Pan-cancer Detection via miRNA Liquid Biopsy and SVM Analysis
Solution Overview
Problem
Current cancer detection methods are invasive and lack sensitivity and specificity, often detecting tumors only after they have grown or metastasized, necessitating a non-invasive, early detection approach for pan-cancer.
Innovation Solution
An early detection and prediction method utilizing a miRNA expression profile database and supervised learning support vector machine (SVM) analysis of liquid biopsy samples to identify cancer through miRNA expression profiling in plasma, serum, or urine, involving data normalization, imputation, scaling, predictive modeling, and cross-validation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If biopsy or surgically obtained tissues are used to diagnose malignant tumors, then diagnostic accuracy is improved, but the detection method becomes invasive
Solution Approach 1:
The patent uses miRNA as an intermediary biomarker that can be detected in liquid biopsy samples (blood, plasma, serum, urine, saliva, or stool). These miRNAs are secreted from cancer cells into the liquid samples, serving as a non-invasive proxy for detecting cancer presence and type, thereby avoiding the need for invasive tissue biopsy while maintaining diagnostic accuracy
Solution Approach 2:
The patent creates a copy of cancer diagnostic information by detecting miRNA expression profiles in liquid samples that mirror the cancer cell's molecular signature. The miRNA expression profile serves as a molecular copy or fingerprint of the cancer, allowing diagnosis without directly examining the original cancer tissue
2Ease of operation
If tumor marker detection is used to determine cancer, then the detection process is simplified, but sensitivity and specificity deteriorate resulting in late detection
Solution Approach 1:
The patent changes the detection parameter from traditional protein-based tumor markers to miRNA expression profiles. By measuring the expression levels of multiple specific miRNAs simultaneously and comparing them against established thresholds and patterns, the method achieves both high sensitivity and specificity while maintaining operational simplicity through automated computational analysis
Solution Approach 2:
The patent uses a composite approach by analyzing multiple miRNA markers together rather than relying on a single tumor marker. The combination of multiple miRNA expression patterns creates a more robust diagnostic signature that improves both sensitivity and specificity compared to individual marker detection
3Loss of time
If traditional cancer screening methods are used, then early detection may be achieved, but the methods are invasive and have limited diagnostic rate
Solution Approach 1:
The patent creates a universal detection method that can identify multiple types of cancer through a single liquid biopsy test. The miRNA expression profiling approach is applicable across different cancer types and can be performed on various liquid samples (blood, urine, saliva), providing a multi-functional screening tool that improves convenience and diagnostic rate while enabling early detection
Data Source
AI summary
An early detection and prediction method of pan-cancer is provided, including the following steps. A microRNA expression profile database of cancer patient populations and healthy populations is established, and an early detection and prediction method of pan-cancer is developed based on the microRNA expression profile database. Next, a microRNA expression profile of a liquid biopsy sample of a subject is analyzed via the early detection and prediction method of pan-cancer, so as to determine whether the subject may already have cancer.