A method for identifying transparent casts and pathological casts in urine sediment

Abstract

The present invention proposes a method for identifying transparent casts and pathological casts in urinary sediment, 1: constructing a data set; 2: preprocessing of the data set; 3: constructing Mobile_Urinet urine formed component recognition network; 4: setting the used The network parameters are as follows; 5: Training is based on the Mobile_Urinet urine formed component recognition model described in step 3; 6: Use the urine sediment image test set obtained in steps 1 and 2 to test the obtained model, and obtain the identification of cast components in the test set results and overall accuracy. Aiming at the characteristics that the cast images are difficult to classify, the present invention adopts data expansion and high-low hat transformation to enhance the contrast, and uses a lightweight network structure to complete the task. The recognition speed is high while the accuracy is high, and the obtained model parameters are few. It is easy and deployable, and has broad application prospects in clinical urine sediment identification testing.

Description

technical field [0001] The invention belongs to the field of medical image processing, in particular to a method for identifying transparent casts and pathological casts in urine sediment. Background technique [0002] Urine sediment refers to the urine formed components detected after the urine is concentrated, centrifuged, and precipitated. Urine examination is also one of the earliest medical examination items. Since the invention of the microscope, urine visible component inspection has been an important part of urinalysis, but in the field of clinical testing, the automation of urine visible component inspection has been relatively slow. With the development of digital image processing technology and computer technology in recent years, various automatic equipment for urine visible component inspection has gradually appeared and put into use. The main reason for manual microscopy is that the accuracy of automated inspection results is not high and the cost is too high,...

AI Technical Summary

Problems solved by technology

The diversity of cast components also makes it difficult to classify and identify them. There are also various cast components such as mucus filaments, pseudocasts, and impurities in urine, which makes accurate identification of casts very difficult. It hinders the development of automatic identification equipment, so the present invention aims to propose related methods to solve the problem of classification and identification of these cast components, mainly for the classification and identification of transparent casts and pathological cast components.

Images