Sample Reevaluation Scoring Using Process Probe Readouts

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Genotyping processes are vulnerable to both process and sample errors, making it difficult to predict whether a rerun of an inconclusive production run will succeed, leading to increased costs and turnaround time due to sample-related failures.

Innovation Solution

A machine learning-based system using a trained retry classifier that analyzes readouts from process probes and call rates to generate a retry success confidence score, determining whether a sample reevaluation will produce conclusive results.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual review of test samples is performed to identify false positives and false negatives, then measurement precision is improved, but loss of time increases due to manual effort required

Engineering Contradiction:
Improveaccuracy of test result interpretationVSAvoidtime for manual review
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent replaces the manual mechanical review process with an automated machine learning model that analyzes test samples, process indicators, and test conditions to predict interpretation accuracy. This substitution eliminates manual effort while maintaining or improving measurement precision through systematic automated evaluation.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent creates a virtual copy of the manual review process through a machine learning model that replicates and enhances human interpretation capabilities. The model learns from historical review data and applies that knowledge to predict outcomes for new samples, effectively copying and improving upon manual review performance without the time cost.

Inventive Principle:
Principle #26Copying

2Measurement precision

If additional process indicators are collected and analyzed to improve prediction accuracy, then measurement precision is improved, but device complexity increases

Engineering Contradiction:
Improveprediction accuracy of success rateVSAvoidcomplexity of data collection and analysis system
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent implements a multi-functional machine learning model that simultaneously processes diverse data types including process indicators, test conditions, sample characteristics, and historical review outcomes. This universal system handles multiple functions (data collection, preprocessing, feature extraction, prediction) within a single integrated framework, managing complexity through consolidation rather than proliferation of separate systems.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Solution Approach 2:

The patent transforms raw process indicators and test conditions into standardized features through preprocessing and normalization. By changing the parameter representation and scaling data appropriately, the system manages complexity in data handling while extracting meaningful patterns that improve prediction accuracy.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP4158636B1Machine learning-based analysis of process indicators to predict sample reevaluation success
Publication Date: 2026.05.13 ILLUMINA INC
  • EP4158636B1 patent drawingFigure 1
  • EP4158636B1 patent drawingFigure 2
  • EP4158636B1 patent drawingFigure 3

AI summary

The technology disclosed relates to systems and methods for scoring whether to reevaluate a sample after one or more inconclusive sample evaluation runs. The scores can be based on combination of call rates and readouts from process probes which generate signals indicative of processing parameters at successive stages of sample processing. The system can include a classifier trained to predict whether an additional sample evaluation run will produce conclusive results for the sample. The system can use a plurality of readouts from probes grouped into sample-dependent and sample-independent process probes. The system can use a plurality of readouts of radiant signals from probes grouped according to three stages of the sample evaluation run.