ScolioDetect revolutionizes scoliosis screening with precise detection. Conventional methods yield 90% false positives and fail to detect early-stage scoliosis—our solution delivers 3° accuracy, safeguarding millions of teenagers while reducing costs.

Your inspiration

Adolescent Idiopathic Scoliosis (AIS) affects 1-3% of teenagers globally, requiring detection for 1.3 billion adolescents. Without early detection, cases can progress to require surgery. Current methods cannot deliver both accuracy and portability at scale. Visual assessments produce 90% false positives and are highly operator-dependent. Advanced equipment achieves >7° error but stays expensive and bulky. Accuracy must meet portability. Our breakthrough: spinal deformities create measurable walking asymmetries detectable before visible symptoms. This inspired our solution: transform walking patterns into precise, radiation-free AIS detection.

How it works

ScolioDetect transforms walking patterns into spine curve measurement through an integrated sensor-AI system comprising wearable sensors, data acquisition, and AI algorithms. IMU sensors at key anatomical landmarks capture spinal dynamics using inertial data during gait. Microcontrollers manage sensor clusters monitoring torso and lower extremities at 100Hz with hardware-triggered synchronization technology achieving low-latency sampling. Patients wear the lightweight array and walk naturally while recording movement data. Our AI system integrates three technologies: musculoskeletal simulation models using reinforcement learning; data-driven algorithms predicting spine curve by detecting compensatory relationship between spine and lower limb mechanics; and cloud-based continuous learning systems enhancing precision. This converts subjective observation into objective measurement, delivering rapid detection without radiation.

Our design process employed a three-phase methodology ensuring engineering rigor and clinical utility for adolescent idiopathic scoliosis detection. Phase I established system specifications including IMU configuration, anatomical placement, sampling frequencies, and communication standards. Engineering challenges centered on achieving synchronization across the multi-channel array through hardware-triggered timing and load distribution. Iterative PCB designs and firmware updates enhanced sampling accuracy and reliability. Phase II focused on clinical integration and user optimization. Multiple sensor deployment configurations were evaluated to minimize setup time while maintaining signal fidelity. Hardware components including cable management, housings, and attachment systems underwent refinements based on practitioner input. Phase III involved validation testing with over 200 adolescent participants. The system acquired synchronized, high-resolution motion data during postural and gait evaluations. Clinical specialists confirmed the system's capability to identify spinal asymmetries and movement patterns. User feedback informed final modifications, yielding a portable, clinically-deployable platform for early detection and monitoring.

ScolioDetect offers a breakthrough in non-invasive scoliosis detection by combining high accuracy, simplicity, and real-world usability. Unlike traditional methods that rely on static posture, it detects spinal asymmetry through dynamic gait analysis. Common tools like the Adam Test and surface topography suffer from low sensitivity, high false positives, operator dependence, and inability to assess movement. ScolioDetect uses synchronized IMUs to capture natural walking data, achieving a 3° mean absolute error—representing nearly twofold improvement over existing non-invasive diagnostic approaches. Assessments take under 3 minutes, require no specialists, and involve no radiation. Validated on over 200 adolescents, the system provides superior early detection of spine curve. Its lightweight, portable design supports use in schools and community settings, enabling large-scale access and timely intervention before deformities worsen.

免责申明：本文中部分图文内容转载自jamesdysonaward，相关版权归原作者及出处所有。若涉及post plant侵权，请及时联系我们，我们将第一时间删除处理。