Electromyogram prediction during anesthesia by using a hybrid intelligent model
- PDF / 1,676,598 Bytes
- 10 Pages / 595.276 x 790.866 pts Page_size
- 14 Downloads / 190 Views
ORIGINAL RESEARCH
Electromyogram prediction during anesthesia by using a hybrid intelligent model José‑Luis Casteleiro‑Roca1 · Marco Gomes2 · Juan Albino Méndez‑Pérez3 · Héctor Alaiz‑Moretón4 · María del Carmen Meizoso‑López1 · Benigno Antonio Rodríguez‑Gómez1 · José Luis Calvo‑Rolle1 Received: 29 October 2018 / Accepted: 18 August 2019 © Springer-Verlag GmbH Germany, part of Springer Nature 2019
Abstract In the search for new and more efficient ways to administer drugs, clinicians are turning to engineering tools. The availability of these models to predict physiological variables are a significant factor. A model is set out in this research to predict the EMG (electromyogram) signal during surgery, in patients under general anaesthesia. This prediction hinges on the Bispectral Index™ (BIS) and the infusion rate of the drug propofol. The results of the research are very satisfactory, with error values of less than 0.67 (for a Normalized Mean Squared Error). A hybrid intelligent model is used which combines both clustering and regression algorithms. The resulting model is validated and trained using real data. Keywords EMG · BIS™ · Clustering · MLP · SVM
1 Introduction
* Héctor Alaiz‑Moretón [email protected] José‑Luis Casteleiro‑Roca [email protected] Marco Gomes [email protected] Juan Albino Méndez‑Pérez [email protected] María del Carmen Meizoso‑López [email protected] Benigno Antonio Rodríguez‑Gómez [email protected] José Luis Calvo‑Rolle [email protected] 1
Dpto. de Ingeniería Industrial, University of A Coruña, A Coruña, Spain
2
ALGORITMI Centre, University of Minho, Braga, Portugal
3
Dpto. de Ingeniería de Sistemas y Automática y Arquitectura y Tecnología de Computadores, University of La Laguna, San Cristóbal de La Laguna , Spain
4
Dpto. de Ingeniería Eléctrica y de Sistemas y Automática, University of León, León, Spain
Interest in the application of advanced computing engineer‑ ing techniques has intensified considerably in recent years (Lemaître et al. 2015). A number of different approaches stand out. For instance, the enormous rise in the digitali‑ zation of information in this field has brought about the use of big data tools for extracting information and gener‑ ating inferences as to its relevance. Artificial intelligence (AI) techniques have given rise to tools for the planning and optimization of resources in healthcare systems. At the same time, the role of AI techniques is becoming widely extended in different clinical procedures and applications. One example is the use of automatic devices for delivering drugs during general anaesthesia, which is of great interest in anaesthesiology. Numerous advanced commercial devices for the infusion of drugs are based predicting the patient’s reactions during anaesthesia to establish the correct dosage of the drugs. Thus, the availability of reliable predictions for the evolution of the patient’s variables is essential to establish the correct dosage. It is necessary to regulate three main variable
Data Loading...
