AI camera app linked to better shoulder pain relief in Spanish trial

Adults with shoulder pain who used an AI app through a phone or computer camera reported larger improvements at some points of a Spanish clinical trial published in 2026. The study followed 46 people with pain linked to the rotator cuff, the group of muscles and tendons that stabilizes the shoulder. The researchers tested whether adding the RehBody app to usual physiotherapy improves pain and function. The results favored the app in several measures, although not in every week of follow-up.

The participants had an average age of 59 years. The team split them into two groups of 23 people each and followed them for 12 weeks. One group received conventional physiotherapy only. The other received the same treatment plus RehBody, which uses the front camera of a cellphone or computer to analyze movements and give feedback in real time.

The trial took place at the rehabilitation and physiotherapy service of the Hospital Universitario Virgen Macarena, in Seville, Spain. It was a randomized, controlled clinical trial conducted at a single center, with blinded assessors. Outcomes were measured at the start and at weeks 2, 4 and 12.

What the 12-week trial measured in 46 patients

Pain reduction in the app group passed the minimum clinically important difference of 1.3 points on the numeric pain rating scale at weeks 4 and 12. That threshold is the smallest change that patients are likely to notice as meaningful. At weeks 2 and 4, the differences between groups favored the app, with average changes of -0.7 and -1.01 points, respectively.

Function of the upper limb improved more in the app group at week 4, with an average difference of -7.3 points on the QuickDASH questionnaire, a tool that measures arm, shoulder and hand function. Fear-avoidance beliefs, measured by the FABQ questionnaire, fell more in the app group at week 12, with an average difference of -7.6 points. After the intervention, fewer patients in the app group needed additional health care: 3, against 10 in the control group.

The researchers recruited 76 patients between March and May 2025. Of these, 46 met the criteria and agreed to take part. Follow-up ended in September 2025, and the analysis followed the intention-to-treat principle, which keeps every participant in the group they were first assigned to.

How the camera-based app tracks movements at home

RehBody identifies body reference points through the camera. It guides the patient through each exercise, recognizes the correct form and timing of repetitions, and records data such as range of motion and adherence. The app needs no extra sensors or special equipment, only the device the patient already owns.

Keep reading: NOAA puts 83% odds on a historic El Niño in October–December 2026

The tool was designed for health professionals to prescribe and monitor therapeutic exercise programs. Professionals follow the patient’s performance on a dashboard and can adjust the plan when needed. According to the study, therapeutic exercise is the first-line treatment for rotator cuff-related pain, but adherence to home programs is often limited by the lack of supervision.

Brazilian surgeons’ view on AI as a complement to care

Eduardo Malavolta, president of the Brazilian Society of Shoulder and Elbow Surgery (SBCOC), commented on the technology. “Artificial intelligence can be an interesting tool to expand patient monitoring during rehabilitation. By using the camera to analyze how exercises are performed and giving real-time feedback, the technology can help patients feel more confident when moving at home and make follow-up between visits or physiotherapy sessions easier,” he said.

Malavolta also stressed the supporting role of the tool. “It is important to prevent patients from using artificial intelligence to treat alone a condition that should be evaluated by a professional. The message we need to convey is a balance between innovation and safety, understanding AI as a complementary tool, not a substitute for the doctor or the physiotherapist.”

Limits of the app and who should guide exercises

Shoulder pain can have several causes, and a symptom alone does not define a diagnosis. An individual assessment, including medical history, physical examination and, when indicated, imaging tests, remains necessary to guide treatment. Patients should not use the app to diagnose their own condition or choose exercises without guidance, since digital tools can provide incomplete or outdated information.

Two authors of the study, Alejandro Suero-Pineda and Olga Villar-Alises, have ties to the company that developed RehBody. The other authors declared no conflict of interest. The trial was registered in the Open Science Framework and approved by the hospital’s ethics committee.

Learn more: Smart home devices keep running offline through local protocols

Why home exercise programs often lose patients

Rotator cuff-related pain is one of the most common musculoskeletal complaints in primary care. It affects the structures that stabilize and move the shoulder and can limit daily activities. Clinical guidelines recommend structured therapeutic exercise as the base of conservative treatment.

Keeping up a program at home is often difficult. Patients report a lack of immediate feedback, trouble advancing through the exercises and no objective monitoring. Telerehabilitation and computer vision tools aim to reduce these limits by offering real-time guidance and performance data to the professional.

The study suggests that adding the app to conventional treatment may improve pain at certain points of follow-up, favor upper limb function and reduce the need for additional care after 12 weeks. The results apply to patients with rotator cuff-related pain treated at a specific hospital center. The app’s data show the professional how often exercises are done and how well they are performed, without requiring a visit for each exercise.

The research did not test the app on its own, without professional follow-up. All participants received conventional treatment, and the experimental group had the app as an addition. The study was carried out in a hospital rehabilitation setting, so applying the findings to other contexts and populations calls for caution. Patients with other causes of shoulder pain, such as advanced arthritis or instability, were not the focus of the research.

Results for pain and function were observed at specific points of the follow-up, not at every stage. For professionals and patients, the main point is integration: a camera-based AI tool can expand monitoring of home exercises, while the definition of the plan and the diagnosis remain with the health team. The study was published in the journal Musculoskeletal Science and Practice, volume 85, article 103626, in October 2026. The main authors are Olga Villar-Alises, Celia Cónstenla Cortés, Manuel Rodríguez-Piñero Durán, Pablo Rodríguez Sánchez-Laulhé, Javier Martinez-Calderon and Alejandro Suero-Pineda.