Spinal Stability & Pelvic Tilt Kinematics
Inertial sensor-based lumbopelvic movement analysis and a clinical motion analysis platform for rehabilitation assessment.
Overview
Posterior pelvic tilt (PPT) is a critical biomechanical parameter in spine rehabilitation, yet its clinical assessment remains largely subjective. Our lab uses IMU sensors (Movella DOT) to quantify pelvic tilt kinematics during functional movements, and has developed SpineView — a web-based clinical platform that transforms sensor data into actionable biomechanical scores.
This research-to-platform pipeline bridges the gap between laboratory kinematics research and bedside clinical application.
1. Pelvic Tilt Kinematics Research
| Project | Stage |
|---|
| Clinical vs. inertial sensor measurements of posterior pelvic tilt | 📜 Published |
| Posterior pelvic tilt and radiographic parameter correlation | 📜 Published |
| Machine learning analysis of squat movements | 📊 Analyzing |
| AI-based prediction of premature posterior pelvic tilt | ✅ Completed |
| Biomechanical analysis of PPT during squats using inertial sensors | 💵 Funded (Ongoing) |
Biomechanical Parameters
- PPA (Posterior Pelvic Angle): pelvis end - pelvis initial
- LL (Lumbar Lordosis): 180° - lumbar pitch + pelvis pitch
- HA (Hip Angle), KA (Knee Angle)
- SPPT-I / SPPT-E (Seated Pelvic Position Test, Initial/End)
Methods
- Sensors: Movella DOT IMU (spine + pelvis dual-sensor)
- Signal Processing: Gaussian smoothing, inflection point detection via gradient analysis
- Analysis: Python (NumPy, SciPy), custom parameter extraction pipeline
SpineView is the clinical translation of our kinematics research into a real-time assessment tool.
Analysis Modules
| Module | Function | Key Outputs |
|---|
| Real-time Streaming | Live BLE sensor visualization | Flexion-extension curves, 60-second rolling window |
| Research Analysis | Inflection point detection, lordosis | PPA, LL, SPPT-I/E, Gaussian smoothing |
| Squat Analysis | Flexion-extension kinematics | ROM, angular velocity, acceleration, jerk, coordination (R²) |
| Sit-to-Stand Analysis | Functional movement scoring | Lordosis score, Hip Hinge score, Trunk Lean score (0-100) |
Clinical Scoring System
| Score | Target | Optimal |
|---|
| Lordosis Index | Spine FE - Pelvis FE | ≤ 0° |
| Hip Hinge Score | Pelvis ROM optimization | 30-60° |
| Trunk Lean Score | Peak spine flexion + timing + smoothness | 0-100 |
Technical Stack
| Component | Technology |
|---|
| Backend | Django 5.2 (Python 3.12) |
| Database | PostgreSQL 16 |
| Data Processing | pandas, NumPy, SciPy |
| Frontend | Tailwind CSS + Plotly.js |
| Sensors | Movella DOT (BLE) |
| Deployment | Docker + docker-compose |
Development Status
| Milestone | Status |
|---|
| Core analysis engine (CSV) | ✅ Completed |
| Real-time BLE streaming | ✅ Completed |
| Squat & sit-to-stand modules | ✅ Completed |
| Docker deployment | ✅ Completed |
| Multi-patient data management | 🔄 In Progress |
| Clinical validation study | 💡 Planned |
Key Presentations
| Year | Title | Venue |
|---|
| 2023 | Relationship between Posterior Pelvic Tilt and Radiographic Parameters | 51st KARM Annual Meeting |
| 2022 | Clinical Versus Inertial Sensor Measurements of Posterior Pelvic Tilt | 50th KARM Anniversary Meeting |
Funding
- Biomechanical Analysis and Clinical Correlation of Posterior Pelvic Tilt During Squats Using Inertial Sensors — PNUH Biomedical Research Institute, 2025 (Ongoing)
- Prediction of Premature Posterior Pelvic Tilt Using AI and Wearable Sensor — Kosin University College of Medicine, 2020-2021
- Smart Posture Correction System — Pukyong National University, 2019-2020