Thyroid Nodule Classifier Algorithm for Malignancy Diagnosis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for distinguishing benign and malignant thyroid nodules are not entirely effective, leading to unnecessary surgeries, as they often classify nodules as indeterminate, and existing gene-based classifiers lack sufficient validation in European populations.
Innovation Solution
A method using a classifier algorithm that evaluates the expression levels of specific genes (DCSTAMP, NOD1, SLC26A7, EMP2, EGR1, MPZL2, ITGA2, MET, SLPI, and TPO) in thyroid nodule samples, combined with clinical features, to accurately differentiate between benign and malignant nodules with high accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If fine-needle aspiration biopsy (FNAB) is used for diagnosis, then the diagnostic procedure is minimally invasive and easy to perform, but the ability to assess follicular cell arrangement and invasive features is lost, leading to indeterminate results
Solution Approach 1:
The patent introduces molecular markers (gene expressions, mutations, and microRNA profiles) as intermediary indicators that bridge the gap between minimally invasive FNAB sampling and accurate malignancy assessment. These molecular markers serve as mediators that provide additional diagnostic information without requiring more invasive procedures, thus maintaining ease of operation while improving measurement precision.
Solution Approach 2:
The patent replaces the mechanical/cytological assessment limitations of FNAB with molecular biological analysis. Instead of relying solely on visual evaluation of cell morphology and arrangement (mechanical observation), the invention uses molecular markers (gene expressions, mutations, microRNA profiles) to substitute and enhance the diagnostic capability, enabling accurate distinction between benign and malignant nodules even with limited cellular material.
2Measurement precision
If molecular markers are used to improve diagnostic accuracy, then the ability to distinguish benign from malignant nodules improves, but the complexity of the diagnostic procedure increases
Solution Approach 1:
The patent segments the molecular diagnostic panel into distinct, manageable components: gene expression analysis (e.g., HNF1A, PAX8, TTF1), mutation detection (e.g., BRAF V600E, RAS mutations), and microRNA profiling. This segmentation allows the complex diagnostic information to be processed in modular fashion, making the overall procedure more manageable and interpretable while maintaining high diagnostic accuracy.
Solution Approach 2:
The patent develops a universal molecular marker panel that can be applied across different thyroid nodule cases regardless of cytological category (B3, B4, or B5). This multi-functional approach allows the same set of molecular markers to provide diagnostic value across various clinical scenarios, reducing the need for multiple different test protocols and thereby managing complexity while improving accuracy.
3Measurement precision
If a larger number of molecular markers are analyzed, then the diagnostic accuracy increases, but the cost and time required for the test increase
Solution Approach 1:
The patent performs preliminary analysis by first categorizing the nodule based on cytological features (Bethesda system classification), then selectively applying molecular marker analysis based on the pre-test probability of malignancy. This preliminary stratification allows the system to focus molecular testing on cases where it will provide the most value, reducing overall diagnostic time while maintaining high accuracy for indeterminate cases.
Solution Approach 2:
The patent optimizes the molecular marker panel by selecting a specific set of markers with high discriminatory power (e.g., 5-10 key genes, specific mutations, and microRNAs) rather than analyzing all possible molecular markers. This parameter optimization balances diagnostic accuracy with testing time and cost, identifying the minimum effective set of markers needed for reliable classification.
4Loss of information
If existing gene classifiers are used, then some diagnostic information is obtained, but their performance is insufficient in European populations due to lack of validation
Solution Approach 1:
The patent creates a dynamic, adaptable molecular marker panel that can be adjusted and validated for different population characteristics. Rather than using a fixed classifier developed in one population, the invention allows for optimization of marker selection and threshold values based on European population data, ensuring the diagnostic system remains reliable and accurate across different genetic backgrounds and epidemiological contexts.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method of distinguishing benign and malignant thyroid nodules in the nodule biopsy sample, classified according to the Bethesda system, including the steps of RNA isolation and the determination of expression level of a normalized gene set, including AMOT, DCSTAMP, NOD1, SLC26A7, EMP2, EGR1, MPZL2, ITGA2, MET, SLPI, TPO. To differentiate thyroid nodules, a classifier algorithm is used, which is trained on gene expression data correlated with the nature of the thyroid nodule. In addition, the probability of malignancy of the test sample is determined.