A joint detection method and system for abnormality of white blood cell scattergram
A combined detection and white blood cell technology, applied in the field of white blood cell detection, can solve the problems of low accuracy, manpower consumption, and high missed detection rate, and achieve the effect of reducing the missed detection rate, improving reliability, and good judgment effect
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0056] combine Figure 4 , the part inside the box is the abnormal part. The result given by the neural network model is negative, and the given confidence is 0.87, the result of the prior knowledge is positive, the given confidence is 0.66, and the weight of the prior knowledge is 2, then the final prior knowledge confidence is 0.66*2=1.32, which is higher than the confidence level of the neural network, the scatter plot is judged to be positive.
Embodiment 2
[0058] The neural network judges it as positive with a confidence level of 0.5, and the prior knowledge judges it as negative with a confidence level of 0.8. If the system uses a sensitivity value of 0.4 to prevent the positive data from being missed, then once there is a neural network or prior knowledge, either party will be Positive and the confidence level is higher than the sensitivity value (the neural network confidence level is 0.5 higher than the sensitivity value), the result is positive.
[0059] The present invention first collects a large amount of white blood cell scattergram data, and adds labels to the data according to the doctor's judgment. On the one hand, the pixel information in the image is counted, and according to the distribution of different color scatter points on the image, appropriate judgment conditions are set up for classification; on the other hand, the data is divided into training set, verification set and test set. The labeled data is used a...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 


