Forensic medicine examination automatic identification system and method based on deep learning

Abstract

The invention discloses a forensic medicine examination automatic identification system based on deep learning. The system is used for examining diatom and comprises a data acquisition module which isused for performing automatic image acquisition and image segmentation on a to-be-examined tissue sample so as to obtain a plurality of electron injection images corresponding to the tissue sample astraining data samples and output the training data samples; a data expansion module which is connected with the data acquisition module and is used for carrying out sample quantity amplification on the input training data samples based on a predetermined sample amplification method; a model iterative training module which is connected with the training data expansion module and is used for carrying out iterative training by adopting a deep learning method based on the training data sample and/or an externally input training sample to form a deep learning model; and a recognition module whichis connected with the model iterative training module and is used for carrying out diatom component inspection on the input electron injection picture based on the deep learning model, so that the full automation of the diatom inspection is realized, and the diatom inspection accuracy is improved.

Description

technical field [0001] The invention relates to the technical field of forensic medical examination, in particular to an automatic identification system and identification method for forensic medical examination of diatoms based on deep learning. Background technique [0002] In the practice of forensic science, diatom detection is the main technical means for diagnosing drowning, and its positive test results can even be used as the gold standard for diagnosing drowning. As a common microorganism in fresh water or sea water, diatoms can be inhaled through the respiratory tract and circulated in different tissues and organs during human drowning, such as the lungs, liver, spleen, kidneys and bone marrow. Because the diatom shell has strong acid resistance, it is not easy to be completely corroded when it remains in the human body. Therefore, the conventional diatom test mainly uses strong acid digestion method to remove the organic matter of the tissue to form a digestive su...

AI Technical Summary

Problems solved by technology

[0003] However, the above-mentioned conventional diatom inspection methods still rely on manual visual observation and identification, and the inspection accuracy largely depends on personal inspection experience, and the inspection accuracy cannot be effectively guaranteed

At the same time, due to the small size of diatoms, they are usually sparsely distributed in tissue samples, and inspectors need to spend a lot of time to find diatom components through manual observation, and the test results often appear false positive or false negative

In addition, due to the strong acidity of tissue sections, it is often difficult to preserve them for a long time, which will have an impact and inconvenience on the retrieval and review of future cases

Images