Öz
Diyabet, insan vücudunda birçok hastalığı tetikleyen kronik bir hastalıktır. İnsan vücudunu olumsuz etkileyen diyabet için kritik şeker seviyeleri, hiperglisemi (yüksek kan şekeri) ve hipoglisemi (düşük kan şekeri) olarak adlandırılır. Kan şekerinin bu seviyelerin üstünde veya altında olması insan vücudunda çeşitli tahribatlara neden olmaktadır. Mevcut sistemler, kan şekerini sürekli ölçerek takip yapmakta olup, kritik seviyeler geçildikten sonra kullanıcıya uyarı vermektedir. Kullanıcının uyarının farketmesi, gerekli ilaç ve/veya tedbirleri almasına takiben kan şekeri istenilen seviyeye indirilebilmektedir. Bu durum, hasta vücudunun belirli bir süre kritik seviyelerde çalışmasına neden olmaktadır. Bu sorunun önüne geçebilmek için, kan şekerinin zaman içindeki değişiminden bir sonraki kan şekerinin değerini tahmin edebilecek ve böylelikle, kritik seviyelerine ulaşmadan hastayı uyarabilecek bir sistem geliştirilmiştir. Önerilen sistem, uzun-kısa süreli bellek (Long Short Term Memory-LSTM) tabanlı tekrarlayan sinir ağı (Recurrent Neural Network-RNN) ile zamanla değişen kan şekeri değerlerinden bir sonraki değeri tahmin edebilmektedir. Diyabet hastalarından elde edilen gerçek veriler ile eğitilen sistem, %95.6 doğruluğa karşılık gelen 3.72 mg/dl'den daha düşük bir hata ile kan şekerini tahmin edebilmiştir. Önerilen sistem, ayrıca, kendi geliştirdiğimiz BffDiabetes adlı Android uygulamamızla birleştirilmiştir. Uygulama, ölçülen kan şekeri değerini bir bulut sistemi üzerinden bir sonraki şeker seviyesini tahmin etmek için sunucuya gönderir. Sunucuda koşturulan LSTM tabanlı tahmin algoritmamız gelen değere bağlı olarak gelecek üç zaman adımı için şeker değerlerini hesaplar. Tahmin sonuçları, kan şekeri seviyesinin kritik aşamalara ulaşma eğiliminde olup olmadığını değerlendirmek için bulut sistemi üzerinden tekrar Android uygulamasına gönderilir. Bu eğilim tespit edilirse, uygulama hastayı gerekli önlemler için bir bildirimle uyarır. Böylelikle, kan şekerini ölçerek kablosuz (Bluetooth) aktarım yapan cihazlarla çalışabilecek bir platform diyabet hastalarının kullanımına sunularak günlük hayat kalitelerinin artırılması amaçlanmıştır.
Anahtar Kelimeler
Yapay zeka Tekrarlayan Sinir Ağı LSTM Akıllı Telefon Android
Kaynakça
- Aliberti, A., Pupillo, I., Terna, S., Macii, E., Di Cataldo, S., Patti, E., & Acquaviva, A. (2019). A multi-patient data-driven approach to blood glucose prediction. J IEEE Access, 7, 69311-69325.
- Association, A. D. (2014). Diagnosis and classification of diabetes mellitus. J Diabetes care, 37(Supplement 1), S81-S90.
- Bunescu, R., Struble, N., Marling, C., Shubrook, J., & Schwartz, F. (2013). Blood glucose level prediction using physiological models and support vector regression. Paper presented at the 2013 12th International Conference on Machine Learning and Applications.
- Daskalaki, E., Prountzou, A., Diem, P., & Mougiakakou, S. G. (2012). Real-time adaptive models for the personalized prediction of glycemic profile in type 1 diabetes patients. J Diabetes technology therapeutics, 14(2), 168-174.
- Firebase. (2021). Firebase Realtime Database. Retrieved from https://firebase.google.com/docs/database
- Kap, Ö., Kilic, V., Hardy, J. G., & Horzum, N. (2021). Smartphone-based colorimetric detection systems for glucose monitoring in the diagnosis and management of diabetes. J Analyst.
- Li, W.-J., Yen, C., Lin, Y.-S., Tung, S.-C., & Huang, S. (2018). JustIoT Internet of Things based on the Firebase real-time database. Paper presented at the 2018 IEEE International Conference on Smart Manufacturing, Industrial & Logistics Engineering (SMILE).
- Mercan, Ö. B. (2020). Deep Learning based Colorimetric Classification of Glucose with Au-Ag nanoparticles using Smartphone. Paper presented at the 2020 Medical Technologies Congress (TIPTEKNO).
- Mercan, Ö. B., Doğan, V., & Kılıç, V. (2020). Time Series Analysis based Machine Learning Classification for Blood Sugar Levels. Paper presented at the 2020 Medical Technologies Congress (TIPTEKNO).
- Mercan, Ö. B., & Kılıç, V. (2020). Fuzzy classifier based colorimetric quantification using a smartphone. Paper presented at the International Conference on Intelligent and Fuzzy Systems.
- Mercan, Ö. B., Kılıç, V., & Şen, M. (2021). Machine learning-based colorimetric determination of glucose in artificial saliva with different reagents using a smartphone coupled μPAD. J Sensors Actuators B: Chemical, 329, 129037.
- Midroni, C., Leimbigler, P., Baruah, G., Kolla, M., Whitehead, A., & Fossat, Y. (2018). Predicting glycemia in type 1 diabetes patients: experiments with xg-boost. Paper presented at the KHD@ IJCAI.
- Moroney, L. (2017). The firebase realtime database. In The Definitive Guide to Firebase (pp. 51-71): Springer.
- Mutlu, A. Y., & Kılıç, V. (2018). Machine learning based smartphone spectrometer for harmful dyes detection in water. Paper presented at the 2018 26th Signal Processing and Communications Applications Conference (SIU).
- Olah, C. (2021). Understanding LSTM Networks. Retrieved from https://colah.github.io/posts/2015-08-Understanding-LSTMs/
- Pala, T., & Yücedağ, İ. (2016). Veri Madenciliği Tekniklerinden Sınıflandırma Kullanılarak Tip 2 Diyabet Tanısı. Paper presented at the International Artificial Intelligence and Data Processing Symposium.
- Pérez-Gandía, C., Facchinetti, A., Sparacino, G., Cobelli, C., Gómez, E., Rigla, M., . . . Hernando, M. (2010). Artificial neural network algorithm for online glucose prediction from continuous glucose monitoring. J Diabetes technology therapeutics, 12(1), 81-88.
- Song, W., Cai, W., Li, J., Jiang, F., & He, S. (2019). Predicting Blood Glucose Levels with EMD and LSTM Based CGM Data. Paper presented at the 2019 6th International Conference on Systems and Informatics (ICSAI).
- Strollo, F., Furia, A., Verde, P., Bellia, A., Grussu, M., Mambro, A., . . . Gentile, S. (2021). Technological innovation of Continuous Glucose Monitoring (CGM) as a tool for commercial aviation pilots with insulin-treated diabetes and stakeholders/regulators: A new chance to improve the directives? %J diabetes research clinical practice, 172.
- Sun, Q., Jankovic, M. V., Bally, L., & Mougiakakou, S. G. (2018). Predicting blood glucose with an LSTM and Bi-LSTM based deep neural network. Paper presented at the 2018 14th Symposium on Neural Networks and Applications (NEUREL).
- Wiley, M. T. (2011). Machine learning for diabetes decision support. Ohio University,
- Xie, J., & Wang, Q. (2018). Benchmark machine learning approaches with classical time series approaches on the blood glucose level prediction challenge. Paper presented at the KHD@ IJCAI.
- Yahyaoui, A., Jamil, A., Rasheed, J., & Yesiltepe, M. (2019). A decision support system for diabetes prediction using machine learning and deep learning techniques. Paper presented at the 2019 1st International Informatics and Software Engineering Conference (UBMYK).
Öz
Diabetes is a chronic disease that triggers many diseases in the human body. Critical sugar levels for diabetes, which negatively affects the human body, are referred to as hyperglycemia (high blood sugar) and hypoglycemia (low blood sugar). Blood sugar being above or below these levels causes various damages in the human body. Existing systems monitor blood sugar continuously, and alert the user after critical levels are passed. After the user notices the warning and takes the necessary medication and/or precautions, the blood sugar can be reduced to the desired level. This situation causes the patient's body to work at critical levels for a certain period of time. In order to avoid this problem, a system has been developed that can predict the value of the next blood sugar from the change in blood sugar over time and thus alert the patient before reaching critical levels. The proposed system can predict the next value from the time-varying blood sugar values with the Long Short Term Memory (LSTM) based recurrent neural network (RNN). Trained with real data from diabetes patients, the system was able to predict blood sugar with an error of less than 3.72 mg/dl, corresponding to 95.6% accuracy. The proposed system is also combined with our custom-designed Android application called BffDiabetes. The application sends the measured blood sugar value via a cloud system to the server to predict the next sugar level. Our LSTM-based prediction algorithm run on the server calculates sugar values for the next three time steps based on the incoming value. The prediction results are sent back to the Android app via the cloud system to assess whether the blood sugar level tends to reach critical stages. If this trend is detected, the application alerts the patient with a notification for necessary precautions. In this way, it is aimed to increase the quality of daily life by offering a platform that can work with devices that make wireless (Bluetooth) transmission by measuring blood sugar.
Anahtar Kelimeler
Artificial Intelligence Recurrent Neural Network LSTM Smartphone Android
Kaynakça
- Aliberti, A., Pupillo, I., Terna, S., Macii, E., Di Cataldo, S., Patti, E., & Acquaviva, A. (2019). A multi-patient data-driven approach to blood glucose prediction. J IEEE Access, 7, 69311-69325.
- Association, A. D. (2014). Diagnosis and classification of diabetes mellitus. J Diabetes care, 37(Supplement 1), S81-S90.
- Bunescu, R., Struble, N., Marling, C., Shubrook, J., & Schwartz, F. (2013). Blood glucose level prediction using physiological models and support vector regression. Paper presented at the 2013 12th International Conference on Machine Learning and Applications.
- Daskalaki, E., Prountzou, A., Diem, P., & Mougiakakou, S. G. (2012). Real-time adaptive models for the personalized prediction of glycemic profile in type 1 diabetes patients. J Diabetes technology therapeutics, 14(2), 168-174.
- Firebase. (2021). Firebase Realtime Database. Retrieved from https://firebase.google.com/docs/database
- Kap, Ö., Kilic, V., Hardy, J. G., & Horzum, N. (2021). Smartphone-based colorimetric detection systems for glucose monitoring in the diagnosis and management of diabetes. J Analyst.
- Li, W.-J., Yen, C., Lin, Y.-S., Tung, S.-C., & Huang, S. (2018). JustIoT Internet of Things based on the Firebase real-time database. Paper presented at the 2018 IEEE International Conference on Smart Manufacturing, Industrial & Logistics Engineering (SMILE).
- Mercan, Ö. B. (2020). Deep Learning based Colorimetric Classification of Glucose with Au-Ag nanoparticles using Smartphone. Paper presented at the 2020 Medical Technologies Congress (TIPTEKNO).
- Mercan, Ö. B., Doğan, V., & Kılıç, V. (2020). Time Series Analysis based Machine Learning Classification for Blood Sugar Levels. Paper presented at the 2020 Medical Technologies Congress (TIPTEKNO).
- Mercan, Ö. B., & Kılıç, V. (2020). Fuzzy classifier based colorimetric quantification using a smartphone. Paper presented at the International Conference on Intelligent and Fuzzy Systems.
- Mercan, Ö. B., Kılıç, V., & Şen, M. (2021). Machine learning-based colorimetric determination of glucose in artificial saliva with different reagents using a smartphone coupled μPAD. J Sensors Actuators B: Chemical, 329, 129037.
- Midroni, C., Leimbigler, P., Baruah, G., Kolla, M., Whitehead, A., & Fossat, Y. (2018). Predicting glycemia in type 1 diabetes patients: experiments with xg-boost. Paper presented at the KHD@ IJCAI.
- Moroney, L. (2017). The firebase realtime database. In The Definitive Guide to Firebase (pp. 51-71): Springer.
- Mutlu, A. Y., & Kılıç, V. (2018). Machine learning based smartphone spectrometer for harmful dyes detection in water. Paper presented at the 2018 26th Signal Processing and Communications Applications Conference (SIU).
- Olah, C. (2021). Understanding LSTM Networks. Retrieved from https://colah.github.io/posts/2015-08-Understanding-LSTMs/
- Pala, T., & Yücedağ, İ. (2016). Veri Madenciliği Tekniklerinden Sınıflandırma Kullanılarak Tip 2 Diyabet Tanısı. Paper presented at the International Artificial Intelligence and Data Processing Symposium.
- Pérez-Gandía, C., Facchinetti, A., Sparacino, G., Cobelli, C., Gómez, E., Rigla, M., . . . Hernando, M. (2010). Artificial neural network algorithm for online glucose prediction from continuous glucose monitoring. J Diabetes technology therapeutics, 12(1), 81-88.
- Song, W., Cai, W., Li, J., Jiang, F., & He, S. (2019). Predicting Blood Glucose Levels with EMD and LSTM Based CGM Data. Paper presented at the 2019 6th International Conference on Systems and Informatics (ICSAI).
- Strollo, F., Furia, A., Verde, P., Bellia, A., Grussu, M., Mambro, A., . . . Gentile, S. (2021). Technological innovation of Continuous Glucose Monitoring (CGM) as a tool for commercial aviation pilots with insulin-treated diabetes and stakeholders/regulators: A new chance to improve the directives? %J diabetes research clinical practice, 172.
- Sun, Q., Jankovic, M. V., Bally, L., & Mougiakakou, S. G. (2018). Predicting blood glucose with an LSTM and Bi-LSTM based deep neural network. Paper presented at the 2018 14th Symposium on Neural Networks and Applications (NEUREL).
- Wiley, M. T. (2011). Machine learning for diabetes decision support. Ohio University,
- Xie, J., & Wang, Q. (2018). Benchmark machine learning approaches with classical time series approaches on the blood glucose level prediction challenge. Paper presented at the KHD@ IJCAI.
- Yahyaoui, A., Jamil, A., Rasheed, J., & Yesiltepe, M. (2019). A decision support system for diabetes prediction using machine learning and deep learning techniques. Paper presented at the 2019 1st International Informatics and Software Engineering Conference (UBMYK).
Ayrıntılar
| Birincil Dil | Türkçe |
|---|---|
| Konular | Mühendislik |
| Bölüm | Makaleler |
| Yazarlar | |
| Yayımlanma Tarihi | 31 Temmuz 2021 |
| Yayımlandığı Sayı | Yıl 2021 Sayı: 26 - Ejosat Özel Sayı 2021 (HORA) |
