Neonates at the intensive care unit of a hospital are very vulnerable. They are very receptive for disorders like brain injury and sepsis. Physicians do not always know exactly what is happening and start general treatments to work with the current condition. UZA, UA and Innocense are developing an algorithm for early detection of brain injury and sepsis. Based on minute-to-minute and second-to-second vital sign monitoring our algoritme is, at this stage, able to detect brain injury and sepsis four hours earlier than the active care team on the ward. This allows the team to start treatment faster and more focused on the disorder. This results in less side effects, shorter length of stay for the neonate and less mortality.

Length of stay determination of ED patients

When you arrive at the emergency department, no one is able to tell you how long you will actually have to stay there. Since many different forms of urgency hit the emergency departments, a triage system is used to categorize them. Based on the outcome of the triage and the present population at the emergency your waiting time is determined and based on your general condition the treatment time is determined or even the decision if you will have to be admitted to the hospital or not. UZA has developed an algorithm than is able to take all these parameters and predict how long you will have to stay at our emergency department and what the chance is you will have to be admitted or not.

VDAB has already multiple AI projects in Production. Guido will present the Ethical AI Program and the different phases a company goes through to become industrialize AI.

In today’s political environment where global trade regulations change at a fast pace, organizations are reconsidering their trade compliance strategy and are looking for ways to optimize processes, minimize manual efforts and focus resources on driving strategic value. Classification is an important part of trade strategy as it determines amongst others, tariff and non-tariff measures, duty rate applicable to an item, etc.

By leveraging Artificial Intelligence (AI) and Machine Learning, the traditional resource intensive classification process can be augmented, accelerated and even fully automated. This will transform how classifiers work: moving from conducting manual classification to focus on reviewing, confirming and auditing classifications.

During this presentation we will talk about the business challenges in this context, how AI can help in managing your Trade Master Data and what the benefits are compared to the traditional approaches.