Lecture#2: DEVICE INFORMATICS - applications in Public Health

#MedicalDevices #AI #LLM #ML #Public #Health

We are in the midst of an artificial intelligence (AI) revolution. It may seem that AI development is recent, but the term AI was coined (1955) not too long after the invention of computing devices. However, the idea that a machine can behave like a human being is even older. The term “automaton” was used for it. Please see:

• https://www.tableau.com/data-insights/ai/historyLinks to an external site.
• https://en.wikipedia.org/wiki/Logic_TheoristLinks to an external site.

What has changed, though, in recent times, is the amount of computation that can be performed. GPUs have made tremendous progress, and it is now possible to do heavy computation with very large amount of memory in real time.

The terms AI and machine learning (ML) are normally used interchangeably though there are differences. We will discuss what is the difference between them.  While AI tries to mimic the human behavior, ML is a set of statistical tools to look at the data to find patterns without direct instructions. In some ways, it is a subset of AI. We will dig a little deeper into how AI works. This will give us a better understanding of what kind of problems it can solve effectively and where it should be avoided, or one needs enhanced or use better tools.

Neural networks are normally used to perform AI calculations: Deep Neural Networks (DNN), Convolutional Neural Networks (CNN), Recurrent Neural Networks (RNN) are some examples. They have their origin in the way the human brain works.

We will discuss the biological basis of neural networks and discuss how these networks are implemented. We will focus on concepts and will develop intuition about how they are trained. We will also see how and when a certain type of network can be used.

We will go over some caution around data analysis and what to watch out for when using AI.

In this lecture, we will take this discussion further. AI/ML is finding more frequent use in medical devices especially wearables. We will briefly discuss how exactly same data can result in very different outcomes and conclusions. We will then discuss differences between making and shipping a generic consumer device and a medical device - why medical devices cost more both in terms of time and money. Finally, we will consider a (imaginary) medical device and take it through the process of designing, validating, and getting it through the FDA for pre-market approval. In this respect, we will review different types of FDA approvals, and strategies for it.

• Add Event to Calendar
  iCal
  Google Calendar

Loading virtual attendance info...

• 5270 California Ave
• Gora Datta
• Irvine, California
• United States 92617
• Building: Bealle Applied Innovation
• Room Number: Emerald Cove

---

• Contact Event Hosts

• Starts 16 October 2025 07:00 AM UTC
• Ends 24 October 2025 08:00 PM UTC
• No Admission Charge

• 

• 14:00 to 14:10 PM : Welcome to IEEE OC EMBS and general introduction
• 14:10 to 14:20 PM : Introduction by Gora Datta, FHL7
• 14:20 to 15:50 PM: Expert Lecture by Dr Md Usman

  • The difference between AI and ML,
  • Biological basis of some AI networks,
  • Basics of how an AI model is trained and what does it mean,
  • Advantages and disadvantages of using AI – why should we use it, and what does it measure anyways?
  • What are large language models (LLM), and their usefulness in medicine and medical devices
  • Will understand some examples of using AI, and
  • Know when to use AI for health measurements.

• 15:55 to 16:00 PM : Wrap Up