Medical Device Step Counting via Orthogonal Acceleration Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current medical devices cannot objectively quantify a patient's physical activity level, which is crucial for assessing the quality of life and treatment efficacy for patients with chronic conditions, due to the complexity and power requirements of existing step counting algorithms and devices.
Innovation Solution
A medical device with an accelerometer unit and processor that processes 2-dimensional acceleration data along orthogonal axes to determine step count using a bandpass filter, summing, smoothing, and identifying local extrema, allowing for accurate step counting within limited power and processing constraints.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing step counting algorithms are used, then accurate step counting is achieved, but device complexity and power consumption increase
Solution Approach 1:
The patent extracts and implements only the essential components needed for step counting within the medical device, rather than incorporating complete fitness tracker algorithms. The solution uses a simplified algorithm that processes acceleration data through bandpass filtering, full-wave rectification, and peak detection to identify steps, removing unnecessary complexity while maintaining accuracy for medical applications.
Solution Approach 2:
The patent adjusts algorithm parameters specifically for medical device constraints, using a bandpass filter with cutoff frequencies optimized for human walking patterns (0.5-10 Hz), and tuning the peak detection thresholds to distinguish genuine steps from noise. These parameter optimizations enable accurate step counting with reduced computational requirements suitable for implantable devices.
2Measurement precision
If existing step counting algorithms are used, then accurate step counting is achieved, but power consumption increases
Solution Approach 1:
The patent removes computationally intensive processing stages from standard step counting algorithms, retaining only the essential operations (bandpass filtering, rectification, peak detection) that can be executed efficiently on low-power microcontrollers. This extraction of core functionality reduces CPU utilization and power consumption while preserving step counting accuracy.
Solution Approach 2:
The patent implements periodic processing where the accelerometer is activated only during expected walking periods rather than continuous operation. The system uses motion detection thresholds to trigger processing intervals, reducing overall power consumption while maintaining accurate step counting during active periods.
3Measurement precision
If accelerometer data is processed continuously, then accurate physical activity measurement is achieved, but processing capacity requirements increase
Solution Approach 1:
The patent applies partial processing by first applying a motion detection threshold to identify periods of significant movement, then applying the full step counting algorithm only during these periods. This two-stage approach processes only the necessary data portions, reducing overall computational load while maintaining measurement accuracy for all physical activity levels.
Solution Approach 2:
The patent performs preliminary bandpass filtering and full-wave rectification of acceleration data before peak detection. These preprocessing steps transform the raw acceleration signal into a form that makes step identification more straightforward, reducing the computational complexity of the subsequent peak detection algorithm and overall processing requirements.
Data Source
AI summary
A medical device for a patient's body includes an accelerometer unit and a processor which are electrically interconnected. The patient's body has two orthogonal axes (YP, ZP). The first orthogonal axis (YP) is given by the craniocaudal or similar axis of the patient's body and the second orthogonal axis (ZP) is given by the dorsoventral or similar axis of the patient's body. The accelerometer unit is configured to determine at least 2-dimensional acceleration data (YD-data, ZD-data) over time along the first orthogonal axis (YP) and the second orthogonal axis (ZP).There is also a respective method for determining the number of steps made by a patient using a medical device, a computer program product and a computer readable data carrier.