Practical Course:
Machine Learning in Medical Imaging

Prof. Nassir Navab, Dr. Shadi Albarqouni, Roger Soberanis, Ashkan Khakzar, Walter Simson, Magda Paschali, Azade Farshad

Type: Master Practical Course Module IN2106
SWS: 6
ECTS: 10 Credits
Location: Seminar Room, 03.13.010
Time: Thursdays, 16-18
Course Language: English

Announcements:

• 23-09-2019: Projects are assigned! Please contact your tutor this week to define a regular weekly meeting.
• 13-09-2019: Projects are announced! Please, send us your preferences (at least three preferences) by email to mlmi@mailnavab.in.tum.de before September, 20th. Please make sure the subject of your email is "MLMI_Projects_Preferences".
• 05-07-2019: Preliminary meeting: Thursday, 18.07.2019 (12:00-13:00) in CAMP Seminar Room, 03.13.010.
• 29-06-2019: Website is up!

About the Course

The aim of the course is to provide the students with notions about various machine learning techniques. The course is subdivided into a lecture/excercises block and a project.

• The lectures will include topics in image processing and analysis, unsupervised methods, and supervised methods. Each lecture will be followed by a practical hands-on exercise (e.g. the implementation in Python).
• The topics of the projects will be distributed at the beginning of the semester. Each topic will be supervised by a different person. The projects are to be realized by couples. The programming language for the projects is to be decided in agreement with your supervisor.

Registration

• Interested students should attend the introductory meeting to enlist in the course.
• Students can only register through TUM Matching Platform themselves if the maximum number of participants hasn't been reached (please pay attention to the Deadlines).
• A maximum number of participants: 20.

Requirements

• Good knowledge of machine learning methods and terminology: knowledge of the meaning of the words classification, regression, data-set, cross-validation ...
• Strong knowledge of Python programming language
• Background in image processing

Schedule

Date	Topic	By	Slides	Homework
18.07.2019	Preliminary Meeting	Shadi	Slides
Online	Project Assignment	Shadi
24.10.2019	Intro. to DL (Tentative)	Shadi
31.10.2019	Neural Networks (NN)	Roger	Slides	Hands-on
07.11.2019	Convolutional Neural Networks (CNN)	Ashkan	Slides	Hands-on
14.11.2019
21.11.2019	Generative Models (CAE, VAE, GANs)	Azade	Slides	Hands-on
28.11.2019 (tentative)	Recurrent Models (RNN, LSTM)	Anjany?
05.12.2019	Robustness	Magda	Slides	Hands-on
12.12.2019	Intermediate Presentation	all tutors	Guidelines
19.12.2019
06.02.2020	Final Presentation Location: MI Hörsaal 2	all tutors	Guidelines

Evaluation

• 60% Project Progress (Done by your tutor -- mainly on your weekly progress on lrz git repository)
• 40% Intermediate and Final Presentation (Done by all tutors -- mainly on your presentation skill, progress so far compared to other groups ...etc.)
• Presentation: Use the CAMP templates for PowerPoint camp-tum-jhu-slides.zip, or Latex: CAMP-latex-template.

Projects

ID	Project	Tutor	Group I	Group II	Material
1	Comparison of Uncertainty Estimation techniques in Deep Learning for Medical Applications	Agnieszka and Shadi	Andrei Diaconu, Springer	Abdelhamid, Esposito	Project Description
2	Nerve Segmentation and Classification on Ultrasound and Optoacoustic Imaging	Magda and Walter	Flynn, Giri	Yagubbayli, Alamleh	Project Description
3	Weakly supervised medical segmentation: How to save the annotation cost for training deep segmentation network?	SeongTae			Project Description
4	SIIM-ACR Pneumothorax Segmentation Challenge	Ashkan and Roger	Jahiri, Beheim	Ergin, Faizan	Project Description
5	Robustness in Wasserstein Generative Adversarial Network	Azade	Naeem, Karaoglu	Khanduja, Deo	Project Description
6	Deep learning with adaptive feature acquisition	Gerome	Raj, Bernecker	Elsharnoby, Mullakaeva	Project Description