chapter nine
This chapter covers:
- A five-step approach to break down our large cancer detection problem into a series of smaller, easier ones.
- The criteria we’re using to select a project structure, and why simpler approaches aren’t ideal for our use case.
- Background information about cancer detection, the data formats used, and why those topics are important to understand when starting a new project.
As you might have guessed, the title of this chapter is more eye-catching implied-hyperbole than anything approaching a serious statement of intent. Let us be precise: our project in part 2 will be to take three-dimensional CT scans of human torsos as input, and produce as output the location of any suspected malignant tumors, if such exist. Prior to the arrival of deep learning, this problem was exclusively tackled by humans. Radiologists had to pore over every single image manually, looking for hints that some part of the patient’s body might be displaying signs of malignancy. Doing the search this way results in a lot of missed warning signs, particularly in the early stages when the hints are more subtle.