Artificial intelligence deep learning method for correcting continuous blood glucose monitoring method

Abstract

The invention belongs to the technical field of blood glucose monitoring, and particularly discloses anartificial intelligence deep learning method for correctinga continuous blood glucose monitoringmethod.The method comprises the specific steps of (1) collecting accurate blood sugar values of a single patient continuously changing for a period of time; (2) performing multiple times ofa fingertipblood sampling method to measure the accurate blood sugar values; (3) correcting blood glucose values measured by a continuous blood glucose monitor with the accurate blood glucose values; (4) performing deep artificial learning correction on comparative data; (5) usingan artificial Intelligent automatic encoder technology to correct the data measured by the continuous blood glucose monitor; (6)summarizinga numerical difference pattern recognized by the artificial intelligence method into a blood glucose correction mode; (7) afterartificial intelligence training is completed, the sustainableblood glucose monitoring is conducted, and the recorded blood glucose values are corrected by using the numerical correction methodobtained by the artificial intelligence training. The method can provide accurate and reliable real-time blood glucose data for diabetic patients to better manage their physical condition.

Description

technical field [0001] The invention belongs to the technical field of blood sugar monitoring, and in particular relates to a monitoring method for correcting continuous blood sugar by an artificial intelligence deep learning method. Background technique [0002] Diabetes has become one of the chronic diseases that seriously threaten human health. According to the statistics of the World Health Organization, the number of diabetic patients is increasing year by year. The number of diabetic patients in all countries in the world has reached as many as 200 million. It has exceeded 100 million, and the incidence of diabetes is as high as 10%, with an average annual increase of 5.5 million cases, an increase of 15,000 cases per day, an increase of 600 cases per hour, and an increase of 10 cases per minute. What is worse is that behind the increasing number of diabetic patients, there are still There are 264 million people with pre-diabetes. Their blood sugar is not within the no...

AI Technical Summary

Problems solved by technology

[0008] The third method is the colorimetric method of blood glucose detection with blood glucose test strips: this method can measure blood glucose concentration without a detector, but it is a semi-quantitative test method, and the test strips can only be used at 18-35°C

Because machine learning requires a large amount of data collection to achieve a good learning effect, and the collected data must have certain accuracy, but because non-invasive blood glucose meters can only measure at a certain point in time, they cannot continuously collect data , which brings a large workload to data collection, the amount of data is small and the data is not accurate enough, so there is no good learning effect

In addition, in addition, such a method cannot achieve personalized effects. In order to make up for the above-mentioned shortcomings of insufficient data, training data will be obtained from many diabetic patients when collecting data, and then use this result to test other patients. Obviously, due to the physical differences between people, the detection effect is not very good, and such a method obviously greatly increases the workload, so it cannot be widely used

Images