Drug Evaluation via Deep Learning Cell Image Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional drug evaluation methods, such as MTT, MTS, and CCK assays, are time-consuming and costly, require separate assays for cell viability determination, and involve harmful tetrazolium-based substances, limiting the ability to perform repeated measurements without damaging cells.
Innovation Solution
A drug evaluation method using an image processing-based deep learning algorithm that predicts molar concentrations from cell images, allowing for the calculation of IC50/EC50 values without additional treatments, excluding harmful substances, and enabling continued cell culture for repeated measurements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional MTT/MTS/CCK assays are used to evaluate drug efficacy and toxicity, then IC50/EC50 values can be obtained through color reaction measurements, but the process requires harmful tetrazolium-based substances, causes cell damage, and prevents repeated measurements
Solution Approach 1:
The patent extracts and removes the harmful tetrazolium-based substances from the assay process. Instead of using MTT/MTS/CCK reagents that cause cell damage, the invention uses alternative colorimetric methods that do not require these toxic substances, thereby eliminating the harmful effects while maintaining the ability to measure cell viability and calculate IC50/EC50 values
Solution Approach 2:
The patent creates a copy or alternative measurement approach that does not damage cells. By using different colorimetric reactions that are non-toxic to cells, the method produces equivalent measurement data (cell viability, IC50/EC50 values) without the need for harmful substances, allowing the same cells to be used for repeated measurements
2Measurement precision
If conventional assays with separate cell viability determination are performed, then IC50/EC50 values can be calculated, but the process requires multiple separate assay steps, increasing time and costs
Solution Approach 1:
The patent merges the drug efficacy evaluation and cell viability determination into a single integrated assay process. The colorimetric method simultaneously provides both the drug response data and cell viability information needed to calculate IC50/EC50 values, eliminating the need for separate viability assays and reducing total experiment time
Solution Approach 2:
The patent develops a universal assay method that serves multiple functions: it evaluates drug efficacy, determines cell viability, and calculates IC50/EC50 values all in one process. This multi-functional approach replaces the need for multiple specialized assays, reducing both time and computational costs
3Measurement precision
If cells are treated with tetrazolium-based dyes for color reaction, then drug effects can be measured, but the cells cannot be used for additional culture or subsequent tests
Solution Approach 1:
The patent converts the harmful effect of tetrazolium substances into a beneficial alternative approach. By using non-toxic colorimetric reagents that do not damage cells, the method maintains cell viability after measurement, allowing the same cells to be reused for additional culture, passaging, or subsequent experiments while still providing accurate drug effect measurements
Data Source
AI summary
Disclosed are a drug evaluation method and apparatus. The drug evaluation method includes acquiring cell images in response to a drug to be evaluated, loading an image processing-based learning model trained so as to predict molar concentrations, outputting molar concentrations of cells corresponding to the cell images by applying the learning model to the acquired cell images, and calculating a drug evaluation value used to evaluate effects of the drug on the cells based on the molar concentrations of the cells corresponding to the cell images output through the learning model.