Filamentous fungus form detection method and device, equipment and medium

Abstract

The invention provides a filamentous fungus form detection method and device, equipment and a medium. The method comprises the following steps: establishing a microscopic image database of various filamentous fungi; preprocessing the microscopic images of the filamentous fungi through a high information area evaluation algorithm; constructing a neural network model based on ResNet, and performing feature extraction on the preprocessed microscopic images serving as test samples; adjusting the neural network model; and preprocessing a to-be-detected image with the filamentous fungus structure characteristics, and inputting the preprocessed to-be-detected image into the trained neural network model to obtain the strain category and morphological description of filamentous fungi in the to-be-detected image. By introducing artificial intelligence technologies such as computer image understanding and machine learning, the detection rate of filamentous fungi can be greatly improved, and the machine can obtain the image identification capability similar to or even higher than that of people.

Description

technical field [0001] The invention relates to the fields of medical detection and image processing, in particular to a detection method, device, equipment and medium for the morphology of filamentous fungi. Background technique [0002] Filamentous fungi, namely brown rot fungi, are filamentous, non-photosynthetic, heterotrophic eukaryotic microorganisms, with well-developed hyphae, which do not produce large fleshy fruiting body structures. The cell wall of chitin is coated, generally colorless and transparent, and the width is mostly 3-10um, which is several to dozens of times the width of bacteria. The cells in the hyphae contain a large number of eukaryotic organelles. The cytoplasmic components in the hyphae tend to concentrate toward the growth point, the older part of the hyphae has a large number of vacuoles, and may be separated from the younger area by a transverse septum, the hyphae has bifurcations, and the branched hyphae interact with each other The staggere...

AI Technical Summary

Problems solved by technology

However, there are several problems in microscopic inspection in clinical microbiology departments: 1. The labor intensity is high, and manual inspection is difficult to observe every field of view under the microscope without omission, and it is easy to cause missed diagnosis after visual fatigue; Higher skill requirements, requiring experienced physicians to test

[0005] Therefore, due to the lack of fungal identification, identification and diagnosis capabilities in the field and grass-roots hospitals, the channels for requesting consultation support from large hospitals are limited or not timely, and the existing diagnostic techniques are lacking. Fungal serological techniques suitable for the shortage of personnel and equipment in grass-roots hospitals need to be developed. and other problems, through the establishment of clinical fungal image database and expert knowledge base, and the introduction of artificial intelligence technology, and then according to image recognition methods such as CNN convolutional neural network, a system for rapid morphological detection of on-site fungi is established to solve the above problems

Images