Using AI-Assisted Cameras to Track Hand Exercises in Scleroderma Patients

Volunteers assigned to the experimental group within the scope of the study will use an AI-integrated, markerless telerehabilitation system that operates with a standard mobile or web camera and requires no external wearable sensors, markers, or complex hardware.

Based on machine learning and computer vision algorithms, this intervention can map twenty-one distinct anatomical reference points on the hand, wrist, and fingers in three-dimensional space in real-time, within milliseconds. The application utilizes movements previously validated for biomechanical accuracy by a physiotherapist; as the patient performs exercises at home, the system continuously compares their current range of motion (ROM) against reference values using algorithms. The patient receives immediate visual and auditory feedback on-screen regarding the accuracy, speed, and angle of the movement; simultaneously, the system creates an autonomous, asynchronous monitoring mechanism by automatically recording daily changes in ROM and session participation rates (treatment adherence) to a secure, cloud-based database.

• AI-supervised exercise

Participants in the control group will perform an eight-week specific hand exercise regimen-identical to that of the experimental group in terms of biomechanics, intensity, and physiological parameters-under the face-to-face supervision of a physiotherapist. The exercise protocol to be administered to both groups will consist of an evidence-based rehabilitation module comprising isometric strengthening, flexor and extensor tendon gliding, and active/passive stretching components, aligned with the pathophysiology of SSc.

Participants in the control group will perform an eight-week specific hand exercise regimen-identical to that of the experimental group in terms of biomechanics, intensity, and physiological parameters-under the face-to-face supervision of a physiotherapist. The exercise protocol for both groups will consist of an evidence-based rehabilitation module incorporating isometric strengthening, flexor and extensor tendon gliding, and active/passive stretching components, aligned with the pathophysiology of SSc.

• Physiotherapist-supervised exercise

The HAMIS test encompasses nine distinct basic motor tasks: finger flexion and extension, thumb abduction, pincer grip, wrist volar and dorsal flexion, and forearm supination and pronation (2). Each movement is scored on a scale from 0 (full, normal range of motion) to 3 (inability to perform the movement), yielding a dysfunction score ranging from 0 to 27. The clinical utility of this test lies in its ability to detect cutaneous fibrosis and limitations in finger range of motion (ROM) with high accuracy (3).

To determine the extent of flexor tendon contractures-commonly observed in scleroderma patients due to excessive collagen accumulation-and the severity of sclerodactyly, the distance between the fingertips (pulp) and the distal palmar crease will be measured both actively and passively in centimeters (cm) using a standard metric ruler. Biomechanically, a value of 0 cm (indicating full contact of the fingertip pulp with the palmar crease) is expected during the full range of finger flexion; positive values directly reflect flexor system dysfunction and joint contracture. The literature indicates that this measurement method is practical, time-efficient, and highly reliable for clinical monitoring and for assessing joint motion in cases of restricted digital movement (particularly when combined with goniometry).

The test is based on the principle of the patient sequentially inserting pegs (small rods) into nine holes on a wooden or plastic board of standard dimensions and then rapidly removing them (6). The test begins with the dominant hand, and the time taken for patients to insert and remove all pegs is recorded in seconds using a stopwatch (7).

Gross hand grip strength will be measured three consecutive times using a Jamar dynamometer-with the patient in a seated position, the shoulder adducted, and the elbow flexed at 90 degrees and held against the trunk-and the average value will be recorded in the system in kilograms (kg) or pounds (lbs).

This questionnaire is a comprehensive, six-dimensional, 57-item tool that specifically assesses not only physical functions but also activities of daily living, work performance, pain, aesthetic perception, and patient satisfaction (9). The MHQ score is normalized to a range of 0 to 100; higher scores in dimensions other than pain indicate a better clinical status and increased functionality.

The Turkish cultural adaptation and validation of the DASH-which consists of 30 items and is scored on a scale of 0 to 100 (with higher scores indicating greater disability)-were introduced to the literature by Düger et al. (2006) (ICC: 0.910). The DASH is a 30-item, self-administered questionnaire designed to measure physical function and symptoms in patients with musculoskeletal disorders of the upper limb. It assesses the patient's ability to perform specific daily activities, alongside the severity of pain, tingling, weakness, and the overall impact of the condition on social activities, work, and sleep. A score of 0 indicates no disability (best possible outcome), while a score of 100 indicates the most severe disability (worst possible outcome).

The SHAQ is a disease-specific, self-administered instrument used to evaluate functional disability and disease burden in patients with systemic sclerosis (scleroderma). It incorporates the standard 20-item Health Assessment Questionnaire Disability Index (HAQ-DI), which assesses difficulty in performing activities of daily living across 8 domains, combined with disease-specific Visual Analog Scales (VAS). These VAS components evaluate the severity of scleroderma-specific symptoms, including Raynaud's phenomenon, digital ulcers, gastrointestinal involvement, breathing difficulties, pain, and overall disease severity.

The standard HAQ-DI component is scored from 0 to 3. The disease-specific VAS items are typically scored on a 0 to 100 scale (or 0 to 150 mm, converted to a 0-3 scale depending on the specific trial methodology).

For both the HAQ-DI and VAS components, higher scores indicate greater functional impairment or more severe symptom burden (worst poss

Consisting of a total of 10 items, this scale rapidly and reliably measures the perceived usability of digital interfaces and interactive systems. The Turkish validity and reliability study of the scale (SUS-TR) was conducted by Demirkol and Şeneler (2018), demonstrating that, with its two-factor structure, it is a valid and highly reliable psychometric tool for Turkish users. The scale comprises Likert-type questions scored from 1 to 5; the resulting scores are converted into a 0-100 point scale for interpretation, with higher scores indicating high system usability. This tool enables an objective analysis of the technological real-time feedback system's usability from the patient's perspective.

the Semmes-Weinstein Monofilament (SWM) test will be used to assess cutaneous sensory loss that may develop secondary to SSc-specific microvascular ischemic damage, digital ulcerations, or neural compression. Nylon monofilaments of varying thicknesses and calibrated buckling forces are applied to the patient's fingertips for 1-1.5 seconds, using enough pressure to create a slight bend (C-shape). In the assessment of tactile sensation, force values between 1.65 and 2.83 are reported as "normal" touch sensation (8), and this test is a validated method for detecting somatosensory impairment caused by cutaneous fibrosis in sensory tissue.