AI Knowledge Transfer from the University to Society

Applications in High-Impact Sectors

José Guadix Martín, Milica Lilic, Marina Rosales Martínez, José Guadix Martín, Milica Lilic, Marina Rosales Martínez

• Publisher: CRC Press
• Year: 2022
• ISBN:
  9781000568172
• Edition: 1
• Topic: Computer Science
• Subtopic: Computer Science General
• Index: Computer Science

AI Knowledge Transfer from the University to Society: Applications in High-Impact Sectors brings together examples from the "Innovative Ecosystem with Artificial Intelligence for Andalusia 2025" project at the University of Seville, a series of sub-projects composed of research groups and different institutions or companies that explore the use of Artificial Intelligence in a variety of high-impact sectors to lead innovation and assist in decision-making.

Key Features

• Includes chapters on health and social welfare, transportation, digital economy, energy efficiency and sustainability, agro-industry, and tourism

• 
  
  

• Great diversity of authors, expert in varied sectors, belonging to powerful research groups from the University of Seville with proven experience in the transfer of knowledge to the productive sector and agents attached to the Andalucía TECH Campus

• 
  
  

1 Health and Social Welfare

DOI: 10.1201/9781003276609-1

The health and well-being of citizens is a key factor that must be guaranteed, as has recently become clear with the global health pandemic. Being aware of this, the University of Seville research groups apply Artificial Intelligence techniques to address existing problems from a multidisciplinary perspective and improve current solutions. The health sector has large amounts of patient data, which, when correctly analyzed and visualized, will increase the efficiency of the applied treatments. For this reason, there are medicine, computer science, or physics groups that address different issues.

An example of this is the application of Artificial Intelligence to optimize focused ultrasound cleaning of implanted shunts in patients with different pathologies. An innovative technology is used for the non-invasive preventive cleaning of shunts, valves, or catheters implanted in patients with different pathologies. Artificial Intelligence on the data of test cases determines the optimal parameters for its application in diverse clinical environments, adapted to the individual circumstances and the specific devices used in neurosurgery, oncology, and other clinical areas.

Likewise, another example of robotics that allows automating medical processes is shown, guaranteeing the quality and safety of the steps carried out during the process. It is intended to improve health processes with the use of emerging technologies, such as blockchain or the robotization of software processes.

Furthermore, there is another situation generated by the extension of life expectancy in modern societies and the social challenge that arises in the prevention and treatment of neurodegenerative diseases suffered by the constantly increasing elderly population. The underlying problem is the investigation of changes in gene expression of individual neurons in response to nerve terminal dysfunction in the brain of genetically modified mice. Within this experimentation, Machine Learning methods are used for the bioinformatic analysis of a data set that includes thousands of cells with thousands of genes each. This issue has a utility to be addressed in future therapeutic strategies.

1.1 Artificial Intelligence for the Optimization of Focused Ultrasound Cleaning of Shunts Implanted in Patients of Different Pathologies

Emilio Gómez-González and Javier Márquez-Rivas

Abstract

Focused ultrasound is an innovative technology for potential non-invasive, preventive cleaning of shunts and infusion systems (valves, catheters) implanted in patients with different pathologies. Artificial Intelligence tools are used to determine optimal parameters for their application in various clinical settings, tailored to the individual circumstances and the specific devices used in neurosurgery, oncology, and other clinical areas.

Introduction

The expanding implementation of technologies based on Artificial Intelligence (AI) represents an authentic revolution in most aspects of daily life, having a very relevant impact on Medicine and Healthcare (Gómez González & Gómez, 2020). Diagnosis and treatment methods powered by AI, and their combinations with augmented reality (AR) devices, evolve toward personalized three-dimensional (3D) models, and to the development of advanced simulation tools for predictive, interactive visualization of the response of organisms under different circumstances and stimuli.

The European Union considers the development of AI an essential area, particularly in Medicine and Healthcare, and has launched different initiatives to lead its scientific and technological advances and to establish the bases for their implementation (Nepelski, 2021). One of these initiatives is the ATTRACT Program, which seeks to identify and support groundbreaking technologies with a clear potential of application for solving societal, clinical, environmental, problems of high impact (ATTRACT Program, 2019). This work presents an extension of a successful project within the aforementioned ATTRACT Program (FUSCLEAN Project, 2020).

State of the Art

The project described in this article addresses the preventive cleaning of fluid infusion or drainage (shunting) systems implanted in the human body to avoid their obstruction by the deposit of residues on their inner surfaces.

Intended cleaning is achieved by means of focused ultrasound beams that generate a certain 3D distribution of energy, safe for the patient but which produces mechanical effects—controlled cavitation—in the fluid which, in turn, disaggregate and remove the deposits from the walls of the conduits and valves.

The initial field of application of this technology is neurosurgery, to avoid the most frequent complications in patients with hydrocephalus, but it has the potential to be extended for additional applications in other clinical areas in which shunting systems are implanted in the human body to infuse or drain fluids, such as pain control, oncology treatments, and anesthesia.

In this research, AI systems are employed to conduct multiple 3D simulations to determine and optimize the parameters required to achieve the desired level and spatial distribution of ultrasound energy volume density in the targeted regions. AR systems are also tested. They include different types of display, from screens or tablets to glasses or helmet-mounted devices, to superimpose “layers” of information on the user’s real field of vision (Baraas et al., 2021). They can also be integrated with positioning and motion tracking elements to enable simultaneous viewing of real objects with data or graphics in different orientations or perspectives.

Currently, there are no references to early detection of debris in fluids flowing through implanted shunts in the human body or to preventive procedures to avoid the deposits of materials in the conduits and valves and their possible obstruction. Therefore, no additional references are found about energy focusing and visualization technologies for the intended cleaning or about similar applications.

Contribution

The described technology is based on the generation and real-time control of a volumetric distribution of ultrasound energy in the region of interest (called “sonication volume”), during the time interval necessary to produce the detachment of deposits adhered to the inner surfaces of conduits and valves. It is important to note that ultrasound waves are mechanical waves (not electromagnetic radiation), without ionizing effects. The intended effect is achieved by concentrating individual beams in the sonication volume, analogously to the concentration of the sun’s rays using a magnifying glass.

The desired energy distribution is achieved by the overlapping of the ultrasound fields generated by independent emitters at certain frequencies, whose powers and orientation are precisely determined. This process requires identifying the exact location of the shunt elements—catheters and valves—under the skin and generating a 3D numerical model of the emission of the various ultrasound beams. The combined ultrasound field produces a controlled cavitation effect in the targeted volume—inside the conduits—that disintegrates the debris in the deposits so that they are evacuated by the circulation of the fluid itself. Likewise, it is necessary to monitor the temperature distribution in the zone surrounding the region of concentration of the ultrasound energy to avoid possible thermal effects. We, therefore, explore the combined use of visible and thermal imaging cameras.

In this technology, AI algorithms are used to optimize the emission patterns of the individual ultrasound beams, considering the various factors inherent to the implanted systems (types of valves and conduits, with their different mechanical properties), the circulating fluid (biofluids, drugs) and the patient. It is also required to include in the calculations the biological variability of elements which may be present in fluids and the state of the corresponding system of each person, as well as their specific conditions (temperature, cardiac and respiratory pulsations, possible concurrent pathologies) at the time of the application of the ultrasound beams. The different types, positions, and circumstances of the implanted valves and conduits must also be considered since the presence of elements such as scar tissue, and the various characteristics of the layers of the skin and subcutaneous fat, determine the effective distribution of ultrasound in the targeted structures, that is, within the implanted devices and components. This multitude of factors is completely different for each clinical setting and person. Multiple 3D simulations must be developed in a variety of physical scenarios, under varying circumstances and with different materials of complex properties. From them, the optimal sonication parameters are extracted using AI tools.

In our work, we explore clinical applications in which the consequences of possible obstructions of implanted systems may be particularly relevant for patient safety and the effectiveness of the treatment, and in which the removal and replacement of the implanted devices are more complex or difficult. These areas of applications include, in addition to patients of hydrocephalus in neurosurgery, treatments based on drug infusion for pain control and chemotherapy in oncology and anesthesia. Such developments are possible within the institutional collaboration framework of this project, with the participation of aggregate agents that complement from the practical, medical, and clinical points of view, the scientific analysis from the academic field.

From a technical perspective, this project is based on the availability of AI tools that allow the optimization of the simulations of the different scenarios and visualization systems using augmented reality devices. AI algorithms mainly belong to the “Machine Learning” type. In general, although they need extended data sets for training, they have a great advantage—for subsequent practical implementation—of their ability to improve by increasing the number and typology of processed cases. We also explore different AR devices to allow for intuitive, user-friendly visualization of the sonication volume, targeted structures, and the calculated fields of ultrasound energy and the corresponding thermal maps. Visualization devices differ in image quality, form factor, and ease-of-use.

In summary, described contribution relies on the combined use of AI technologies with AR devices for the optimization of parameters for the application, from outside the body, of focused ultrasound beams to achieve preventive cleaning of deposits adhered to the inner surfaces of shunt systems (valves, catheters) implanted in patients with various pathologies.

Acknowledgments

This project has been carried out in collaboration with FISEVI and University Hospital V. Rocio (Seville, Spain) and with the support of the following researchers: Manuel A. Perales-Esteve, Francisco J. Muñoz-Gonzalez, Desiree Requena-Lancharro, Isabel Fernández-Lizaranzu, Pedro Gil-Gamboa, Maria Jose Mayorga-Buiza and Mónica Rivero-Garvía.

Works Cited

1. ATTRACT Program. (2019). https://attract-eu.com/
2. Baraas, R. C., Imai, F., Yöntem, A.Ö. & Hardeberg, J. Y. (2021). Visual perception in AR/VR. Optics & Photonics News (April 2021), 34–41.
3. FUSCLEAN Project. (2020). https://phase1.attract-eu.com/showroom/project/combined-optical-imaging-and-ultrasound-focusing-for-hand-held-non-invasive-cleaning-of-implanted-cerebrospinal-fluid-shunting-devices-in-patients-of-hydrocephalus-initial-design-and-proof-of-concep/
4. Gómez González, E., & Gómez, E. (2020). Artificial Intelligence in medicine and healthcare: Applications, availability and societal impact. JRC Science for Policy Report (EUR 30197 EN), European Commission. https://doi.org/10.2760/047666
5. Nepelski, D. (Ed.). (2021). How can Europe become a global leader in AI in health? European Commission. https://knowledge4policy.ec.europa.eu/file/how-can-europe-become-global-leader-ai-health_en.

1.2 Assuring the Quality and Security of Medical Robotics Process Automation

María-José Escalona and José Navarro-Pando

Abstract

In recent years, the application of new technologies to the healthcare environment is a common practice. However, the COVID-19 pandemic has shown us that this application is a critical necessity for society. The application of disruptive techniques such as Artificial Intelligence or Machine Learning in the healthcare environment is something necessary but not sufficient. It is necessary to take another qualitative leap. This paper presents a reflection on the use of emerging technologies, such as blockchain or the robotization of software processes to further improve healthcare processes.

Introduction

In recent years, many lines of research and innovation have been developed, aimed at improving the health environment through ICT (Information and Communication Technologies) and, very specifically, through t...

Table of contents

• Cover
• Half-Title
• Title
• Copyright
• Dedication
• Contents
• Foreword Artificial Intelligence: The New Paradigm to Boost Society 5.0
• Editors
• Contributors
• Introduction
• 1 Health and Social Welfare
• 2 Energy Efficiency and Sustainable Construction
• 3 Digital Economy
• 4 Mobility Logistics and Advanced Industry Linked to Transportation
• 5 Endogenous Land-Based Resources, Agroindustry, and Tourism
• Conclusion
• Index