Ct Feature Engineering for Local Infection Forecasting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing models for forecasting disease outbreak case counts at a local level do not effectively utilize Cycle Threshold (Ct) data from PCR tests, which correlates with viral load, limiting the accuracy of predictions.
Innovation Solution
Generating and incorporating Ct features into machine learning models, specifically using mean, smoothed mean, skewness, smoothed skewness, incidence rates, and effective reproduction rate curves derived from Ct data to improve case count forecasting performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If only binary positive/negative PCR results are used for forecasting, then the forecasting model is simpler to implement, but the accuracy of case count predictions deteriorates
Solution Approach 1:
The patent transforms the binary positive/negative PCR result into a continuous parameter by incorporating the Cycle Threshold (Ct) value. The Ct value provides a quantitative measure of viral load that varies continuously, allowing the forecasting model to capture gradient information about infection intensity rather than just binary states. This parameter transformation enables more nuanced predictions while maintaining model tractability.
Solution Approach 2:
The patent introduces Ct values as an intermediary variable that mediates between the binary PCR result and the case count forecast. The Ct value serves as a bridge that carries additional information about viral load magnitude, allowing the model to infer incident rates and reproduction rates more accurately without directly observing these unmeasured quantities.
2Measurement precision
If Ct data is incorporated into the forecasting model, then the accuracy of case count predictions improves, but the complexity of the model increases
Solution Approach 1:
The patent performs preliminary processing of Ct data by calculating derived metrics such as mean Ct, smoothed mean Ct, skewness, and smoothed skewness before feeding them into the forecasting model. This preprocessing step transforms raw Ct values into meaningful features that capture temporal trends and distribution characteristics, reducing the computational burden on the main forecasting algorithm while preserving predictive power.
Solution Approach 2:
The patent employs feedback mechanisms by using smoothed versions of Ct metrics (smoothed mean, smoothed skewness) that incorporate historical information. These smoothed values provide feedback about recent trends in viral load distributions, allowing the model to adapt to changing epidemic conditions while filtering out noise from individual measurements.
3Measurement precision
If Ct data is used to generate multiple features (mean, smoothed mean, skewness, smoothed skewness), then the forecasting accuracy improves, but the data processing requirements increase
Solution Approach 1:
The patent calculates a selective set of Ct features (mean, smoothed mean, skewness, smoothed skewness) rather than all possible statistical moments and transformations. This partial action approach focuses computational resources on the most informative features that have been shown to improve forecasting accuracy, avoiding the excessive computation that would result from calculating every possible Ct statistic.
Solution Approach 2:
The patent segments the Ct data processing into distinct computational stages: calculating basic statistics (mean, skewness), applying smoothing operations, and then feeding these processed features into the forecasting model. This segmentation allows each processing step to be optimized independently and enables parallel computation of different feature types, reducing overall processing time.
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
Enhances the accuracy of high-resolution case count forecasting at a local level by leveraging Ct data, providing early warnings for epidemic spread and improving logistical planning.
Implementation Method 1
A PCR test performs thermal cycling on a biological sample. The cycling amplifies DNA corresponding to a target sequence if that sequence is present in the sample.
Implementation Method 2
Because the PCR process amplifies DNA, the cycle at which a sequence is first detectable is, on average, inversely proportional to the amount of a given DNA sequence initially present in a given sample volume.
Data Source
AI summary
Methods for forecasting case counts for a future date in one or more geographic areas of persons infected by a disease is disclosed. The presence of the disease in a biological sample is testable by a polymerase chain reaction (PCR) test. A load of one or more pathogens associated with the disease correlates with a PCR cycle which indicates presence of the one or more pathogens, and is referred to as a threshold cycle (Ct). Data relevant to forecasting the case counts including Ct data and other data is received. The Ct data comprises Ct values from PCR tests of biological samples from persons within the one or more geographic areas. Arrays of feature data for processing by a trained machine learning model are generated, comprising Ct features and other features obtained from the data. A forecasted number of infected persons are generated by processing the arrays using machine learning.