Banu Isbilen Basok, Ipek Deveci Kocakoç, Veli Iyilikci, Selena Kantarmaci, M. Fidan
{"title":"利用深度学习模型预测凝血检验的消耗量","authors":"Banu Isbilen Basok, Ipek Deveci Kocakoç, Veli Iyilikci, Selena Kantarmaci, M. Fidan","doi":"10.5937/jomb0-40244","DOIUrl":null,"url":null,"abstract":"Background: Laboratory professionals aim to provide a reliable laboratory service using public resources efficiently while planning a test’s procurement. This intuitive approach is ineffective, as seen in the COVID-19 pandemic, where the dramatic changes in admissions (e.g. decreased patient admissions) and the purpose of testing (e.g. D-dimer) were experienced. A model based on objective data was developed that predicts the future test consumption of coagulation tests whose consumptions were highly variable during the pandemic. \nMethods: Between December 2018 and July 2021, monthly consumptions of coagulation tests (PTT, aPTT, D-dimer, fibrinogen), total-, inpatient-, outpatient-, emergency-, non-emergency -admission numbers were collected. The relationship between inputs and outputs was modeled with the external input nonlinear autoregressive artificial neural network (ANN) (NARX) using MATLAB. Monthly test consumptions between January-July 2021 were used to test the models’ prediction power. \nResults: According to the cointegration analysis, total-, emergency-, and non-emergency admission numbers plus the number of working days per month were included in the model. When aPTT and fibrinogen consumptions were estimated, it was possible to predict the other tests. Fifty months of data were used to predict the next six months, and the NARX prediction was the more robust approach for both tests. \n \nConclusions: The deep learning model gives better results than the intuitive approach in forecasting, even in the pandemic era, and it shows that more effective and efficient planning will be possible if ANN-supported decision mechanisms are used in forecasting tests’ consumptions in the procurement process.","PeriodicalId":16175,"journal":{"name":"Journal of Medical Biochemistry","volume":"74 14","pages":""},"PeriodicalIF":2.0000,"publicationDate":"2023-12-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Forecasting The Consumptions of Coagulation Tests Using A Deep Learning Model\",\"authors\":\"Banu Isbilen Basok, Ipek Deveci Kocakoç, Veli Iyilikci, Selena Kantarmaci, M. Fidan\",\"doi\":\"10.5937/jomb0-40244\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Background: Laboratory professionals aim to provide a reliable laboratory service using public resources efficiently while planning a test’s procurement. This intuitive approach is ineffective, as seen in the COVID-19 pandemic, where the dramatic changes in admissions (e.g. decreased patient admissions) and the purpose of testing (e.g. D-dimer) were experienced. A model based on objective data was developed that predicts the future test consumption of coagulation tests whose consumptions were highly variable during the pandemic. \\nMethods: Between December 2018 and July 2021, monthly consumptions of coagulation tests (PTT, aPTT, D-dimer, fibrinogen), total-, inpatient-, outpatient-, emergency-, non-emergency -admission numbers were collected. The relationship between inputs and outputs was modeled with the external input nonlinear autoregressive artificial neural network (ANN) (NARX) using MATLAB. Monthly test consumptions between January-July 2021 were used to test the models’ prediction power. \\nResults: According to the cointegration analysis, total-, emergency-, and non-emergency admission numbers plus the number of working days per month were included in the model. When aPTT and fibrinogen consumptions were estimated, it was possible to predict the other tests. Fifty months of data were used to predict the next six months, and the NARX prediction was the more robust approach for both tests. \\n \\nConclusions: The deep learning model gives better results than the intuitive approach in forecasting, even in the pandemic era, and it shows that more effective and efficient planning will be possible if ANN-supported decision mechanisms are used in forecasting tests’ consumptions in the procurement process.\",\"PeriodicalId\":16175,\"journal\":{\"name\":\"Journal of Medical Biochemistry\",\"volume\":\"74 14\",\"pages\":\"\"},\"PeriodicalIF\":2.0000,\"publicationDate\":\"2023-12-05\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Journal of Medical Biochemistry\",\"FirstCategoryId\":\"3\",\"ListUrlMain\":\"https://doi.org/10.5937/jomb0-40244\",\"RegionNum\":4,\"RegionCategory\":\"医学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q4\",\"JCRName\":\"BIOCHEMISTRY & MOLECULAR BIOLOGY\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Medical Biochemistry","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.5937/jomb0-40244","RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q4","JCRName":"BIOCHEMISTRY & MOLECULAR BIOLOGY","Score":null,"Total":0}
Forecasting The Consumptions of Coagulation Tests Using A Deep Learning Model
Background: Laboratory professionals aim to provide a reliable laboratory service using public resources efficiently while planning a test’s procurement. This intuitive approach is ineffective, as seen in the COVID-19 pandemic, where the dramatic changes in admissions (e.g. decreased patient admissions) and the purpose of testing (e.g. D-dimer) were experienced. A model based on objective data was developed that predicts the future test consumption of coagulation tests whose consumptions were highly variable during the pandemic.
Methods: Between December 2018 and July 2021, monthly consumptions of coagulation tests (PTT, aPTT, D-dimer, fibrinogen), total-, inpatient-, outpatient-, emergency-, non-emergency -admission numbers were collected. The relationship between inputs and outputs was modeled with the external input nonlinear autoregressive artificial neural network (ANN) (NARX) using MATLAB. Monthly test consumptions between January-July 2021 were used to test the models’ prediction power.
Results: According to the cointegration analysis, total-, emergency-, and non-emergency admission numbers plus the number of working days per month were included in the model. When aPTT and fibrinogen consumptions were estimated, it was possible to predict the other tests. Fifty months of data were used to predict the next six months, and the NARX prediction was the more robust approach for both tests.
Conclusions: The deep learning model gives better results than the intuitive approach in forecasting, even in the pandemic era, and it shows that more effective and efficient planning will be possible if ANN-supported decision mechanisms are used in forecasting tests’ consumptions in the procurement process.
期刊介绍:
The JOURNAL OF MEDICAL BIOCHEMISTRY (J MED BIOCHEM) is the official journal of the Society of Medical Biochemists of Serbia with international peer-review. Papers are independently reviewed by at least two reviewers selected by the Editors as Blind Peer Reviews. The Journal of Medical Biochemistry is published quarterly.
The Journal publishes original scientific and specialized articles on all aspects of
clinical and medical biochemistry,
molecular medicine,
clinical hematology and coagulation,
clinical immunology and autoimmunity,
clinical microbiology,
virology,
clinical genomics and molecular biology,
genetic epidemiology,
drug measurement,
evaluation of diagnostic markers,
new reagents and laboratory equipment,
reference materials and methods,
reference values,
laboratory organization,
automation,
quality control,
clinical metrology,
all related scientific disciplines where chemistry, biochemistry, molecular biology and immunochemistry deal with the study of normal and pathologic processes in human beings.