How artificial intelligence, robotics and Big Data can support pandemic situations
Head of Artificial Intelligence and Robotics. Atos Research and Innovation
Artificial Intelligence and Robotics Associated Head. Atos Research and Innovation
Head of Health Unit. Atos Research and Innovation
Posted on: 17 April 2020
Governments, healthcare authorities and staff are working hard to fight and overcome the Covid-19 situation worldwide. Beyond that, innovative technologies can be a powerful additional resource to support the current efforts – let us explore some of these opportunities:
Predictive models for the spread of infectious diseases
The development of models that provide short and medium term forecasting of the number of new cases is a powerful tool to estimate the need for clinical resources and personnel, especially when healthcare systems are overloaded.
In today’s situation, daily growth rate models are being published, relying on statistical models or time-series forecasting using Long Short-Term Memory (LSTM) neural networks. Further, at the beginning of the crisis, Big Data and AI techniques like Natural Language Processing have been used to identify and alert early about the new virus.
We have observed that most of the current models rely on the number of daily cases, but do not take into account the specific characteristics of each region or other relevant variables that may influence the evolution of the disease. For example, clustering regions can help create more accurate and specific predictive models. AI and Big Data can also be used to create accurate and detailed geospatial spread models, complementing the capabilities of traditional epidemiological compartment approaches.
Risk stratification and personalized treatments
The publicly available datasets for Covid-19 do not include all the needed clinical information regarding patients’ preexisting conditions, such as age, sex, region etc. While worldwide healthcare organizations state that the disease affects mainly elderly people or those with underlying health problems, there is a notable number of deaths outside of these higher-risk groups.
Electronic health records would help find hidden correlations and more finely-tuned risk stratification, resulting in a map of clinical and non-clinical factors that can enable implementation of appropriate mitigation actions while protecting the most vulnerable groups. However, most hospital information systems worldwide do not share clinical data among each other, therefore becoming silos of independent data. This barrier can be overcome with the use of eHealth standards and medical terminologies with a two-fold objective:
- Provide semantic interoperability (that is, common descriptive terminology)
- Build a global dataset with information from different countries, and with much more details on the effectiveness of experimental treatments. This can enable medical staff to prescribe personalized treatments according to the specific conditions of the patient, and also avoid conflicts and controversies with respect to the adverse side-effects of some drugs.
Robotic platforms to assist isolated patients
Hospitalized patients in isolation do not have the possibility to receive visits or be accompanied, due to the strict restrictions to stop the spreading of the virus. Moreover, in many cases, the overload of healthcare systems reduces the time that clinical staff can dedicate to each patient. Therefore, they spend most of their time alone which may further cause physical and psychological deterioration.
Many reports and scientific studies already warn of the negative impact effect that social distancing and isolation measures have on mental health. Cases of anxiety, depression and stress may also result in a post-traumatic condition after the end of quarantine periods.
The usage of autonomous robotic platforms can be helpful to extend the capabilities of hospitals in crises, in order to offer continuous interaction and personalized attention to isolated patients. Those robots could incorporate capabilities to recognize and adapt their behavior to human emotions and moods, and to naturally interact with patients. While performing non-critical routine tasks (e.g. food delivery), they could offer an alternative source of entertainment and distraction to the patient, monitor non-clinical parameters such as emotional state and level of activity, and report unsafe or at-risk situations or behaviors.
Robotic platforms for disinfection
As recently seen in China, autonomous robotic platforms for disinfection of healthcare facilities can play a fundamental role, by preventing and reducing the spread of infectious diseases, virus, bacteria and other types of harmful organic microorganisms in the environment. They are particularly effective when equipped with ultraViolet (UV) light emitters which break down the DNA-structure of viruses. In other cases, robots and drones have been used to spray sanitizers in public spaces to mitigate the possible expansion of the infection.
Although most of these robots have some autonomous capabilities they must generally be triggered or even controlled by human operators, which strongly reduces the benefit of automation. For instance, in the case of UV disinfecting robots, cleaning staff must use an application to order an actuation, and drones must be typically piloted by humans. To deliver the full potential of robotic platforms, semi-autonomous and distributed decision-making must be implemented in the robots, so that they can operate in a standalone way with minimum supervision, following the high-level goals defined by human authorities.
Global pandemic situations suppose a challenge that affects all aspects of modern societies. The technological progress of the last decades in areas like Big Data, artificial intelligence or robotics systems must be used to support the ongoing efforts to control and overcome the spread of the disease, to mitigate its economic consequences and to prevent the appearance of new outbreaks.