Objective To investigate the expression changes of histidine decarboxylase(HDC)and H+,K+-ATPase in gastric mucosa during the healing of experimental gastric ulcer in rats.Methods The ulcers were caused by applying acetic acid to the serosal surface of the anterior face of the rat gastric body.At different time points during ulcer healing,HDC and H+,K+-ATPase mRNA and protein expressions were studied by using reverse transcription-polymerase chain reaction and Western blot respectively.Results An ulcer crater developed on the anterior face of the gastric body on day 1 after the induction of ulcers,and the ulcer area was biggest on day 3.On day 12,most of the gastric ulcers had healed.Compared with the control group,the HDC and H+,K+-ATPase mRNA expression in the gastric mucosa of ulcer rats showed a decrease on day 1,and increased back to initial level on day 9.The protein expression of HDC and H+,K+-ATPase in gastric mucosa of ulcer rats decreased immediately on day 1,more on day 6,and returned to the initial levels on day 12.Conclusion The mRNA and protein expressions of HDC and H+,K+-ATPase decrease in the healing process of gastric ulcers,resulting in accelerated ulcer healing through inhibiting gastric acid secretion.

Pulse recognition is an important part of the objectification and intelligence of TCM.This non-invasive and fast diagnostic method has great clinical value,however,data imbalance and cumbersome feature extraction are still challenging problems.The feature vectors were extracted from the one-dimensional pulse signal obtained after the Butterworth bandpass filter using the tsfresh library.And 9 columns of medical auxiliary features selected by exploratory data analysis were added.The feature filtering is performed jointly to derive 21 columns of feature vectors,which are used as input to the weighted soft voting fusion model.The data imbalance problem is solved by Borderline SMOTE algorithm.Construct a weighted soft-voting fusion model based on four types of machine learning:XGBoost,RF,LGBM,and GBDT.Eventually,the models will output specific pulse categories and demonstrate the performance by evaluating the metrics accuracy,precision,recall and F1 score.The experimental results show that the screened 21 feature vectors for a total of six types of pulse signal test sets achieve an accuracy of 90.04%in the five-fold cross-validation and take only 65.9466 seconds.It can provide a more accurate and intelligent auxiliary reference for pulse signal recognition,with lower operational complexity and higher accuracy compared to commonly used pulse recognition methods.The shorter training time also makes it more clinically useful in the recognition of multiple pulse signals.

The objectification of facial diagnosis has been developed through recent years and has become a multidisciplinary research topic.However,many studies are still limited to the adjustment of algorithms and the design of data collection environment and equipments,few studies focus on facial diagnosis zoning.The purpose of this paper is to discuss the problems in the current research literature on machine learning-based intelligent TCM facial diagnosis zoning to build a foundation for subsequent related research.The study uses bibliometric methods and text analysis to clarify and analyze the current intelligent TCM facial diagnosis zoning methods,which mainly include facial feature point-based,facial feature block-based and complete face-based method;then by analyzing the influencing factors of facial diagnosis zoning research and summarizing the common machine learning algorithms,the advantages and disadvantages of different machine learning algorithms and the corresponding common facial diagnosis zoning methods are obtained;Finally,we discuss three aspects of the current phase of facial diagnosis zoning research:dataset construction,advantages of deep learning,and embodiment of facial diagnosis theory.