Microsoft director introduces completely new concept for American healthcare28. October 2019
Microsoft director introduces completely new concept for American healthcare
New York, 29.12.2010
In the current issue of TIME, Geralyn Miller, Director of AI and Research at Microsoft, presents a concept for the American healthcare of the future. It is not only perfect but also Nobel priced.
Miller initially asks readers to imagine a “not-so-distant future” where “the focus is no longer on disease, but on how we stay healthy.” With that she provokes a series of questions: The key is probably: is that even possible?
No doubt, yes, because Miller would get a thorough, multi-faceted initial profile right after birth, including screening for genetic and rare diseases. Then, cost-effective, minimally invasive clinical-use devices could accurately monitor a range of biometric data such as heart rate, blood pressure, temperature, and glucose levels over their lifetime, in addition to environmental factors such as exposure to pathogens and toxins, behavioral factors such as sleep and activity patterns. These biometric, genetic, environmental, and behavioral information could be linked to social data and used to create AI models. In this way, the risk of disease could be predicted, an early warning of life-threatening conditions such as stroke and myocardial infarction triggered and warned of possible adverse drug reactions “.
But that is not the end. Because every American can assume that his health remains under constant control. And this goes like this: “Smart bots could be integrated into their home / apartment via digital assistants or smartphones to screen, educate, and advise on symptoms and ensure that they comply with drug regulations.”
But Miller not only thinks about the potential sick but also about the often overburdened doctors. She says Artificial Intelligence could also reduce doctors’ burn-out and increase the reach of doctors in underserved areas. The following options are already available: AI writers could help doctors take clinical notes, and bots could help medical expert teams get together and discuss challenging cases. Computer vision could assist radiologists in detecting tumors and dermatologists in identifying skin lesions. It could also be used for routine eye exams. Miller emphasizes, “All this is possible with the technology available today or in development”.
Because such far-reaching changes can not bring about AI alone, Miller calls for us to “undertake a social transformation that includes trusted, responsible and inclusive policies and governance on AI and data. That is only possible if there is effective cross-sector cooperation, comprehensive training for the public, professionals and civil servants. “
These conditions are particularly relevant to health care, because it is inherently complex. Even a misstep can have serious consequences. One challenge will undoubtedly be to “reconcile individual rights with the health and safety of the general population and to find out how resources can be equitably and efficiently distributed across geographic areas”.
The crucial starting point for their concept is data. She explains: “So we need to invest in the creation and collection of data while ensuring that the value created by the use of this data falls to those who are data. To protect and preserve the integrity of this information, we need trusted, responsible, comprehensive regulatory and regulatory policies and a governance framework. “
As an example, Miller calls the DSGVO (General Data Protection Regulation). In the EU, the GDPR came into force in May 2018 and is already helping to ensure that the health care industry deals responsibly with information from individuals.
It is also clear that even commercial companies could not solve these problems on their own because “they need partnerships with authorities, universities and non-profit organizations”.
From all this follows for them: “We must ensure that our computer scientists, data scientists, physicians, lawyers and policy makers through relevant training will acquire the unique skills of AI and have an understanding of the risks. This type of training can take place in professional societies such as the American Society of Human Genetics and the American Association for the Advancement of Science, which have the necessary reach and infrastructure. “
Diversity is important. By that she means “a variety of developers who write the algorithms, a variety of data scientists who create the models, and a variety of the underlying data. This means that in order to be truly successful with AI, we must overlook the things that historically characterize us, such as race, gender, age, language, culture, socioeconomic status, and expertise. It will not be easy in the face of this story. But if we want the full potential of AI to be used to address the urgent needs of global health care, we need to do that. “