Low back pain (LBP) is a common problem with complex causes, of which some are modifiable. Physical factors like strength, movement, and pain play a big role, but measuring all these factors accurately is tricky. This is where Artificial Intelligence (AI) comes in. This projects aims to develop an AI solution (in the form of a mobile application) that can measure four key components of the physical factor of LBP, such as how quickly you can stand up five times, your spine's flexibility, how you walk, and your pain levels while moving. The measurements taken by the mobile application will be compared against those of trained physiotherapists to ensure its accuracy. If successful, this AI solution will be a game-changer. Physiotherapists will be able to remotely track the progress of their LBP patients. The data gained from the remote tracking will allow physiotherapists to have a better understanding of the individual profile of each LBP patient and adjust their treatment accordingly, hence allowing for better care and more effective LBP management. In short, this project aims to harness the power of AI to make managing LBP easier for both patients and physiotherapists.
Background: Low back pain (LBP) is a complex condition and its causes are multifactorial, of which the physical, lifestyle, cognitive and emotional factors are potentially modifiable. Due to the complexity of LBP, Artificial Intelligence (AI) can be used to accurately measure and analyze large amounts of data from different sources to aid in the assessment and management of LBP. Objective: Development of an AI model that accurately assesses and measures 4 core components that comprise the Physical factor of LBP. The 4 core components are functional activity (measured using the 5 times sit-to-stand task - 5xSTS), trunk range of motion (ROM), gait pattern and pain levels during movement. Methods: The project aims to recruit 120 LBP patients receiving care at SGH Physiotherapy. For the first (primary) study (n=103), we will compare the measurements (5xSTS, trunk ROM, gait pattern and pain levels during movement) taken by the AI model against that of a trained assessor/physiotherapist. For the second study (n=17), following integration of the AI model with our industry partner's platform, a pilot study will be conducted to assess the feasibility and usability of a minimum viable product. Planned Analysis: For the first study, the Bland-Altman plot will be used to compare the measurements taken by the AI model against that of a trained assessor/physiotherapist. If our hypothesis is correct, the results should show narrow limits of agreement between the 2 methods of measurement. Descriptive statistics will be used for the second study. We anticipate that there will be positive feedback and satisfaction from use of the minimum viable product. Discussion: Successful development of our solution will allow accurate remote tracking of the progress made by LBP patients. This will support/assist physiotherapists in clinical decision-making, hence allowing for more effective management of LBP.
Study Type
OBSERVATIONAL
Enrollment
120
This intervention involves developing an artificial intelligence (AI) model to objectively assess four physical parameters relevant to low back pain (LBP): 1) sit-to-stand performance, 2) trunk range of motion, 3) gait pattern, and 4) facial expression-based pain levels during movement. The AI model processes video recordings of participants performing these tasks to extract movement and facial data, providing standardized measurements. The tool is designed to assist physiotherapists in clinical decision-making by offering consistent and accurate assessments compared to traditional observational methods.
Singapore General Hospital
Singapore, Singapore
5 times sit-to-stand
The test provides a method to quantify functional lower extremity strength and/or identify movement strategies a patient uses to complete transitional movements. The score is the amount of time (to the nearest decimal in seconds) it takes a patient to transfer from a seated to a standing position and back to sitting five times.
Time frame: Baseline
Trunk Range Of Motion (ROM)
Measurement of trunk (lumbar spine) ROM for: 1. Flexion (forward bending) and Extension (backward bending) in the sagittal plane 2. Right Lateral Flexion (right tilt) and Left Lateral Flexion (left tilt) in the frontal plane 3. Right Rotation (right turn) and Left Rotation (left turn) in the transverse plane
Time frame: Baseline
Gait pattern
Walking pattern of patients with low back pain
Time frame: Baseline
European Quality of Life Questionnaire (EQ-5D-5L)
Patient Reported Outcome Measure (PROM) that represents an estimation of the patient's perceived quality of life and state of health.
Time frame: Baseline, 3 months and 6 months
Pain Catastrophizing Scale (PCS)
Patient Reported Outcome Measure (PROM) describing thoughts and feelings that individuals might experience when in pain.
Time frame: Baseline, 3 months and 6 months
Hospital Anxiety and Depression Scale (HADS
Patient Reported Outcome Measure (PROM) used to evaluate depression and anxiety. It has 14 items over 2 subscales: depression and anxiety.
Time frame: Baseline, 3 months and 6 months
Depression Anxiety Stress Scales 21 (DASS21)
Patient Reported Outcome Measure (PROM) designed to measure the emotional states of depression, anxiety and stress. It has 21 items over 3 subscales: depression, anxiety and stress.
Time frame: Baseline, 3 months and 6 months
Short version of Örebro Musculoskeletal Pain Screening Questionnaire (ÖMPSQ-SF)
Patient Reported Outcome Measure (PROM). 10 item questionnaire that was developed from the Örebro Musculoskeletal Pain Screening Questionnaire (ÖMSPQ). The ÖMSPQ "was developed as a tool to assist in the early identification of yellow flags and patients risking the development of work disability due to the pain".
Time frame: Baseline
STarT Back Screening Tool (SBST)
9 item screening tool that is designed to subgroup low back pain patients for prognostic indicators that are relevant to their care.
Time frame: Baseline
Numeric Rating Scale (NRS)
NRS is a single 11-point numeric scale for rating pain intensity.
Time frame: Baseline, 3 months and 6 months
Patient Specific Functional Scale (PSFS)
PSFS is a self-reported, patient specific measure, that is designed to assess functional change in patients with musculoskeletal disorders. Patients are asked to identify up to five functional activities that are important to them and of which they have difficulty performing, after which the patients then rate each activity on an 11-point scale on the current level of difficulty they have performing the activity.
Time frame: Baseline, 3 months and 6 months
This platform is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare professional.