The treatment used to combat COVID-19 disease is based on the combination of several drugs: hydroxychloroquine/chloroquine, azithromycin and kaletra. All of them can delay the ventricular repolarization process, thus prolonging the QT interval, and can cause serious arrhythmias that result in the death of the patient. This circumstance is conditioning the action protocol of the clinical staff in COVID-19 patients, in such a way that the duration of the QT interval of all patients with the aforementioned treatment is being permanently analyzed, which implies an additional burden for the staff clinical.
This project aims to automate the monitoring of QT intervals, through the use of signal processing techniques and artificial intelligence, to significantly reduce this load, offering the possibility of increasing the frequency of the EKGs performed and the consequent reduction in the risk of death. due to severe arrhythmias in patients with COVID-19. Additionally, it is important to note that the application to be developed with this project is useful for any treatment that includes drugs that can prolong the QT interval individually or in combination.
The results of the project may be applied free of charge within a period of three months from the start of the project, not only in the collaborating hospital centers, but also in any health center that so wishes, which will receive technical support for the implementation functioning. In this sense, it is also important to highlight that the application can be used even in those centers where there are no specialized clinical staff.
One of the main causes of the appearance of serious arrhythmias, such as ventricular fibrillation, which can cause the death of the patient, is due to the existence of a prolonged QT interval in the morphology of the heart beat. In this sense, it is important to highlight the existence of drugs that can produce said prolongation. Examples of these drugs can be hydroxychloroquine / chloroquine, azithromycin and kaletra. Although the information available on the antiviral action of hydroxychloroquine/chloroquine comes from non-vitro studies and series of patients with size and methodology limitations, its use constitutes a potential treatment for COVID-19 and is being used extensively in clinical practice. in these patients in association with azithromycin and kaletra. As previously discussed, hydroxychloroquine can prolong the QT interval of the electrocardiogram (EKG), and this risk increases with high doses and with the simultaneous use of azithromycin or other drugs with the potential to prolong the QT interval. Circumstance also suggested by a study carried out in patients infected by SARS-CoV-2. Thus, according to the Spanish Agency for Medicines and Health P roducts (AEMPS), extreme caution should be exercised, closely monitoring the patient.
On the other hand, the application of signal processing techniques has made great contributions regarding the extraction of EKG characteristics, such as the QRS complex, the RR interval and the QT interval. In this sense, it is worth highlighting the work carried out by Pan and Tompkins for the detection of QRS complexes, the development of which has been the basis for the extraction of information related to the morphology of the heartbeat. Regarding the extraction of the QT interval, there are also previous studies. However, since the detection of the T wave and the location of its end represents a certain difficulty since it can be confused with the isoelectric line or when there are alterations in the morphology, its extraction requires a particular analysis and treatment, being, in addition, necessary the estimate of the corrected QT interval to, in this way, refer the duration of the QT interval to 60 ppm.
Additionally, for the establishment of risk levels of suffering a serious arrhythmia, in addition to the duration of the QT interval, there are other factors that must be taken into consideration, such as the patient's age, sex, medical history, and current treatments. This circumstance, together with the greater or lesser prolongation of the QT interval mentioned above, establishes an environment of uncertainty and imprecision when making the decision about the continuity or not of the treatment used for COVID-19. In this sense, Soft-Computing techniques are intended to operate in environments subject to uncertainty and imprecision. Its application can be beneficial in inherently dynamic and uncertain environments such as the one described above. Specifically, it focuses on the application of Fuzzy Logic and Evolutionary Computing for the design of rule-based fuzzy meta-planners that incorporate knowledge acquisition strategies for their machine learning.
The objective of this project is to have an application that allows us to analyse the risk of severe arrhythmias automatically in patients treated for COVID-19, due to QT interval prolongation, which could lead to death. Such application would significantly reduce the burden on clinical staff, offering the possibility of increasing the frequency of EKGs performed and consequently reducing the risk of death from severe arrhythmias in patients with COVID-19. This objective can be subdivided into several partial sub-objectives:
- Sub-Obj. 1: Design and implementation of a reliable algorithm for the determination of QT interval duration.
- Sub-Obj. 1.1: EKG signal extraction.
- Sub-Obj. 1.2: QT interval extraction.
Sub-Obj. 2: Design and implementation of an expert system for the determination of arrhythmia risk levels based on QT interval duration and different patient profiles according to age and gender, among others. patient profiles according to age and sex, among others. The design of the expert system will take into consideration aspects related to the interpretability of the knowledge bases to be used for the used for the estimation of the risk of serious arrhythmia
- Sub-Obj. 2.1: Design of the expert system databases.
- Sub-Obj. 2.2: Design and implementation of the expert system.
Sub-Obj. 3: Implementation of an application that allows the analysis of EKG images, and the subsequent real-time sending of the alarm status for risk detection. The application located on the server will be based on the expert system.
- Sub-Obj. 3.1: Integration of the different sub-systems designed and implemented previously in one application.
- Sub-Obj. 3.2: Validation of the implemented application.
Sub-Obj. 4: Free distribution of the applications to health centres. In this way, the aim is to the generalised use of the system to be developed, providing technical support for installation and installation and start-up.
Sub-Obj. 5: As a consequence of the developed software, digital EKG records will be available for inclusion in the patient's medical record. In addition, a database per hospital centre will be available and will be used in the final part of the project to analyse the impact of the treatment on COVID19 patients in terms of the evolution of the QT interval according to the different profiles of patients treated by COVID-19.
- Sub-Obj. 1: Design and implementation of a reliable algorithm for the determination of QT interval duration.
As for the expected results, firstly, it will implement an application that will allow the estimation of the risk of serious arrhythmia occurrence in COVID-19 patients, which will be made available to all health centres, for whose installation, an image of the application will be created and deployed in a Docker container platform. which will be deployed on a Docker container platform.
Another relevant result to be obtained focuses on interpretability, as the result of its analysis will not only be the benefits of its analysis will not only have the benefits of the application itself, but also a better understanding of the mechanisms that are triggered prior to the onset of severe arrhythmias.
The present research project is characterised by its enormous potential for publication due, among other things, to its multi-disciplinary nature, as it deals with aspects related to different areas of knowledge such as soft-computing, cardiology and signal processing. As specific activities for the dissemination and exploitation of the project results, the following will be carried out:
Relations with other international institutions. The project's research team has extensive experience in cooperating with other international institutions to achieve results and maintains active collaboration with members of other institutions, with several stays and publications that have resulted. Thus, this project will establish links with national and international research centres in biomedicine and cardiology, such as, for example, the National Centre for Cardiovascular Research, the Spanish Society of Cardiology and the American Heart Association.
Publication of results in the most widely read journals in the fields of artificial intelligence, soft-computing, cardiology, and signal processing will be encouraged, e.g. Applied Soft-Computing (Elsevier), IEEE Transaction on Fuzzy Systems, IEEE Transaction on Biomedical Engineering, Computers in Cardiology, IEEE Signal Processing, etc.
In order for the research results to be used by other national and international research groups, the software will be published in a GitHub repository as the various versions are released.
Efforts will be made to present and disseminate the working prototypes to the specialised press. working prototypes. These presentations will also include invitations to companies in the sector that may have an interest in the work have an interest in the work developed.
Results will be presented at the main international conferences in the areas of work, e.g.: IEEE International Conference on Fuzzy Systems, IEEE areas of work, e.g.: IEEE International Conference on Fuzzy Systems, IEEE International Conference on Emerging Technologies, etc. In addition, special sessions will be organised special sessions will be organised at international scientific conferences in the field.
The application for European projects, European Structural Funds and ERA-NETs investment funds will be promoted.
Profiling in CORDIS for international consortium outreach.
Research social networks such as Researchgate will be used to disseminate partial results of the research project. results of the research project. A group will be set up on the professional social network LinkedIn to disseminate the partial results of the project. In addition, the results of the research and the research results and implementations in LinkedIn groups of professionals related to the progress made in each case.
The TIC-188 research group's website will be used to disseminate publications and other work related to the development of the research project. In addition, a web page will be developed to disseminate and publicise the results of the project.
The completion of Doctoral Theses with European/international mention of the staff hired in charge of the project will be encouraged, thus promoting training and the search for novel results that may not only help to achieve the objectives proposed in the project but also present a contribution to science.
News in professional magazines, blogs or other portals, as well as in international digital newspapers where relevant.
Press releases highlighting the benefits of the research.
Distribution of promotional materials, including brochures, flyers, posters, etc., as well as videos created about the project on a YouTube channel.
Organisation of visits to potential customers: demonstrations based on the use cases and results of the project will be presented.
Demonstrations at various venues (workshops, events and specialised international trade fairs).
With regard to transfer, it must be said that the possibilities for transferring the research carried out in the research project are great and immediate, as the aim is to distribute the application developed free of charge to the different health centres so that they can incorporate it into their action protocols.