Alcohol User Prediction With Deep Learning Methods From Electronic Health Record Data

Abstract

Alcohol consumption has negative effects on individuals and societies in various areas, including health, economic, social and cultural aspects. Alcohol use prediction is a very important research topic to prevent the negative effects of alcohol. While dose-dependent alcohol use disorder is usually predicted in the literature, in this study, unlike the literature, dose-independent alcohol users are predicted. This prediction is made from electronic health record data using popular deep learning methods. The dataset used in the study consists of 24 different attributes including personal characteristics and health parameters of 991346 individuals collected from the National Health Insurance Service in Korea. The data were optimised after digitisation and normalisation preprocessing steps. A certain amount of training and test separation was applied to the dataset. Then, an alcohol user prediction model was developed using artificial neural networks, LSTM and CNN method. According to the results obtained, although the models achieved close prediction success, artificial neural networks achieved the best result. After artificial neural networks, CNN ranked second, and LSTM ranked last. By using more than one deep learning method together in the study, a conclusion about the general success of deep learning methods on the current problem has been made and a method that will make an important contribution to the solution of the problem has been put forward.

Keywords

• Alcohol user
• Deep learning
• LSTM
• CNN
• ANN

Elektronik Sağlık Kaydı Verilerinden Derin Öğrenme Yöntemleri ile Alkol Kullanıcısı Tahmini

Abstract

Alkol tüketiminin bireyler ve toplumlar üzerinde sağlıksal, ekonomik, sosyal ve kültürel yönler de dâhil olmak üzere çeşitli alanlarda olumsuz etkileri vardır. Alkol kullanımının öngörülmesi, alkolün olumsuz etkilerini önlemek için çok önemli bir araştırma konusudur. Literatürde genellikle doza bağlı alkol kullanım bozukluğu tahmin edilirken, bu çalışmada literatürden farklı olarak dozdan bağımsız alkol kullanıcısı tahmini yapılmaktadır. Bu tahmin popüler derin öğrenme yöntemleri kullanılarak elektronik sağlık kaydı verilerinden yapılmaktadır. Çalışmada kullanılan veri kümesi, Kore'deki Ulusal Sağlık Sigortası Hizmetinden toplanan 991346 bireye ait kişisel özellikler ve sağlık parametrelerini içeren 24 farklı öznitelikten oluşmaktadır. Veriler, sayısallaştırma ve normalizasyon ön işleme adımlarından sonra optimize edilmiştir. Veri kümesine belirli miktarda eğitim ve test ayrımı uygulanmıştır. Ardından, yapay sinir ağları, LSTM ve CNN yöntemi kullanılarak bir alkol kullanıcısı tahmin modeli geliştirilmiştir. Elde edilen sonuçlara göre modeller birbirine yakın tahmin başarısı elde etse de en iyi sonucu yapay sinir ağları elde etti. Yapay sinir ağlarından sonra CNN ikinci sırada, LSTM ise son sırada yer aldı. Çalışmada birden fazla derin öğrenme yöntemi bir arada kullanılarak derin öğrenme yöntemlerinin mevcut problem üzerindeki genel başarısı hakkında bir sonuca varılmış ve problemin çözümüne önemli katkı sağlayacak bir yöntem ortaya konulmuştur.

Keywords

• Alkol kullanıcısı
• Derin öğrenme
• LSTM
• CNN
• ANN

References

1. Centers for Disease Control and Prevention. (n.d.). Alcohol screening and brief intervention (SBI). Centers for Disease Control and Prevention. https://www.cdc.gov/alcohol-pregnancy/hcp/alcoholsbi/index.html
2. Sacks, J. J., Gonzales, K. R., Bouchery, E. E., Tomedi, L. E., & Brewer, R. D. (2015). 2010 national and state costs of excessive alcohol consumption. American Journal of Preventive Medicine, 49(5). https://doi.org/10.1016/j.amepre.2015.05.031
3. Mumtaz, W., Vuong, P. L., Xia, L., Malik, A. S., & Rashid, R. B. (2016). Automatic diagnosis of alcohol use disorder using EEG features. Knowledge-Based Systems, 105, 48–59. https://doi.org/10.1016/j.knosys.2016.04.026
4. Ebrahimi, A., Wiil, U. K., Andersen, K., Mansourvar, M., & Nielsen, A. S. (2020). A predictive machine learning model to determine alcohol use disorder. 2020 IEEE Symposium on Computers and Communications (ISCC), 1–7. https://doi.org/10.1109/iscc50000.2020.9219685
5. Sisodia, D. S., Agrawal, R., & Sisodia, D. (2018). A comparative performance of classification algorithms in predicting alcohol consumption among secondary school students. Advances in Intelligent Systems and Computing, 523–532. https://doi.org/10.1007/978-981-13-0923-6_45
6. Kinreich, S., Meyers, J. L., Maron-Katz, A., Kamarajan, C., Pandey, A. K., Chorlian, D. B., Zhang, J., Pandey, G., Subbie-Saenz de Viteri, S., Pitti, D., Anokhin, A. P., Bauer, L., Hesselbrock, V., Schuckit, M. A., Edenberg, H. J., & Porjesz, B. (2019). Predicting risk for alcohol use disorder using longitudinal data with multimodal biomarkers and family history: A machine learning study. Molecular Psychiatry, 26(4), 1133–1141. https://doi.org/10.1038/s41380-019-0534-x
7. Dhillon, A., Singh, A., Vohra, H., Ellis, C., Varghese, B., & Gill, S. S. (2020). IoTPulse: Machine learning-based Enterprise Health Information System to predict alcohol addiction in Punjab (India) using IOT and fog computing. Enterprise Information Systems, 16(7). https://doi.org/10.1080/17517575.2020.1820583
8. Narkbunnum, W., & Wisaeng, K. (2022). Prediction of depression for undergraduate students based on imbalanced data by using data mining techniques. Applied System Innovation, 5(6), 120. https://doi.org/10.3390/asi5060120

9. Saxena, S., Deep, V., & Sharma, P. (2018). Liver disorder prediction due to excessive alcohol consumption using slave. Advances in Intelligent Systems and Computing, 193–202. https://doi.org/10.1007/978-981-13-1951-8_18
10. Tseng, Y.-J., Wang, Y.-C., Hsueh, P.-C., & Wu, C.-C. (2022). Development and validation of machine learning-based risk prediction models of oral squamous cell carcinoma using salivary autoantibody biomarkers. BMC Oral Health, 22(1). https://doi.org/10.1186/s12903-022-02607-2
11. Patel, J. S., Su, C., Tellez, M., Albandar, J. M., Rao, R., Iyer, V., Shi, E., & Wu, H. (2022). Developing and testing a prediction model for periodontal disease using machine learning and Big Electronic Dental Record Data. Frontiers in Artificial Intelligence, 5. https://doi.org/10.3389/frai.2022.979525
12. Qiu, C., Su, K., Luo, Z., Tian, Q., Zhao, L., Wu, L., Deng, H., & Shen, H. (2024). Developing and comparing deep learning and machine learning algorithms for osteoporosis risk prediction. Frontiers in Artificial Intelligence, 7. https://doi.org/10.3389/frai.2024.1355287
13. Ahmad, H., Umar Khan, M., & Azam, M. (2024). Comparative analysis of machine learning methods for enhancing sleep efficiency and prediction. Information Systems Engineering and Management, 3–15. https://doi.org/10.1007/978-3-031-66854-8_1
14. Unlu, A., & Subasi, A. (2024). Substance use prediction using artificial intelligence techniques. Journal of Computational Social Science, 8(1). https://doi.org/10.1007/s42001-024-00356-6
15. Soo.Y. (2023, August 30). Smoking and drinking dataset with body signal. Kaggle. https://www.kaggle.com/datasets/sooyoungher/smoking-drinking-dataset
16. Subramaniam, K., & Naganathan, A. (2024). Enhancing retinal fundus image classification through active gradient deep convolutional neural network and red spider optimization. Neural Computing and Applications, 36(26), 16607–16619. https://doi.org/10.1007/s00521-024-09989-0