The first ATHIKA symposium focused on the identification and analysis of challenges in the e-Health sector, specifically concerning the Internet of Things (IoT), artificial intelligence (AI) and data analytics, and governance and ethics. As such, there were panels on each of these topics as well as an opening presentation on the ATHIKA project and its connection to smart health and the development of smart cities.
The panel on the Internet of Things included insights from a variety of players in the healthcare IoT industry. The panelists presented examples of the IoT working to produce better health outcomes, including a noncritical patient remote monitoring system and a social/smart housing system. However, these IoT solutions to traditional healthcare issues are not without problems. Three of the main challenges identified in this sector were how to reorganize healthcare industry interactions to work with the IoT, the cost and quality balance of IoT technology, and the different paces of innovation and adoption within the sector. Reorganizing interactions within the healthcare industry presents a challenge as the Internet of Things has the potential to eradicate the current doctor-patient-hospital relationship. There must be a framework in place to handle the new doctor-patient-hospital relationship that will result from the widespread use of IoT technology in healthcare. As with any other emerging technology, there exists a key challenge in finding the correct balance between the cost and quality of IoT technology applied to e-Health. It is up to those in the industry, both businesses and researchers, to compete and produce IoT technologies that are high-quality and cost-effective. Without the correct balance in these regards, IoT in e-Health will be dead in the water. Arguably, the most critical challenge of IoT and other technologies for e-Health is the vastly different pace of innovation and adoption. Adoption, which includes both regulatory approval and widespread acceptance by the public, takes much longer than innovation. As a result, in the time it takes to adopt an IoT device or technology in the healthcare industry, a new, more efficient and effective version of the same solution has most likely been developed. The way to combat this issue is to increase the speed of regulatory approval and public adoption so that the industry is able to take advantage of the greatest technology available at the time.
The panel on artificial intelligence and data analytics included insights from experts in both industry and academia. The panel began with the presentation of an existing solution for the need of personal health data for scientific studies looking to improve patient outcomes. The Salus Common Good Data License allows members of the public to share their personal health data with researchers without fear of commercialization or loss of privacy. Data analytics for e-Health depends on the generation of usable personal health data and access to it. This dependence, as well as other things, contribute to the challenges currently facing data analytics for e-Health. The challenges identified for data analytics include the generation of high-quality, standardized health data, and the privacy of the patients generating the data. The ownership of personal health data presents a challenge because it has the potential to promote research at the expense of privacy or to promote patient privacy at the expense of crucial scientific studies. However, the panel also dealt with artificial intelligence in healthcare as the development and implementation of AI systems require the use of data analytics. AI is mainly used in healthcare for diagnostics and for performing time-consuming administrative tasks. The majority of challenges in this area are concerned with the use of artificially intelligent systems for disease diagnosis. The challenges identified by the panel include the adversarial vulnerability of AI systems, the black-box nature of these systems, and fostering a healthy relationship between doctors and artificial intelligence. Adversarial vulnerability refers to the drastic differences in output determination that results from targeted, imperceptible perturbations to the input image. For example, the application of a slight distortion to an image of a benign tumor can cause the neural networks that make up the AI system to classify the new image as a malevolent tumor when the unaltered image was correctly classified. This is a challenge for the developers of these systems, but it must be addressed to protect end users from malevolent attacks and to protect the accuracy of the system’s diagnoses. The black-box nature of AI systems presents a major challenge because the public has a right to know what went into the determination of their diagnosis. Currently, the inner workings of artificially intelligent systems are imperceptible to almost anyone except their developers. The e-Health industry must move towards explainable AI as doctors and patients need to know how the systems are making decisions that have significant impacts on their lives. Along this same line exists the challenge of creating a healthy relationship between AI systems and doctors. This relationship is necessary so that doctors are able to trust in the diagnoses made using artificial intelligence and endorse them in front of the patients receiving the results. This challenge is tied into the eradication of AI’s black-box nature but will also be addressed as time progresses, doctors gain more experience with artificial intelligence, and they see the improved health outcomes that result from the implementation of AI systems.
The panel on ethics and governance focused on the engineering mindset, the need to go beyond healthcare to promote individual wellbeing, and ethical access to health data for scientific research. The mindset instilled in engineers in the past has been one of purely utilitarian problem-solving. The issue with this mindset is that it fails to recognize the creation of new problems from solutions until they are the most important thing to be dealt with. The challenge here involves changing engineering education to move towards a deontological mindset that promotes awareness of the problem-solution interdependency that is critical to effective engineering, especially in the e-Health sector. The need to go beyond traditional healthcare for total wellbeing included discussions of the Pulse Project and the concept of Health in All Policies (HiAP). The Pulse Project is an attempt to transform public health into a predictive system based on both health sector factors and social determinants of health. A key takeaway from the project is that only 15% of health is linked to healthcare while the majority is tied to social, economic, and individual factors combined with individual behaviors. The challenge here is changing governance to take into account the wide range of factors that determine public health. One possible and promising solution is the concept of Health in All Policies. HiAP is a strategy for improving public health by incorporating health considerations into the policy decisions made in every level of governance. The relationship between e-Health and a wider approach to public health is crucial for the future of the health sector. Another critical aspect of the future of the health sector is the use of personal health data for scientific research. The challenge in this area is making sure governance is able to guarantee the ethical acquisition and use of this data by researchers. An example of a current solution to this problem is the PADRIS Program that has been undertaken by the Government of Catalonia. The program´s mission is to make health data available for research and innovation given that those receiving the data comply with ethical standards including respect, transparency, and protection of sensitive data. It is the prerogative of governments to provide a regulatory framework for the ethical use of data in the e-Health sector so that commercial actors or other parties are unable to take advantage of citizens. The challenge to be addressed is how to provide this framework and how governance should be used to promote ethics in the e-Health sector while also fostering research and innovation.