Adaptive Bit Depth Compression for Biological Signal Storage
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Implantable medical devices (IMDs) face limitations in storing wide dynamic range and high-resolution biological signals like ECG/EGM due to limited memory, processing, and power resources, leading to inadequate resolution and potential clipping of signal features.
Innovation Solution
The use of an N-bit analog-to-digital converter (ADC) in IMDs to convert biological signals into N-bit data values, followed by compression using a compression curve with lookup tables (LUTs) to produce M-bit data values, which are then stored efficiently, allowing for later expansion to reproduce the original signal with high resolution and wide dynamic range.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If 8-bit sampled data is stored to conserve memory and power, then memory storage efficiency is improved, but signal resolution and dynamic range are degraded
Solution Approach 1:
The patent applies parameter changes by dynamically adjusting the bit depth of data storage based on signal characteristics. During low-amplitude signal segments (e.g., P-waves, T-waves), the system increases bit depth to capture fine details, while during high-amplitude segments (e.g., QRS complexes), it uses lower bit depth. This adaptive parameter adjustment resolves the contradiction by optimizing both memory efficiency and signal resolution according to actual signal needs.
Solution Approach 2:
The system implements dynamics through real-time adaptation of sampling parameters. The IMD continuously monitors signal amplitude and dynamically adjusts the number of bits used for storage, transitioning between different resolution modes based on signal conditions. This dynamic approach allows the system to maintain high resolution when needed while conserving memory during periods of lower signal complexity.
2Measurement precision
If gain of the sensing front end is increased to increase P-wave amplitude, then P-wave detection capability is improved, but R-wave amplitude increases to the point of clipping or saturation
Solution Approach 1:
The patent resolves this contradiction through parameter changes in the data representation domain rather than analog gain adjustment. By using variable bit depth and adaptive scaling, the system can digitally enhance the representation of low-amplitude P-waves without physically amplifying the signal to a level that would cause R-wave clipping. The parameter adjustment occurs in the digital processing stage, allowing independent optimization of different signal features.
Solution Approach 2:
The patent introduces an intermediary mechanism in the form of adaptive digital signal processing between the sensing front end and storage. This intermediary layer applies dynamic scaling and variable precision arithmetic to preserve P-wave details without requiring analog gain changes that would affect the entire signal range. The intermediary processing stage decouples the detection of low-amplitude features from the risk of high-amplitude clipping.
3Measurement precision
If N-bit ADC with N>8 is used to convert biological signals, then signal resolution is improved, but memory storage requirements and processing resources increase
Solution Approach 1:
The patent applies parameter changes by using variable bit depth storage instead of fixed N-bit storage. The system stores data with adaptive precision, using more bits only when signal complexity requires it, and fewer bits during simpler signal periods. This parameter adjustment significantly reduces average memory requirements while maintaining the capability to represent high-resolution signal features when present.
Solution Approach 2:
The patent implements discarding and recovering by selectively reducing precision in certain data segments while preserving full precision in critical segments. Less significant bits are discarded during low-complexity periods to save storage, and precision is recovered during high-complexity periods when signal details are more important. This selective discarding and recovering strategy optimizes the trade-off between storage efficiency and signal fidelity.
Data Source
AI summary
Described herein are methods, devices and systems for efficiently storing data for sensed biological signals. A sensed biological signal, or an amplitude and/or filtered version thereof, is provided to an N-bit ADC of an IMD to produce an N-bit data value indicative of an amplitude of the biological signal at a point in time. One of a plurality of chords of a compression curve is selected, based on a magnitude of the N-bit data value, and used to produce an M-bit data value, which is a compressed version of the N-bit data value, wherein M<N. The M-bit data value is stored as an M-bit data slice within memory of the IMD, and can be expanded to a reproduced N-bit data value after being uploaded to a non-implanted device or system.