Forecasting Weekly Dengue Cases by Integrating Google Earth Engine-Based Risk Predictor Generation and Google Colab-Based Deep Learning Modeling in Fortaleza and the Federal District, Brazil.

Publisher: MDPI Country of Publication: Switzerland NLM ID: 101238455 Publication Model: Electronic Cited Medium: Internet ISSN: 1660-4601 (Electronic) Linking ISSN: 16604601 NLM ISO Abbreviation: Int J Environ Res Public Health Subsets: MEDLINE

Original Publication: Basel : MDPI, c2004-

Dengue*/epidemiology ; Deep Learning*; Humans ; Brazil/epidemiology ; Artificial Intelligence ; Search Engine ; Forecasting

Efficient and accurate dengue risk prediction is an important basis for dengue prevention and control, which faces challenges, such as downloading and processing multi-source data to generate risk predictors and consuming significant time and computational resources to train and validate models locally. In this context, this study proposed a framework for dengue risk prediction by integrating big geospatial data cloud computing based on Google Earth Engine (GEE) platform and artificial intelligence modeling on the Google Colab platform. It enables defining the epidemiological calendar, delineating the predominant area of dengue transmission in cities, generating the data of risk predictors, and defining multi-date ahead prediction scenarios. We implemented the experiments based on weekly dengue cases during 2013-2020 in the Federal District and Fortaleza, Brazil to evaluate the performance of the proposed framework. Four predictors were considered, including total rainfall (R sum ), mean temperature (T mean ), mean relative humidity (RH mean ), and mean normalized difference vegetation index (NDVI mean ). Three models (i.e., random forest (RF), long-short term memory (LSTM), and LSTM with attention mechanism (LSTM-ATT)), and two modeling scenarios (i.e., modeling with or without dengue cases) were set to implement 1- to 4-week ahead predictions. A total of 24 models were built, and the results showed in general that LSTM and LSTM-ATT models outperformed RF models; modeling could benefit from using historical dengue cases as one of the predictors, and it makes the predicted curve fluctuation more stable compared with that only using climate and environmental factors; attention mechanism could further improve the performance of LSTM models. This study provides implications for future dengue risk prediction in terms of the effectiveness of GEE-based big geospatial data processing for risk predictor generation and Google Colab-based risk modeling and presents the benefits of using historical dengue data as one of the input features and the attention mechanism for LSTM modeling.

PLoS One. 2020 Aug 4;15(8):e0235697. (PMID: 32750051)
Emerg Infect Dis. 2006 May;12(5):800-6. (PMID: 16704841)
PLoS Negl Trop Dis. 2015 Apr 30;9(4):e0003632. (PMID: 25928673)
Nature. 2013 Apr 25;496(7446):504-7. (PMID: 23563266)
PLoS Negl Trop Dis. 2021 Jun 3;15(6):e0009475. (PMID: 34081695)
J R Soc Interface. 2020 Oct;17(171):20200691. (PMID: 33109025)
PLoS Comput Biol. 2017 Jul 20;13(7):e1005607. (PMID: 28727821)
PLoS One. 2018 Apr 2;13(4):e0195065. (PMID: 29608586)
Sci Adv. 2021 Jun 16;7(25):. (PMID: 34134985)
BMC Med. 2019 Sep 2;17(1):171. (PMID: 31474220)
PLoS Negl Trop Dis. 2022 Jun 13;16(6):e0010509. (PMID: 35696432)
Chaos Solitons Fractals. 2020 Nov;140:110212. (PMID: 32839642)
BMC Med. 2018 Aug 6;16(1):129. (PMID: 30078378)
Int J Environ Res Public Health. 2022 Jan 18;19(3):. (PMID: 35162056)
PLoS One. 2014 Apr 21;9(4):e95606. (PMID: 24752230)
J R Soc Interface. 2021 Jun;18(179):20201006. (PMID: 34129785)
PLoS Comput Biol. 2021 Feb 12;17(2):e1008618. (PMID: 33577550)
Infect Dis Poverty. 2021 Mar 25;10(1):40. (PMID: 33766145)
Spat Spatiotemporal Epidemiol. 2020 Nov;35:100372. (PMID: 33138951)
Int J Environ Res Public Health. 2022 Mar 01;19(5):. (PMID: 35270552)
PLoS Negl Trop Dis. 2020 Sep 24;14(9):e0008056. (PMID: 32970674)
BMC Infect Dis. 2020 Mar 12;20(1):208. (PMID: 32164548)
PLoS One. 2018 Jan 3;13(1):e0189988. (PMID: 29298320)
PLoS Negl Trop Dis. 2020 Oct 16;14(10):e0008710. (PMID: 33064770)
Sci Rep. 2021 Jan 13;11(1):923. (PMID: 33441598)
PLoS Negl Trop Dis. 2016 Feb 19;10(2):e0004473. (PMID: 26894570)
PLoS Negl Trop Dis. 2019 Mar 28;13(3):e0007213. (PMID: 30921321)
PLoS Negl Trop Dis. 2018 Dec 3;12(12):e0006990. (PMID: 30507968)
J Med Entomol. 2009 Sep;46(5):1117-24. (PMID: 19769043)
Neural Comput. 1997 Nov 15;9(8):1735-80. (PMID: 9377276)
BMC Infect Dis. 2018 Apr 17;18(1):183. (PMID: 29665781)
Biology (Basel). 2022 Jan 21;11(2):. (PMID: 35205036)
Cien Saude Colet. 2020 May;25(5):1641-1652. (PMID: 32402044)
PLoS One. 2019 Dec 30;14(12):e0226841. (PMID: 31887118)
Int J Environ Res Public Health. 2020 Jan 10;17(2):. (PMID: 31936708)
Sci Data. 2022 May 10;9(1):198. (PMID: 35538103)
J Med Entomol. 2006 Nov;43(6):1112-20. (PMID: 17162942)
Parasit Vectors. 2018 Apr 24;11(1):262. (PMID: 29690906)
BMC Infect Dis. 2019 Mar 21;19(1):272. (PMID: 30898092)
Sci Data. 2022 May 16;9(1):208. (PMID: 35577816)

42061134019 National Natural Science Foundation of China; QYZDB-SSW-DQC005 Key Research Program of Frontier Sciences of the Chinese Academy of Sciences; E0V00110YZ Institute of Geographic Sciences and Natural Resources Research, Chinese Acad-emy of Sciences

Keywords: Google Colab; Google Earth Engine; big geospatial data; cloud deep learning; dengue risk prediction

Date Created: 20221027 Date Completed: 20221028 Latest Revision: 20230117

20230120

PMC9603269

10.3390/ijerph192013555

36294134