Artificial Intelligence in Physical Therapy and Rehabilitation

<jats:p>Artificial intelligence is used for various purposes in the field of physical therapy and rehabilitation, such as assessment, treatment planning, motion analysis, patient education, and remote monitoring. This section discusses research findings on AI-supported applications to evaluate the potential of these technologies in rehabilitation. Camera- and sensor-based systems can contribute to motion recognition, the measurement of joint movements, and the assessment of exercise performance. Meanwhile, virtual and augmented reality applications can make treatment more interactive and provide personalized feedback. Robotic rehabilitation systems can support the performance of repetitive movements and the regular monitoring of patient performance. Systems that monitor at-home exercises can track the patient’s participation in the exercises and whether they are performing the movements correctly. Virtual physical therapist applications can also be useful for demonstrating exercises, guiding the patient, and providing real-time feedback. However, there is insufficient evidence to support these systems making clinical decisions independently. Digital twins aim to integrate a patient’s clinical and functional data into a digital model. This approach can contribute to monitoring the course of the disease and evaluating the potential outcomes of different treatment options. On the other hand, large language models and agent-based artificial intelligence systems can be used in the areas of patient education, summarization of clinical records, access to information, and decision support. Although current studies show promising results, there are significant differences in terms of sample sizes, follow-up periods, and the methods used. Data security, algorithmic bias, explainability, ethical responsibility, and clinical oversight must also be taken into account. In conclusion, artificial intelligence should be viewed not as a replacement for physical therapists, but as a supportive tool that aids in assessment and treatment processes. There is a need for studies with larger sample sizes, longer follow-up periods, and conducted under real-world conditions.</jats:p>