Day 1: Masterclass Deep Learning
After an introduction, participants of SSIMA:Tech got an introduction in deep learning. We started with the neuro-mathematics of deep learning and ended with the layers of a neural network.
Day 2: Unsupervised Learning in AI
My favourite lecture this day was about unsupervised learning. We learned about eigenframes and eigenvectors. Further, the object of interest was a big topic. Objects have consistent, well-formatted, symmetric properties. Shape, appearance, and motion are very important in perceptual grouping.
Day 3: Medical Image Computing
The most interesting thing I've heard on this day was the word "overdiagnosing". It means, that cancer is detected very early, which sounds very good at the beginning. But, by treating a person a major change in life quality is done as the treatments have a lot of side effects. The conclusion here is that we wouldn't need to cut out so much. For more information, I can recommend the blogpost: Overdiagnosis: when finding cancer can do more harm than good.
Day 4: MedTech in hospital and gala dinner
This day was all about best practices. Speakers from Israel, the UK and the Netherlands were talking about their experiences in hospitals. In the evening, we had a gala dinner with traditional Romanian music in the Casa Oamenilor de Stiinta (house of scientists).
Day 5: TensorFlow - Keras programming course
On the last day, we had a tutorial class, where we learned to analyse medical images with the help of TensorFlow and Keras.
Besides the lectures, there was time for networking. I am glad to meet all these people at SSIMA who helped me with my neverending questions regarding my bachelor's project about the image analysis of cells.