Feature Computational Module for Biological Signal Encoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods struggle to efficiently encode biological sample signals into informative features for health condition detection, due to high data volume and uncertainty around informative metrics and regions.
Innovation Solution
The feature computational module processes biological sample signals by computing metrics based on marker information within specified windows in health-condition informative regions, using functions that analyze patterns and occurrences of marker information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If the number of features is increased to preserve relevant information, then information loss is reduced, but computational model complexity increases
Solution Approach 1:
The analyte sequence is divided into multiple windows of consecutive sites, where metrics are computed for each window independently. This segmentation allows the system to process large amounts of biological data in manageable chunks, extracting relevant information without requiring the model to handle the entire sequence at once, thus reducing computational complexity while preserving local patterns.
Solution Approach 2:
The patent transforms the one-dimensional sequence data into multi-dimensional feature space by computing multiple metrics (count, pattern, entropy) across multiple windows. This dimensional transformation allows the system to capture complex biological patterns through structured numerical representations, enabling efficient machine learning processing while retaining information density.
2Device complexity
If the volume of data is reduced into fewer features, then computational complexity decreases, but relevant information may be lost
Solution Approach 1:
The patent applies multiple mathematical transformations to convert raw marker information into different metric types (count metrics, pattern metrics, entropy metrics). Each metric type captures different aspects of the biological data, allowing the system to compress data into fewer features while preserving diverse information dimensions through parameter transformation rather than simple reduction.
3Device complexity
If metrics average information over the entire region, then computational requirements are reduced, but local patterns are lost
Solution Approach 1:
Instead of computing a single metric across the entire analyte region, the patent divides the region into multiple overlapping or non-overlapping windows and computes metrics for each window separately. This local segmentation preserves spatial patterns and regional variations in the biological data, allowing the model to detect location-specific biomarkers while maintaining computational efficiency through localized processing.
Data Source
AI summary
A feature computational module encodes a signal generated by processing a biological sample, given one or more health-condition-informative regions related to an analyte, by using metrics based on marker information occurring within specified windows within a sequence of sites of interest within the health-condition informative regions related to the analyte. Each window has a specified position within a sequence of sites of interest in the health-condition informative region, and a specified size. The size is specified in terms of a number of consecutive sites of interest within the analyte. A metric is thus computed for a plurality of positions within the health-condition informative region. In each metric, a first function of respective marker information for an instance of an analyte for a window is used to compute a respective value for each instance of the analyte in the window. A second function of these respective values is computed to provide one or more values for the one or more metrics for the window.