Online Magazine
TechTalk Audio: Responsible AI in healthcare
How does biased data affect AI-based diagnosis of a skin rash? And what regulations do hospitals and pharmaceutical companies need to adhere to when it comes to the responsible use of AI? Rudraksh Bhawalkar, Global Delivery Lead for Accenture's internal Responsible AI programme, answers these and other questions around responsible AI in the fourth episode of TechTalk Audio on responsible AI.
Tobias Imbach spoke with Rudraksh Bhawalkar
To the point:
- In healthcare, AI is used to diagnose diseases, for treatment purposes and in drug development.
- We want to make sure that an AI that makes a decision about a person's life without human intervention processes the relevant data responsibly – that it doesn't misdiagnose or prescribe the wrong medication.
- For example, an app that uses image recognition to diagnose skin rashes needs data from people with different skin colours so that a reliable diagnosis can be made for everyone.
- Unbiased data is also important for medical statistics because those are used to make medical predictions and diagnoses.
- The Health Insurance Portability and Accountability Act (HIPAA) already regulates the handling of data. Further regulations relating to AI are in the works and are expected to be enacted in 2024.
- For hospitals, responsible data handling has two advantages in particular: First, they comply with the law, and second, they build a trustworthy reputation.
HOW INNOVATIVE MINDS USE AI IN THE HEALTHCARE SECTOR:
A start-up uses AI to speed up disease diagnosis: Listen to the podcast.
Viktoria Prantauer wants to fight breast cancer with more data: Read more in this article.
AlphaFold has revolutionized computer-assisted drug discovery: Read the article here.
