Pixelated Monolithic PPG Sensor With Spiking Neural Noise Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing monolithic photoplethysmographic (PPG) sensors face challenges in obtaining spatial information of optical signals due to varying light inputs from PDs, requiring complex signal processing and high power consumption, and are unable to effectively separate noise from ambient light.
Innovation Solution
A pixelated monolithic PPG sensor with a spiking neural network structure that generates spike signals based on spatial information from each pixelated light sensing neuron, combining outputs in a column or row direction without additional adders and using a synchronizer to convert spike signals into digital data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a plurality of PD outputs is added by an analog circuit or a plurality of LDC outputs is added by a digital adder to generate a sensor output, then the signal-to-noise ratio of the PPG signal is maximized, but the device complexity and power consumption increase
Solution Approach 1:
Multiple photodiode outputs are combined into a single integrated output node through direct electrical connection, eliminating the need for separate adder circuits. The patent integrates signals from multiple PDs (e.g., PD1-PD4) directly to a common output node, achieving signal combination without additional circuitry, thus reducing device complexity while maintaining signal-to-noise ratio optimization
Solution Approach 2:
The photodiode array structure itself performs the summation function through its inherent electrical connection topology. The patent designs the pixel array such that multiple PDs naturally sum their outputs at a shared node, making the combining operation self-service rather than requiring external adder circuits, thereby reducing overall system complexity
2Loss of information
If the CMOS image sensor technique is used to individually extract output of each PD, then spatial information can be obtained, but a lot of power is necessary and the circuit implementation is complex
Solution Approach 1:
The patent extracts only the essential spatial information needed for PPG measurements by maintaining individual PD output nodes that preserve spatial distribution, while removing the power-intensive CMOS image sensor processing circuitry. The spatial information is preserved in the differential signal configuration without requiring full image sensor functionality, thus reducing power consumption while retaining necessary spatial data
Solution Approach 2:
The patent applies different processing approaches to different parts of the sensor array. Individual PDs maintain their spatial characteristics locally through differential connections, while the overall system uses simplified combining logic rather than full image sensor processing. This local quality approach preserves spatial information where needed while avoiding global power-intensive processing
3Loss of information
If the CMOS image sensor technique is used to individually extract output of each PD, then spatial information can be obtained, but the device complexity increases
Solution Approach 1:
The patent merges the functions of multiple PD outputs into a simplified differential signal structure that preserves spatial information. By combining multiple PD outputs through differential pairs and using a single integrated output node, the patent reduces circuit complexity while maintaining the spatial distribution information needed for PPG measurements, avoiding the need for complex CMOS image sensor circuitry
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The sensor effectively removes ambient light noise through simple matrix operations, preserving spatial information and reducing power consumption by eliminating the need for complex signal processing techniques.
Implementation Method 1
A main principle thereof is to measure changes in optical signals according to the change in blood flow so that the PPG sensor includes a photodiode (PD) which converts an optical signal into a current signal
Data Source
AI summary
A pixelated monolithic photoplethysmographic (PPG) sensor with a spiking neural network structure and a driving method thereof are disclosed. The pixelated monolithic photoplethysmographic (PPG) sensor according to an exemplary embodiment of the present disclosure includes a plurality of input neurons which generates spike signals and is disposed in at least one of a row direction and a column direction, a plurality of output neurons which adds the spike signals applied from the plurality of input neurons based on the column direction or the row direction and generates an output spike signal when the added result is equal to or higher than a predetermined reference value, and a synchronizer which converts each of the output spike signals into a digital signal.