00690-6.
7. Lyu M, Chen WH, Ding X, Wang J. Knee exoskeleton enhanced with artificial intelligence to provide assistance-as-needed. Rev Sci Instrum. 2019;90(9):094101. doi:10.1063/1.5091660.
8. Luvizutto GJ, Silva GF, Nascimento MR, et al. Use of artificial intelligence as an instrument of evaluation after stroke: a scoping review based on international classification of functioning, disability and health concept. Top Stroke Rehabil. 2022;29(5):331–346. doi:10.1080/10749357.2021.1926149.
9. Lopez-Meyer P, Fulk G, Sazonov E. Automatic detection of temporal gait parameters in post-stroke individuals. IEEE Trans Inf Technol Biomed. 2011;15(4):594–601.
10. Moon S, Ahmadnezhad P, Song HJ, et al. Artificial neural networks in neurorehabilitation: a scoping review. NeuroRehabilitation. 2020;46(3):259–269. doi:10.3233/NRE-192996.
11. Khera P, Kumar N. Role of machine learning in gait analysis: a review. J Med Eng Technol. 2020;44(8):441–467. doi:10.1080/03091902.2020.1822940.
12. Kianifar R, Lee A, Raina S, Kulic D. Automated assessment of dynamic knee valgus and risk of knee injury during the single leg squat. IEEE J Transl Eng Health Med. 2017;5:2100213. doi:10.1109/JTEHM.2017.2736559.
13. Quinn L, Riley N, Tyrell CM, et al. A Framework for movement analysis of tasks: recommendations from the Academy of Neurologic Physical Therapy's Movement System Task Force. Phys Ther. 2021;101(9):pzab154. doi:10.1093/ptj/pzab154.
14. Khan O, Badhiwala JH, Grasso G, Fehlings MG. Use of machine learning and artificial intelligence to drive personalized medicine approaches for spine care. World Neurosurg. 2020;140:512–518. doi:10.1016/j.wneu.2020.04.022.
15. Chae SH, Kim Y, Lee KS, Park HS. Development and clinical evaluation of a web-based upper limb home rehabilitation system using a smartwatch and machine learning model for chronic stroke survivors: prospective comparative study. JMIR Mhealth Uhealth. 2020;8(7):e17216. doi:10.2196/17216.