Vaginal secretion detection device and method

Abstract

Embodiments of the invention disclose a vaginal secretion detection device and method. The detection device comprises a base and a control circuit board; a bar code scanning head, an upper card assembly and an incubation assembly are sequentially arranged on one side of the base; the other side of the base is sequentially provided with a test tube rack assembly, a consumable replacement assembly,a slide upward and downward moving assembly and a microscope electric platform; the upper card assembly is provided with a card storage box; a reagent card clip is arranged in the card storage box; afirst optical detection assembly and a second optical detection assembly are arranged above the incubation assembly; a reagent card recovery box is arranged on the outer side of the incubation assembly; a Y-direction driving assembly is further arranged on the base; a liquid shifting assembly is arranged on the Y-direction driving assembly; a microscope is arranged above the microscope electric platform; a slide recovery box is arranged on the outer side of the microscope electric platform; and a slide clip is arranged on the slide upward and downward moving assembly. According to the vaginalsecretion detection device and method, the detection speed and the recognition accuracy are improved; and manual feature extraction and image preprocessing are not needed, so that the misdiagnosis isreduced.

Description

technical field [0001] The invention relates to the technical field of biological and medical detection, in particular to a vaginal secretion detection device and method. Background technique [0002] The detection of vaginal secretions is generally divided into two parts, dry chemical detection and microscopic examination. Dry chemical detection is a semi-quantitative detection. Generally, when the dry chemical test results show that the patient's physical indicators are abnormal, a microscopic examination is performed on the patient to complete the confirmation of the test results. However, the current microscopic examination of vaginal secretions is mainly through manual detection. [0003] After the tester gets the patient sample, he rinses the sample. After rinsing, he makes a manual smear and puts it under the microscope for testing. The formed components in the film are classified and counted. Formed components mainly include: white blood cells, Candida, clue cell...

AI Technical Summary

Problems solved by technology

The method of manual microscopic examination not only requires manual smears, constantly switching the field of view to count, but also corresponding the results of microscopic examination with the results of dry chemical testing, and manual input, which is time-consuming and laborious, and the detection efficiency is low, and the counting and input process is prone to errors

[0004] At present, for microscope inspection, there are also automatic detection devices on the market, but the devices are complicated, the recognition rate of formed components is low, and it takes a long time

Moreover, the pictures taken are stored locally, and the detection and identification of the formed components of the sample is limited to the local area, without effective connection with the cloud platform

The low detection rate of instruments on the market is mainly due to the traditional classification method used. The traditional neural network classification method not only requires a lot of pre-processing of pictures, but also needs to artificially select some features when constructing and training the neural network. The selected features are not necessarily suitable for computer recognition, and are prone to overfitting

Moreover, since the detection of formed components is limited to the local area, the large number of samples cannot correct the detection model in a timely and effective manner, and the detection accuracy will not increase with the increase of detection samples.

Images