Algorithm for automatically recognizing fetal heart rate baseline

Abstract

The invention discloses an algorithm for automatically recognizing a fetal heart rate baseline. The algorithm comprises the following steps: S1, collecting fetal heart rate signals; S2, performing curve optimization to the collected fetal heart rate signals; S3, performing feature extraction to the fetal heart rate signal curve after optimization, and clustering the extracted feature points into two feature parts through a clustering analysis algorithm; S4 distinguishing the two feature parts into a baseline part and a non-baseline part according to the distinguishing conditions; and S5, performing smooth filtering to the recognized baseline part to obtain the final estimated fetal heart rate baseline. The algorithm for automatically recognizing the fetal heart rate baseline can determine the fetal heart rate baseline more accurately by combining the fetal movement information particularly under the condition of close results of clustering, provide more intelligent and more reliable fetal heart monitoring results for clinic, and can be widely applied to the fields of biomedical signal processing, artificial intelligence and electronic fetal monitoring.

Description

technical field [0001] The invention relates to the fields of biomedical signal processing, artificial intelligence and electronic fetal monitoring, in particular to an automatic recognition algorithm of fetal heart rate baseline. Background technique [0002] Fetal heart rate monitoring is an important part of fetal electronic monitoring. The change of fetal heart rate is the result of the adjustment of fetal central nervous system. Therefore, fetal heart rate monitoring reflects the functional state of fetal central nervous system. The fetal heart rate curve is a record of fetal heart rate changes over a period of time, which contains certain pattern characteristics, such as baseline, acceleration, deceleration and fine variation (including long variation and short variation), through the reading and identification of these characteristics, obstetrics Medical staff can grasp the status of fetal development in utero in time and make clinical decisions. However, a large nu...

AI Technical Summary

Problems solved by technology

[0003] In terms of automatic recognition of the fetal heart rate baseline, the existing algorithm mainly has the following problems: 1. For the situation of continuous acceleration or continuous deceleration, the baseline obtained by the existing algorithm will be affected by the continuous acceleration or deceleration component and appear to increase or decrease. 2. For the case where the fetal heart rate acceleration or deceleration time is long, and the baseline part of the fetal heart rate curve is similar to the acceleration and deceleration part in time distribution, the existing algorithm may wrongly estimate the position of the baseline, or can only get A baseline that runs right through the middle of the FHR signal, which is clinically meaningless

Images