Neural-network-based bone status evaluation model building method

Abstract

The invention discloses a neural-network-based bone status evaluation model building method. The method comprises the steps that evaluation data of multiple groups of users with different bone statuesis obtained, wherein each group of the evaluation data comprises image data and text data; multiple groups of the evaluation data are preprocessed, multiple groups of standard evaluation data are obtained, and a training sample set is built; a bone status evaluation preset model is trained based on the training sample set and bone health standards, and parameters of the bone status evaluation preset model is updated, and the bone status evaluation model is obtained, wherein the bone status evaluation model is used for outputting current bone statuses of the users according to data input by the users. The invention further provides a neural-network-based bone status evaluation model building device.

Description

technical field [0001] The invention relates to the technical field of health state assessment, in particular to a method for constructing a neural network-based bone state assessment model. Background technique [0002] Bones are important, hard organs of the human body that serve multiple functions. These include support functions, which maintain body posture; hematopoietic functions, which produce blood cells through hematopoiesis; storage functions, which store important minerals in the body, such as calcium and phosphorus; and motor functions, which generate and transmit force to move the body. Bone health is closely related to an individual's daily life and quality of life. [0003] If bones are damaged, bone mass decreases, bone microarchitecture degrades, and metabolic bone diseases such as osteoporosis occur, resulting in decreased bone strength, and fractures are prone to occur during minor trauma or daily activities, that is, fragility fractures. This type of fr...

AI Technical Summary

Problems solved by technology

The evaluation methods in medical measurement only use bone density or simple influencing factors as the evaluation basis. More than half of the women with fractures do not meet the standard of osteoporosis, and the evaluation basis is single; the evaluation method based on questionnaires does not take into account Due to the interrelated effects of different risk factors, it is impossible to discover the potential risk of the combination of weakly correlated factors and the risk of osteoporosis, and only rely on manual assessment, which is highly subjective; assessment methods based on machine learning exist The evaluation model is relatively simple, the evaluation data is relatively single, and it is impossible to accurately fit the characteristics of the data

Images