A pet/ct automatic lung cancer diagnosis and classification system based on feature fitting and its construction method

Abstract

The invention discloses a PET / CT automatic lung cancer diagnosis and classification system and construction method based on feature fitting. The system includes a feature extraction unit, which is used for extracting and obtaining the first feature of the tumor according to the PET / CT image; the feature fitting unit, The PET / CT image is used for fitting the features in the pathological image corresponding to the PET / CT image to obtain the second feature of the tumor; the lung cancer diagnosis and classification unit is used for combining the first feature and the second feature to perform lung cancer diagnosis and classification. By using the PET / CT features to fit the matching pathological features, the present invention improves the diagnostic accuracy obtained by simply using the PET / CT image information features, and is beneficial to the popularization and application of the existing intelligent diagnosis based only on imaging. Improve the diagnostic efficiency of clinicians. Through the present invention, the condition of the tumor can be better understood before the patient undergoes surgery, so that the doctor can be better assisted to provide the patient with a precise treatment plan, and the trauma of the patient can be reduced.

Description

technical field [0001] The invention relates to the field of medical imaging, in particular to a PET / CT automatic lung cancer diagnosis and classification system based on feature fitting and a construction method. Background technique [0002] PET / CT is of great value in modern clinical medicine in the diagnosis of benign and malignant pulmonary nodules, the staging of lung cancer and the evaluation of lung cancer after treatment. 18F-FDG (18 fluorine-labeled glucose analog) is currently the most widely used tracer in PET / CT scans. It can maintain cellular energy supply by increasing glucose uptake and glycolysis according to the abnormal proliferation of malignant tumors. Different types of tumors show differences in the degree of glucose uptake on glucose metabolism imaging images. The tracer decays and annihilates in the patient, producing a pair of 511keV gamma photons emitting about 180° in opposite directions. The detector picks up where and when these gamma photons r...

AI Technical Summary

Problems solved by technology

Although existing PET / CT-based automatic lung cancer diagnostic classification models have achieved good classification accuracy, there is still a long way to go for clinical requirements.

For the existing pathological-based automatic lung cancer diagnosis classification model, although high classification accuracy can be obtained, the diagnosis result is slow and traumatic, which is a disadvantage for the rapid diagnosis of lung cancer screening

Images