Óscar completed the TFM titled Processing and automatic learning from multivariate time series to predict and analyze the appearance of antimicrobial multiresistance in the ICU.
Can you summarize your Master's Thesis for us?
The Final Master's Project (TFM) focuses on the use of data science and artificial intelligence tools for the prediction of the appearance of Antimicrobial Multidrug Resistance (AMR) in the Intensive Care Unit (ICU) of the University Hospital of Fuenlabrada (HUF), as well as in the extraction of clinical knowledge about the acquisition of AMR through the visualization of the data. For this, information from the Electronic Clinical Record (EHR) of patients who have been admitted to the ICU-HUF from 2004 to 2020 (both years included) has been used. The EHR provides significant information about the state and evolution of the patient's health. The use and analysis of this type of data is challenging, due to the complexity and the existence of irregular patterns in clinical data. In this sense, the TFM models the EHR data as a Multivariate Time Series (MTS), considering temporal information both on the patient and on the overall state of the ICU. This last aspect, not always considered in the literature, has a significant impact on the ability to make relevant predictions. From the data science point of view, the fundamental contribution of the TFM is the use of advanced tools to model the relationships (similarities and distances) between MTS, including Time Cluster Kernel (TCK), Dynamic Time Warping (DTW), as well as ad-hoc methods based on learning similarities/distances between MTS. The previous tools have been complemented with the use of feature engineering schemes, unbalanced classification and dimensionality reduction. Likewise, different supervised architectures have been designed for prediction tasks, including logistic regressor, random forests and support vector machines. From the point of view of clinical application, the objective of this work is twofold: to predict the appearance of AMR and to analyze, through the use of advanced visualization tools (nonlinear dimensionality reduction and spectral clustering), possible groupings that allow the extraction of knowledge. .
Why did you apply for the awards call and how do you think your work contributes to the SDGs?
The World Health Organization (WHO) statement on the Antimicrobial Multidrug Resistance (AMR) as one of the main threats highlights the urgent need to address this problem in a multidisciplinary way. I firmly believe that artificial intelligence and data science have a fundamental role in the fight against AMR . My work focuses on designing AI techniques and data analysis to analyze the spread of AMR and contribute to an early detection of this problem. By using machine learning algorithms and predictive models, we are able to identify patterns and trends in data related to antimicrobial resistance. My work seeks to improve the ability of health systems to rapidly detect and address AMR . By identifying the areas of highest risk and anticipating possible outbreaks, we can implement more effective preventive and control strategies. This can help save lives, reduce the spread of antimicrobial-resistant infections, and ensure more responsible use of antimicrobial medicines. In short, I aspire to actively contribute to the achievement of the Sustainable Development Goals, in particular, the Health and Well-being SDG, by addressing one of the greatest threats to global health.
What are you doing now, and what would you like to do in the future?
I am currently working on a doctoral thesis on machine learning and data science. My work consists of exploring new techniques and algorithms to improve the accuracy and efficiency of machine learning models in various contexts, as well as their application to address real problems in society. I am excited to dive into the world of research and contribute to the advancement of artificial intelligence. In the future, I would like to continue with research work. I want to continue exploring new ways to use data to obtain valuable information and apply that knowledge in various sectors, such as medicine, transportation, etc. I also look forward to collaborating with other researchers and practitioners to drive the development and implementation of machine learning-based solutions that have a positive impact on society. Ultimately, my goal is to continue learning, innovating, and contributing to scientific and technological progress in the exciting field of machine learning and data science.
What do you recommend to future students?
Artificial intelligence is an incredibly powerful tool that allows us to address a wide range of problems in reality. From medicine to environmental sustainability, through economics and industry, AI is transforming the way we approach challenges in multiple fields. In this sense, it is interesting to note that taking a master's or doctorate in artificial intelligence will give you the opportunity to immerse yourself in a constantly evolving field and to be at the forefront of technology. You will be able to explore and develop new techniques and algorithms, and apply them in research projects that have a real impact on society. AI is revolutionizing the way we live and work, and being a part of that change is extremely exciting and rewarding.