Based on research by Leo Petersen-Khmelnitski

When looking at the AI solutions that are already available as well as the pipeline of ideas in healthcare-related AI, three phases of AI adoption in healthcare can be distinguished.

First, AI solutions are likely to address routine, repetitive and administrative tasks that otherwise take time of doctors and other healthcare personnel. In addition, AI applications related to image recognition and analysis are already in use in specialties such as radiology, pathology, and ophthalmology.

The second phase is likely to encompass AI solutions that support the shift from hospital-based to home-based care, such as remote monitoring, AI-powered alerting systems and virtual assistants. This phase includes a broader employment of NLP solutions and a wider use of AI in more specialties such as oncology, cardiology or neurology.

In this phase, AI will be embedded in clinical workflows and thus successful adoptions in this phase require not only a combination of the key concepts in AI (e.g. deep learning and NLP), but just as importantly a cultural change and capacity building within healthcare providers.

The third phase of AI adoption in healthcare focuses on improved clinical decision-support (CDS). AI is set to become an integral part of the healthcare value chain, such that broader data sets are integrated across healthcare providers. There will be continuous efforts to improve data quality and regulations will be in place to ensure that organizations, practitioners, patients and the public can have confidence in both the AI solutions and their ability to manage the associated health and data risks.