Welcome back to this third and final course of this specialization. This is a multi-part course where we will cover a few applications of AI, first to medical treatment and then to information extraction. One goal of personalizing medicine is to estimate the effects of particular treatments on a given patient. In the first week, you learn how to build and evaluate models for estimating treatment effects. The first application you learn about in week two is question answering. If, as a patient, you wanted to look up the negative side effects of a particular treatment, you can ask that question to an AI model and get an answer to your question. Specifically, this model can read a medical abstract and answer basic questions from it. You get to build a system using one of the biggest recent advances in the application of deep learning to text, the BERT model. In week two, you will also learn how you can automatically have a machine read radiology reports and extract the mention of diseases from those reports. In course one, you had seen how you can build a chest x-ray classification model from labeled data. In this course, you will learn how you can get to that label data from unstructured text in radiology reports. Finally, in week three, you learn about how to interpret machine learning models. Explainability is important for medical practice. If an algorithm renders a diagnosis, can it explain that diagnosis to a doctor? I think more interpretable or explainable models will be important to accelerate AI's adoption in medical practice. We will switch gears and cover information extraction. There's a vast amount of unstructured text that is being generated today as part of electronic medical records and the medical literature. We'll look at techniques that make use of all of that unstructured data. We'll look at different methods of interpreting models. We'll look at both the interpretation of deep learning models you've built in course one and machine learning models you've built in course two. For instance, you will learn about methods that you can use to determine how different features are contributing to a model's decision. You'll apply those to the machine learning models you built in course two. You'll also learn about interpreting the decisions of a deep learning algorithm by visualizing its hidden representations, and you'll apply that to the chest X-ray algorithm you built in course one. So just one more course to completing this specialization. After you finish this course, you'll be on your way to become an expert in AI in medicine and we'll have seen applications of AI to medical diagnosis, prognosis, and treatment. Let's dive in.