Introduction
Idiopathic normal pressure hydrocephalus (iNPH) is a reversible neurological condition characterized by gait disturbance, cognitive decline, and urinary incontinence. While permanent ventriculoperitoneal shunt surgery is the standard treatment, identifying which patients will benefit remains a clinical challenge. The External Lumbar Drain (ELD) serves as a temporary "test drive" to predict surgical success. Traditionally, improvements after ELD are assessed using subjective clinical scales or simple stopwatches, which are prone to observer bias. This study investigates the use of wearable inertial sensors to provide objective, high-resolution "digital biomarkers" for predicting surgical outcomes.
Methods
The study included 36 patients with suspected iNPH and 20 healthy controls. Patients underwent a 48-hour ELD procedure. Gait and balance were assessed before and after the drain using the APDM Movement Monitoring system.
The system utilizes "Opal" sensors containing 3-axis accelerometers, gyroscopes, and magnetometers. These were placed on the wrists, feet, sternum, and lumbar spine (L5). Participants performed two-minute walkway tests on a 7-meter path. Based on clinical improvement assessed by a multidisciplinary team, patients were categorized as "responders" (n=20) or "non-responders" (n=16).
Key Results and Insights
The study identified several critical differences between responders and non-responders:
- Significant Improvement in Responders: Following ELD, responders showed marked improvements in cadence, gait speed (increasing from an average of 0.46 m/s to 0.58 m/s), stride length, and foot strike angle ($P < 0.05$).
- Turning Efficiency: Difficulty in turning is a hallmark of iNPH. Sensors revealed that responders significantly reduced the number of steps required to turn, a change not observed in the non-responder group.
- Z-Score Thresholds: By normalizing data against healthy controls, researchers found that an absolute change in Z-score > 1 in five key parameters (cadence, velocity, double support, stride length, and turning) strongly correlated with clinical response, providing a potential quantitative cutoff for surgical candidacy.
- Stability in Non-Responders: Non-responders showed no significant change in gait parameters, suggesting that the ELD—and likely a permanent shunt—would be ineffective for this group.
- Patient Perception: Objective sensor data correlated with patient-reported improvements (Global Rating of Change scores), suggesting the sensors captured changes meaningful to the patients' quality of life.
Discussion and Conclusion
Wearable inertial sensors like the APDM system offer a sensitive, objective method for evaluating the ELD response in iNPH patients. While some parameters like "elevation at mid-swing" did not improve within the 48-hour window—possibly requiring longer recovery—the overall gait profile provided by the sensors clearly distinguished surgical candidates. This quantitative approach reduces observer bias and provides clinicians with a robust tool for identifying patients most likely to benefit from shunt surgery. Future research with larger cohorts is needed to validate these digital biomarkers as standard predictors for long-term surgical success.
Summary updated