Research Article
BibTex RIS Cite

Genel Anestezide Kullanılan Propofolün Başlangıç Dozunu Tahmin Eden Yapay Sinir Ağları Model

Year 2020, , 1482 - 1495, 30.09.2020
https://doi.org/10.31202/ecjse.764719

Abstract

Anestezide ilaç dozunun doğru hesaplanması çok önemli bir rol oynamaktadır. Preoperatif anestezide, bir anestezi uzmanı, hipnotik ilaçların dozlarını hastanın faktörlerine göre hesaplamakta ve bunları klinik ortamda bir başlangıç ve devam dozu şeklinde uygulamaktadır. Bu çalışmada, hipnotik bir ajan olan propofolün başlangıç dozu (mg), premedikasyon veya ek ilaç kullanılmadığı varsayılarak çok katmanlı ileri beslemeli yapay sinir ağı (ÇKYSA) yapısı kullanılarak tahmin edilmiştir. Yaş (yıl), ağırlık (kg), boy (m) ve eşlik eden hastalık faktörleri önerilen öngörücü ağın girdilerini oluşturmuştur. Bu çalışma için veri seti 299 hasta örneği ile uzman anestezistler tarafından oluşturulmuştur. En iyi tahmin ediciyi bulmak için farklı hiperparametrelerle tasarlanan birçok YSA modeli test edilmiş ve sonuçları kaydedilmiştir. Elde edilen sonuçlara göre, en iyi tahminci % 92'nin üzerinde başarı oranlarıyla propofolün başlangıç dozunu tahmin etmiştir. Bu model sayesinde, potansiyel anestezik ilaçların başlangıç dozlarının YSA tarafından hesaplanabileceği kanıtlanmıştır. Bu çalışmada, uygulamanın anestezistleri asiste edebileceği önerilmektedir.

Thanks

Projede veri setinin oluşturulmasını sağlayan ve değerli bilgilerini bizlerle paylaşan Çankırı Karatekin Hastanesi Anesteziyoloji ve Reanimasyon bölümü çalışanlarına teşekkürlerimizi sunarız.

References

  • [1] M. Buscema, G. Massini, M. Breda, W.A. Lodwick, F. Newman, M. Asadi-Zeydabadi, Artificial neural networks, Stud. Syst. Decis. Control. 131 (2018) 11–35.
  • [2] M. Mhatre, F. Siddiqui, M. Dongre, P. Thakur, A Review paper on Artificial Neural Networks: A Prediction Technique., Int. J. Sci. Eng. Res. 8 (2017) 1–3.
  • [3] V. Sutariya, A. Groshev, P. Sadana, D. Bhatia, Y. Pathak, Artificial Neural Network in Drug Delivery and Pharmaceutical Research., Open Bioinforma. J. 7 (2014) 49–62. [4] A.O. Basile, A. Yahi, N.P. Tatonetti, Artificial Intelligence for Drug Toxicity and Safety, Trends Pharmacol. Sci. 40 (2019) 624–635.
  • [5] G. Camps-Valls, B. Porta-Oltra, E. Soria-Olivas, J.D. Martin-Guerrero, J.J. Perez-Ruixo, N.V. Jimenez-Torres, Prediction of cyclosporine dosage in patients after kidney transplantation using neural networks, IEEE Trans. Biomed. Eng. 50 (2003) 442–448.
  • [6] M.E. Brier, J.M. Zurada, G.R. Aronoff, Neural Network Predicted Peak and Trough Gentamicin Concentrations, Pharm. Res. An Off. J. Am. Assoc. Pharm. Sci. 12 (1995) 406– 412.
  • [7] C. Pfitzner, S. May, A. Nüchter, Neural network-based visual body weight estimation for drug dosage finding, in: M.A. Styner, E.D. Angelini (Eds.), Med. Imaging 2016 Image Process., 2016: p. 97841Z.
  • [8] O. Caelen, O. Cailloux, D. Ghoundiwal, A. Alexander, Real-time prediction of an anesthetic monitor index using machine learning, Artif. Intell. Med. (2011).
  • [9] C.S. Lin, Y.C. Li, M.S. Mok, C.C. Wu, H.W. Chiu, Y.H. Lin, Neural network modeling to predict the hypnotic effect of propofol bolus induction., Proc. AMIA Symp. (2002) 450–454. [10] Y. Sakuma, R. Kohno, A Dynamic Model Estimation Scheme for Model Predictive Control of Anesthesia Using Recurrent Neural Network, in: 2018 12th Int. Symp. Med. Inf. Commun. Technol., IEEE, 2018: pp. 1–5.

Artificial Neural Network Model Estimating the Initial Dose of Propofol Used in General Anesthesia

Year 2020, , 1482 - 1495, 30.09.2020
https://doi.org/10.31202/ecjse.764719

Abstract

The right dosing of drugs has a pivotal role in anaesthesia. In preoperative anaesthesia, an anaesthesiologist, calculates the doses of hypnotic drugs according to the patient's factors and implements them in the clinical setting in the form of an initial and continuation dose. In this study, the initial dose of a hypnotic agent propofol (mg) was estimated using multilayer feed forward artificial neural network (MNN) structure, assuming no premedication or additional medication was used. The factors of age (year), weight (kg), height (m) and concomitant diseases have constituted the inputs of the proposed predictive network. Data set for this study consists of 299 patient samples and was created by expert anaesthesiologists. Many ANN models designed with different hyperparameters were tested to find the best estimator, and the results were recorded. According to the obtained results, the best estimator has estimated the initial dose of propofol with success rates over 92%. Thanks to this model, it has been proven that the initial doses of potentially anaesthetic drugs can be calculated by ANN, so that the application can be considered as an aid to anaesthesiologists.

References

  • [1] M. Buscema, G. Massini, M. Breda, W.A. Lodwick, F. Newman, M. Asadi-Zeydabadi, Artificial neural networks, Stud. Syst. Decis. Control. 131 (2018) 11–35.
  • [2] M. Mhatre, F. Siddiqui, M. Dongre, P. Thakur, A Review paper on Artificial Neural Networks: A Prediction Technique., Int. J. Sci. Eng. Res. 8 (2017) 1–3.
  • [3] V. Sutariya, A. Groshev, P. Sadana, D. Bhatia, Y. Pathak, Artificial Neural Network in Drug Delivery and Pharmaceutical Research., Open Bioinforma. J. 7 (2014) 49–62. [4] A.O. Basile, A. Yahi, N.P. Tatonetti, Artificial Intelligence for Drug Toxicity and Safety, Trends Pharmacol. Sci. 40 (2019) 624–635.
  • [5] G. Camps-Valls, B. Porta-Oltra, E. Soria-Olivas, J.D. Martin-Guerrero, J.J. Perez-Ruixo, N.V. Jimenez-Torres, Prediction of cyclosporine dosage in patients after kidney transplantation using neural networks, IEEE Trans. Biomed. Eng. 50 (2003) 442–448.
  • [6] M.E. Brier, J.M. Zurada, G.R. Aronoff, Neural Network Predicted Peak and Trough Gentamicin Concentrations, Pharm. Res. An Off. J. Am. Assoc. Pharm. Sci. 12 (1995) 406– 412.
  • [7] C. Pfitzner, S. May, A. Nüchter, Neural network-based visual body weight estimation for drug dosage finding, in: M.A. Styner, E.D. Angelini (Eds.), Med. Imaging 2016 Image Process., 2016: p. 97841Z.
  • [8] O. Caelen, O. Cailloux, D. Ghoundiwal, A. Alexander, Real-time prediction of an anesthetic monitor index using machine learning, Artif. Intell. Med. (2011).
  • [9] C.S. Lin, Y.C. Li, M.S. Mok, C.C. Wu, H.W. Chiu, Y.H. Lin, Neural network modeling to predict the hypnotic effect of propofol bolus induction., Proc. AMIA Symp. (2002) 450–454. [10] Y. Sakuma, R. Kohno, A Dynamic Model Estimation Scheme for Model Predictive Control of Anesthesia Using Recurrent Neural Network, in: 2018 12th Int. Symp. Med. Inf. Commun. Technol., IEEE, 2018: pp. 1–5.
There are 8 citations in total.

Details

Primary Language English
Subjects Engineering
Journal Section Makaleler
Authors

Esra Sivari 0000-0002-5708-7421

Zafer Civelek 0000-0001-6838-3149

Publication Date September 30, 2020
Submission Date July 6, 2020
Acceptance Date August 26, 2020
Published in Issue Year 2020

Cite

IEEE E. Sivari and Z. Civelek, “Artificial Neural Network Model Estimating the Initial Dose of Propofol Used in General Anesthesia”, ECJSE, vol. 7, no. 3, pp. 1482–1495, 2020, doi: 10.31202/ecjse.764719.