Author(s): Andrés Gronda, MD, Cherie Chen, DO, Ariana Schanzer, MD, Emily Sears, DO, Salvador Portugal, DO, MBA
Originally published: September 2, 2026 Last updated: September 2, 2026
Jump to:

Overview and Description

What is AI?

Artificial intelligence (AI) is a broad field of technology that simulates human intelligence to help machines perform complex tasks. Foundational technologies behind AI tools include machine learning (ML), artificial neural networks, and natural language processing (NLP).

ML builds algorithms that “learn” from data, identifying patterns to predict complex relationships over time. A powerful form of ML is deep learning, which uses artificial neural networks modeled after human neurons to handle highly complex tasks like motion tracking and medical imaging analysis. Natural language processing enables computers to understand and interpret human language; comprehending context, reasoning through problems, and generating human like text.

AI and PM&R

AI is increasingly influencing the field of physical medicine and rehabilitation (PM&R), introducing new possibilities in diagnosis, clinical decision-making, and personalized treatment planning. By combining machine learning, data analytics, and intelligent systems, AI is reshaping how rehabilitation professionals assess, monitor, and enhance patient recovery across diverse conditions and settings.

Neurorehabilitation increasingly integrates artificial intelligence to enhance diagnosis, prognosis, and treatment across traumatic brain injury, spinal cord injury, movement disorders, and stroke. AI-driven analytics and imaging, combined with brain–computer interfaces, robotics, virtual reality, and computer-assisted cognitive rehabilitation, enable more precise, personalized, and accessible recovery strategies across inpatient and home-based care settings.

AI is transforming musculoskeletal, sports, and pain medicine through advanced imaging, decision support, and procedural guidance. It enhances X-ray, CT, and MRI interpretation for fractures, osteoarthritis, and soft tissue pathology; personalizes rehabilitation and opioid-risk stratification; and augments ultrasound and fluoroscopy-guided interventions while requiring cautious clinician oversight.

Complex and oncologic rehabilitation increasingly benefits predictive models and continuous monitoring systems designed to identify risk, track progress, and align interventions with patient-specific needs. Such technologies hold value in managing patients with cancer, neuromuscular conditions, or challenging developmental profiles.

Across all areas, AI-driven systems are improving accessibility and continuity of care through virtual and telerehabilitation platforms, intelligent assistive and prosthetic devices, and streamlined clinical documentation and decision-support tools. Collectively, these developments signal a shift toward more integrated, data-informed models of rehabilitation that emphasize adaptability, precision, and patient-centered outcomes.

Relevance to Clinical Practice

Neurorehabilitation and AI

AI Methods to Enhance Prognosis

Neurorehabilitation is comprised of, but not limited to, traumatic brain injury, spinal cord injury, movement disorders and stroke. Traumatic brain injuries are leading causes of hospital deaths and disabilities worldwide numbering over 50 million per year globally with increasing morbidity and mortality.1 Patients with TBI present prognostic challenges given a wide array of risk factors combined with methods of injury and complexity of pathological neurological damage that can be overcome in part by the implementation of AI.2 In one recent study, random forest, decision tree, generalized linear model and ten-fold validation algorithms were used to identify significant variables in predicting prognosis including the motor component of the Glasgow coma scale, pupillary and cistern condition.2 It was found that distinct algorithms are valuable for prognostication from short-term mortality to long-term survival. Machine learning has additionally provided valuable input for predicting functional status and outcome in spinal cord injuries and ambulatory prognosis for those with incomplete injuries.3 These prognostic data can assist physicians in focusing rehabilitation strategies and neuromodulation techniques for the appropriate patients while forecasting potential outcomes.

Imaging/Lesion Mapping

AI in neurorehabilitation aims to improve diagnostics, clinical decision making and prognosis in patients with stroke, TBI, dementia, MS, Parkinson’s Disease, neuro-oncology and intracranial aneurysms and has been widely utilized with CT and MRI in stroke patients for image optimization and lesion mapping. AI assists in enhancing image resolution and has aided in correcting artifact to enhance accuracy of diagnosis. Radiologic AI assists physicians in providing time-sensitive care and in predicting complications such as hemorrhagic transformation in ischemic infarcts. AI uses segmentation in spinal pathologies to delineate vertebral compression fractures, disc compressions and scoliosis by subdividing regions at the pixel level for more accurate diagnostics and treatments.4

Rehab Tech: Cognitive and motor platforms, BCI

Brain computer interfaces, robotics, non-invasive brain stimulation and sensors are being utilized as therapeutic rehabilitation modalities to improve patient care. Computer assisted cognitive rehabilitation (CACR) utilizes multimedia and informatics to aid expression in patients with neurocognitive impairments.5 It has been shown to improve cognitive outcomes with patients with TBI, improve access and cut costs by caring for patients in their homes and improve memory. This is especially valuable as rehabilitation hospitals are often costly for patients, limited in availability and inaccessible for many patients globally. Robotic assisted rehabilitation has proven valuable in gait training and limb function while augmenting therapy intensity and providing real-time motor function feedback. Further fine motor skills can benefit from developing technologies. Additionally, virtual reality has been found to improve socialization, reduce anxiety and improve communication in TBI patients.2 Personalized rehabilitation platforms are emerging to provide a broader framework to provide individualized care to patient populations. Multimodal machine monitoring techniques are increasingly being used to monitor physiological processes in real time to contribute to personalized treatment programs and adapt recovery methods.2 These include electroencephalography (EEG), neuroimaging including CT and MRI, intracranial pressure (ICP), cerebral blood flow and brain oxygen saturation monitoring in traumatic brain injury patients. Machine learning algorithms can combine these data to assess clinical risks to prevent adverse outcomes for patients susceptible to variable pathological events and adjust treatment plans accordingly.2

Musculoskeletal medicine, sports medicine, and pain management

Imaging Interpretation

AI has emerged as a valuable tool in the interpretation of X-ray, CT, and MRI imaging for use in musculoskeletal, sports, and pain medicine. Applications include assistance with the detection of fractures, grading of osteoarthritis, and characterization of bony and soft tissue pathology on cross-sectional imaging of the spine, shoulder, hips, and knees, among other uses.6,7 AI tools can improve the efficiency of imaging interpretation and, in certain clinical contexts, may also improve diagnostic accuracy. Of note, the benefits of AI in imaging interpretation may be more pronounced in non-radiology specialists compared to radiologists. As such, AI radiology tools could assist physiatrists in their independent interpretation of diagnostic imaging.8,9

Clinical Decision-making and Treatment Plan Adjustment

The ability to gather, organize, and reframe clinical information allows AI to assist in clinical decision-making and in customizing rehabilitation treatment plans to patient circumstances. Some examples of AI applications in this area include opioid dependence risk stratification, predicting responsiveness to procedures such as spinal cord stimulator implantations, prediction models for return to sports after sports-related concussions or injuries such as ACL tears, and AI-based applications that modify exercise recommendations, patient education, and cognitive behavioral therapy based on patient progress, wearable device data, and motion tracking data.10,11,12

Procedural Imaging Enhancement

AI has demonstrated notable applications in enhancing imaging-guided interventions. For ultrasound-guided interventions, AI can support physiatrists through enhancement of needle visualization and automatic segmentation to help distinguish anatomical structures. Fluoroscopy-guided procedures can also be facilitated through AI-powered anatomical landmark detection and optimization of X-ray beam collimation to reduce overall radiation exposure. Radiofrequency ablation performed through either imaging guidance modality can also be enhanced by AI-powered thermal mapping algorithms that can improve the precision of ablations and decrease the risk of harming nearby tissues. It is worth noting that the efficacy of these AI tools can be challenged by use in patients with rare disorders, atypical anatomy, or implanted hardware or devices. As such, there remain risks to overreliance on AI for procedural guidance.13

Oncological, pediatric, and medicine in rehab

Cancer Rehab

Cancer rehabilitation plays a vital role in cancer survivorship and helps improve patient’s quality of life. To date, AI has been used in the field to: screen and identify patients who may benefit from rehabilitation, predict responses to exercise training and related outcomes, enhance patient engagement and support behavior change through tools such as feedback systems, coaching, and conversational agents, and enable more precise, individualized exercise prescriptions. By analyzing large volumes of health data, AI can help predict future risks and tailor interventions to each patient’s needs. It can also support clinicians in designing programs that adapt over time, making rehabilitation more flexible and responsive. Early studies show promising improvements in physical activity levels; however, significant challenges remain for large-scale implementation. These include ethical and legal concerns, the need for ongoing human oversight, and difficulties integrating AI into existing clinical workflows.14

Pediatric Rehab

There is a growing body of research on AI-based interventions aimed at improving participation among children and youth with disabilities or other diagnosed health conditions, highlighting the potential of AI in pediatric rehabilitation. Most interventions incorporate multiple AI applications, particularly robotics and human–machine interfaces (HMI). Other approaches, such as machine learning (ML) and natural language processing (NLP), are less common but show promising benefits for participation-focused interventions.15

AI technologies—such as machine learning, computer vision, and wearable sensors—allow efficient, objective assessment of children’s motor, sensory, and cognitive functions. They hold potentially helpful roles in therapy sessions. Particularly, integrating artificial intelligence (AI) into pediatric occupational therapy (OT) represents a significant advancement with the potential to enhance clinical practice and improve patient outcomes. AI-driven tools support more personalized interventions, expand telehealth capabilities, and facilitate interdisciplinary collaboration, helping therapists optimize care delivery, improve access to services, and better support children with diverse developmental needs.16

Cardiac and Pulmonary Rehab

Digital technologies (such as internet, mobile apps) have the potential to increase access and participation in cardiac rehabilitation by mitigating the challenges associated with traditional, facility-based Cardiac rehabilitation.17 Beyond exercise training, comprehensive health counseling and comorbidity management are important components in cardiac rehab. Smart devices and applications can extend beyond traditional healthcare and provide real-time health data and fosters patient empowerment.18 The integration of artificial intelligence into cardiac rehab offers the potential to shift cardiac rehab from a temporary care delivery model to a longitudinal delivery model, providing unique insights into patient health and wellness through digital data collected as part of patients’ daily lives.19

Recent studies have found that AI integration to pulmonary rehab can significantly improve exercise capacity in individuals with chronic respiratory diseases. Particularly, patients in an AI assisted pulmonary rehab group had a significant improvement in 6-minute walk distance. Current studies have not been done on participants with different etiologies for pulmonary rehab. While individualization with AI has potential in improving clinical use results, there will still need a balance between standardization and personalization.20

AI hardware and software

AI to Enhance Patient Independence – Prosthetics, Exoskeletons and Wheelchairs

AI is revolutionizing assistive devices, offering patients greater function and independence in the community. Machine learning embedded prosthetics can both gather data about a patient’s function but also adapt over time and integrate more seamlessly with a wearer’s physiology.21 AI enabled prosthetics and exoskeletons have the ability to better predict intended movement, augment joint strength and improve gait mechanics.22 These devices have also allowed for real-time adaptive algorithms to adjust robotic level of assistance with the intention to improve motor relearning in stroke. Prosthetics have also been equipped with smart sockets that use sensors to detect changes in a residual limb overtime and auto adjust the socket for a safe and secure fit.21 Wheelchairs can now process complex environmental data through advanced AI, cameras, and LiDAR sensors, smart wheelchairs and offer intelligent, environment aware mobility that enables autonomous navigation and collision avoidance.22 At the most cutting edge are devices that allow for brain – computer interface control such as symbiotic neuroprosthetics and devices under myoelectric control. For example, devices that link brain computer interface to functional electrical stimulation have allowed patients with tetraplegia to perform coordinated grasps and other upper extremity movements.22 These technologies give hope to patients even with the most severe functional impairments.

Telerehabilitation and Remote Monitoring

AI has allowed for the extension of high-quality rehab care into the home, allowing for more frequent and real-world therapeutic intervention for patients.23 “Virtual physiotherapy assistants” can use AI computer vision and wearable sensors to monitor daily exercises, providing patients with real-time, corrective feedback on their posture and exercises, without requiring a physical clinic visit.23 Wearable devices are able to capture massive amounts of continuous metrics on data such as gait metrics, balance, and vital signs in the community setting and then generate better objective functional recovery goals.24 Telerehab with remote controlled therapy robots has also been proposed. These home robots can enable therapists to dynamically interact and make hands on adjustments during exercises, from afar.25 By reducing the need for in person appointments, patients may be more compliant with their prescribed rehabilitation programs.

Integration of AI into the rehab clinician’s workflow

Introduction of AI into clinical workflow can make the day to day caring for patients run smoother and with greater insights.26 This integration has already happened at major institutions across the US and across medical specialties. Rehab patients are seen by many healthcare providers – OT, PT, SLP, social work, psychology, nurses, physicians and more. Natural Language Processing has been utilized to rapidly mine the massive volumes electronic health record notes to extract hidden clinical patterns, social determinants of health, and functional risk factors that may be missed.26 Additionally, ambient clinical intelligence systems have been deployed to passively listen to patient provider encounters, instantly generating structured note drafts and reducing the documentation burden that frequently drives physician burnout.27 Reducing the significant amount of time to extract data from an electronic health record and for documentation can allow clinicians to focus on higher level decision making and hands on patient care.26,27

As artificial intelligence continues to evolve, its influence on physical medicine and rehabilitation (PM&R) is expanding from isolated research efforts to integrated clinical practice. Across neuro-rehabilitation, musculoskeletal and pain management, complex and oncologic care, and assistive technology systems, AI is enabling earlier detection, more accurate prognostication, and smarter, personalized interventions. The convergence of advanced imaging, predictive modeling, and adaptive therapeutic platforms is redefining what is possible in functional recovery and long-term outcomes. Continued collaboration among clinicians, data scientists, and engineers will be essential to ensure these innovations are validated, equitable, and aligned with patient-centered goals—ultimately advancing the science and delivery of rehabilitation medicine.

AI in risk stratification and healthcare administration

AI has demonstrated value in analyzing hospital readmission risk for patients being discharged from an acute hospital or a post-acute care sites including Inpatient Rehab Facilities (IRF) and a Skilled Nursing Facilities (SNF).28,29 Moreover, within the rehab facility setting, AI has been applied in predicting the level of functional recovery in certain populations such as in stroke patients.30 These risk-stratification and outcome prediction capabilities have the potential to inform transitions in care and optimal placement to ensure safe and effective rehabilitation and reintroduction to the community.

AI also has the potential to facilitate prior authorization for post-acute care admissions. Electronic prior authorizations have been shown to reduce the median time from request to decision from 18.7 hours to 5.7 hours.31 Driven by recent federal mandates to speed up approval turnaround and a widespread industry focus on administrative cost reduction, AI adoption has become mainstream in recent years. Strategies to optimize AI enhanced prior authorizations include auto-populating pertinent information from patient charts including functional status, therapy progress and interdisciplinary team conference items.32 AI tools can further assist via prediction algorithms utilized in claim denials and facilitating success in peer-to-peer reviews.33 Expediting prior authorization with AI tools can free up bed space for acute care floors and maximize the benefits of early treatment potentials for rehabilitation patients.

AI in medical education

AI tools show strong potential to help physical rehabilitation doctors learn and improve. Modern chatbots now answer about 83–84% of specialty exam questions, similar to doctors-in-training, and can quickly generate treatment plans or act as 24/7 study support.34,35 However, they still make errors, especially in complex cases, and require proper guidelines, training, and ethical safeguards before widespread use.36 While AI assisted learning can accelerate a learner’s acquisition of knowledge, it also carries significant risk. Overreliance on AI can foster a pattern of passive engagement which does not allow for development of foundational skills such as diagnostic reasoning.37 Those seeking to integrate AI into PM&R education should set up safeguards and oversight to make sure learning objectives and fundamental competencies are met.

Cutting Edge/Unique Concepts/Emerging Issues

AI in PM&R is advancing toward multimodal prognostic models that fuse imaging, biomechanics, and sensor data to predict functional recovery and personalize rehabilitation dosing across neuro, MSK, and oncologic populations. This evolution connects with emerging issues such as AI-driven telerehabilitation with real-time kinematic feedback, adaptive robotics and exoskeletons, and precision “digital twin”– style simulations for optimized therapy intensity and progression.38 In parallel, interest is growing in generative and ambient AI for automating documentation, decision support, and workflow optimization, as well as in equitable deployment to expand access for underserved groups in cancer and complex rehab settings.

Among notable issues for AI use in medical practice is its propensity to generate plausible but erroneous outputs often referred to as “hallucinations” or “misinformation”.39 These errors can manifest in many ways, including abnormalities noted on imaging, fabricated data, or incorrect information in large language models.40 Given the plausibility of these hallucinations, they can go unnoticed and may negatively impact clinician decision-making and patient safety. Moreover, the problems generated by hallucination can be compounded when multiple AI systems work together, as errors stack on top of each other when one system uses unreliable information to generate its own flawed output.40 Multiple strategies to reduce the impact of hallucinations have been explored; however, fully eliminating these errors may not be possible.40 As such, AI tools must be used with caution by trained clinicians who can verify AI-generated contributions with reliable medical sources.

Gaps in Knowledge/Evidence Base

Despite rapid innovation, most AI models in rehabilitation are developed on small, homogeneous datasets with limited standardization, external validation, or multimodal inputs, constraining generalizability and real-world impact. Many studies prioritize algorithmic accuracy over clinically meaningful endpoints such as function, participation, and long-term outcomes, and often lack robust comparators or health-economic evaluation. Critical gaps include transparency and explainability, integration into routine workflows, regulatory clarity, and rigorous evaluation of AI-supported robotics, motion-analysis platforms, and digital therapeutics across neuro, MSK, and oncologic rehab. Evidence on equity, bias, and fair performance across diverse populations in PM&R remains especially limited.

References

  1. Khalili, H., et al., Prognosis prediction in traumatic brain injury patients using machine learning algorithms. Sci Rep, 2023. 13(1): p. 960.
  2. Wei, Z., L. Meng, and W. Chong, Advancements in the application of multimodal monitoring and machine learning for the development of personalized therapeutic strategies in traumatic brain injury. Front Hum Neurosci, 2025. 19: p. 1695336.
  3. Draganich, C., et al., Predictive modeling of ambulatory outcomes after spinal cord injury using machine learning. Spinal Cord, 2024. 62(8): p. 446–453.
  4. Fiani, B., et al., Current uses, emerging applications, and clinical integration of artificial intelligence in neuroradiology. Rev Neurosci, 2022. 33(4): p. 383–395.
  5. Wang, Q., et al., Development and Application of Medicine-Engineering Integration in the Rehabilitation of Traumatic Brain Injury. Biomed Res Int, 2021. 2021: p. 9962905.
  6. Glinkowski, W.M., et al., Artificial Intelligence in Orthopaedics: Clinical Performance, Limitations, and Translational Readiness-A Review. J Clin Med, 2026. 15(5).
  7. Bousson, V., et al., Application of artificial intelligence to imaging interpretations in the musculoskeletal area: Where are we? Where are we going? Joint Bone Spine, 2023. 90(1): p. 105493.
  8. Prucker, P., et al., Real-world clinical impact of three commercial AI algorithms on musculoskeletal radiography interpretation: A prospective crossover reader study. Int J Med Inform, 2026. 205: p. 106120.
  9. Guermazi, A., et al., Improving Radiographic Fracture Recognition Performance and Efficiency Using Artificial Intelligence. Radiology, 2022. 302(3): p. 627–636.
  10. Bensel, V.A., et al., Artificial intelligence in spine care: A scoping review of treatment applications. N Am Spine Soc J, 2026. 25: p. 100827.
  11. Gopal, J., et al., Machine learning predicts spinal cord stimulation surgery outcomes and reveals novel neural markers for chronic pain. Sci Rep, 2025. 15(1): p. 9279.
  12. Faysal, J.A., et al., Developing and Validating a Machine Learning Algorithm to Predict the Risk of Incident Opioid Use Disorder Among OneFlorida+ Patients: Prognostic Modeling Study. J Med Internet Res, 2026. 28: p. e79482.
  13. Dubey, A., et al., Role of Artificial Intelligence in Musculoskeletal Interventions. Cancers (Basel), 2025. 17(10).
  14. Bland, K.A., et al., Artificial Intelligence Meets Cancer Rehabilitation: Emerging Evidence for Exercise and Physical Activity Interventions. Cancer Control, 2026. 33: p. 10732748261432280.
  15. Kaelin, V.C., et al., Artificial Intelligence in Rehabilitation Targeting the Participation of Children and Youth With Disabilities: Scoping Review. J Med Internet Res, 2021. 23(11): p. e25745.
  16. Sharma, N., Use of AI in pediatric occupational therapy: A review. Rehabil Sci, 2024. 9(2): p. 21–26.
  17. Wongvibulsin, S., et al., Digital Health Interventions for Cardiac Rehabilitation: Systematic Literature Review. J Med Internet Res, 2021. 23(2): p. e18773.
  18. Pepera, G., et al., Comprehensive and personalized approach is a critical area for developing remote cardiac rehabilitation programs. World J Clin Cases, 2024. 12(12): p. 2009–2015.
  19. Hamilton, D.E., et al., Digital Technologies and Artificial Intelligence in Cardiac Rehabilitation: A Narrative Review. J Cardiopulm Rehabil Prev, 2025. 45(3): p. 169–180.
  20. Cinkavuk, E., E. Calik, and N. Vardar-Yagli, The effect of artificial intelligence-assisted pulmonary rehabilitation on exercise capacity: A systematic review and meta-analysis. Int J Med Inform, 2026. 211: p. 106336.
  21. Abbas, G.H., et al., AI-Driven Rehabilitation Robotics: Advancements in and Impacts on Patient Recovery. Cureus, 2025. 17(10): p. e94273.
  22. Sumner, J., et al., Artificial intelligence in physical rehabilitation: A systematic review. Artif Intell Med, 2023. 146: p. 102693.
  23. Olawade, D.B., et al., Enhancing home rehabilitation through AI-driven virtual assistants: a narrative review. Ann Transl Med, 2025. 13(5): p. 61.
  24. Gu, B., et al., Advancements in Wearable Sensor Technologies for Health Monitoring in Terms of Clinical Applications, Rehabilitation, and Disease Risk Assessment: Systematic Review. JMIR Mhealth Uhealth, 2026. 14: p. e76084.
  25. Rasa, A.R., Artificial Intelligence and Its Revolutionary Role in Physical and Mental Rehabilitation: A Review of Recent Advancements. Biomed Res Int, 2024. 2024: p. 9554590.
  26. Scharp, D., et al., Natural Language Processing Applied to Clinical Documentation in Post-acute Care Settings: A Scoping Review. J Am Med Dir Assoc, 2024. 25(1): p. 69–83.
  27. Guo, Y., et al., What Do Clinicians Edit in Ambient AI-Drafted Clinical Documentation? A Qualitative Content Analysis. medRxiv, 2026: p. 2026.01. 05.26343473.
  28. Brown, Z., et al., Augmenting a Transitional Care Model With Artificial Intelligence Decreased Readmissions. J Am Med Dir Assoc, 2023. 24(7): p. 958–963.
  29. Howard, E.P., et al., Machine-Learning Modeling to Predict Hospital Readmission Following Discharge to Post-Acute Care. J Am Med Dir Assoc, 2021. 22(5): p. 1067–1072.e29.
  30. Miyazaki, Y., et al., Improvement of predictive accuracies of functional outcomes after subacute stroke inpatient rehabilitation by machine learning models. PLoS One, 2023. 18(5): p. e0286269.
  31. Clayton, D., et al., Evaluation of the Fast Prior Authorization Technology Highway. J Am Pharm Assoc (2003), 2022. 62(6): p. 1843–1847.
  32. Young, D.L., et al., Dynamic Prediction of Post-Acute Care Needs for Hospitalized Medicine Patients. J Am Med Dir Assoc, 2024. 25(7): p. 104939.
  33. Ifejika, N.L., et al., Improving access to stroke rehabilitation and recovery: a policy statement from the American Heart Association/American Stroke Association. Stroke, 2025. 56(9): p. e218–e233.
  34. Bitterman, J., et al., Advancements in large language model accuracy for answering physical medicine and rehabilitation board review questions. Pm r, 2025. 17(9): p. 1091–1096.
  35. Ayyıldız, A., et al., Artificial intelligence versus physical medicine and rehabilitation residents: Can ChatGPT compete in clinical exam performance? Pm r, 2026. 18(1): p. 52–59.
  36. Silver, J.K., M.R. Dodurgali, and N. Gavini, Artificial Intelligence in Medical Education and Mentoring in Rehabilitation Medicine. Am J Phys Med Rehabil, 2024. 103(11): p. 1039–1044.
  37. Hosseini, S.M., AI misuse and passiveness of students in medical education. Advances in Physiology Education, 2025. 49(4): p. 1009–1013.
  38. Lauer-Schmaltz, M.W., Digital Twin in Healthcare: Patient System Modelling for Rehabilitation by Exoskeleton. 2024.
  39. Hatem, R., B. Simmons, and J.E. Thornton, A call to address AI “hallucinations”          and how healthcare professionals can mitigate their risks. Cureus, 2023. 15(9).
  40. Granstedt, J., et al., Hallucinations in medical devices. Artificial Intelligence     in the Life Sciences, 2025: p. 100145.

Author Disclosure

Andrés Gronda, MD
Nothing to Disclose

Cherie Chen, DO
Nothing to Disclose

Ariana Schanzer, MD
Nothing to Disclose

Emily Sears, DO
Nothing to Disclose

Salvador Portugal, DO, MBA
Nothing to Disclose