TU Wien:Visualisierung medizinischer Daten 1 VU (Raidou)

Aus VoWi
Zur Navigation springen Zur Suche springen
Ähnlich benannte LVAs (Materialien):

Daten[Bearbeiten | Quelltext bearbeiten]

Vortragende Daniel PahrRenata Georgia Raidou
Alias Selected Chapters from Medical Visualization (en)
Letzte Abhaltung 2023W
Sprache English
Mattermost visualisierung-medizinischer-daten-1RegisterMattermost-Infos
Links tiss:186105, eLearning, [see https://www.cg.tuwien.ac.at/courses/MedVis/VU/2023W Homepage]
Masterstudium Biomedical Engineering Wahlmodul Modul: Advances in Medical Physics & Imaging
Masterstudium Visual Computing Wahlmodul Advanced Visualization
Masterstudium Medizinische Informatik Pflichtmodul Visualisierung medizinischer Daten
  • homepage beginnt nicht mit http:// oder https://.

Inhalt[Bearbeiten | Quelltext bearbeiten]

  1. Image Acquisition
  2. Preprocessing
  3. Segmentation & Registration
  4. Surface Rendering & Marching Cubes
  5. Volume Rendering

Ablauf[Bearbeiten | Quelltext bearbeiten]

6 regular lectures of 2 hours

1 set of invited lectures of 3 hours (mandatory, -10 points if you miss it)

5 assignments during the semester

Project at end of semester in groups of 2-3 students including implementation, summary, presentation


Benötigte/Empfehlenswerte Vorkenntnisse[Bearbeiten | Quelltext bearbeiten]

Not necessary but helpful:

  • Basics of medical image acquisition (basic knowledge of modalities and their pros/cons)
  • Basics of visual computing (filters, edge detection, sampling, ...)
  • Python/C++

Vortrag[Bearbeiten | Quelltext bearbeiten]

in English

Übungen[Bearbeiten | Quelltext bearbeiten]

5 assignments with one week time (Assignment 4 has two weeks)

  1. Acquisition and Preprocessing: MeVisLab
  2. Segmentation: MeVisLab
  3. Surface Rendering: MeVisLab
  4. Volume Rendering: Python
  5. Transfer functions: ImageVis3D

W23: Four assignments and the project. The fourth assignment had two components (the second one was a creative topic)

Prüfung, Benotung[Bearbeiten | Quelltext bearbeiten]

  • Assignments 25 points
  • Project 30 points + 10 points presentation
  • Exam 35 points
  • Active participation 5 bonus points

at least 50 % of each component to pass

> 88 points 1
> 75 points 2
> 63 points 3
>= 50 points 4

WS21: exam oral due to COVID

WS23: written exam + oral repeat if you fail or miss the written one

The written exam comprised only open-ended questions from all lecture units, and it was split into two parts. Part one was theoretical, and part two contained one large and a few smaller practical examples to solve. Most questions in part one were split into smaller sub-parts that each rewarded a small number of points (mainly 0,5 to 1 point per task), so it's no big deal if you couldn't answer one or two of them. The questions were all aimed at checking understanding rather than asking formulas or definitions. There was only one question where you needed to recite a slide from the first lecture (the goals of visualization) but without adding any details. Part two contained three practical examples. In the first one, an application example was given, and you had to choose suitable methods for solving the problem and explain your choice. Then there were two more smaller examples. In the first one, you needed to calculate a few pixel's intensity values using two compositing methods (MIP, average), and also explain what the methods do, how they work, and mention some pros/cons. The second practical example was less practical. A few color maps were given (greyscale, diverging colors, rainbow, single color with hue low to high) and you had to describe when you would/wouldn't use them and what their benefits and drawbacks are. All these were handled in the lecture. See here for more details on the examples.

Dauer der Zeugnisausstellung[Bearbeiten | Quelltext bearbeiten]

W23: Grades after 3 days, certificate 9 days after the exam.

Zeitaufwand[Bearbeiten | Quelltext bearbeiten]

very high: almost 6 ECTS

Everything has almost twice the effort.

WS21: 4,2 ECTS

WS23: Moderate to high (more than 3 ECTS), depending on how much you do for the exercises and project

Unterlagen[Bearbeiten | Quelltext bearbeiten]

Slides and recordings of lectures (except invited lecture) in TUWEL

Based on the book of B. Preim and C.P. Botha "Visual Computing for Medicine" (2nd ed.) (http://medvisbook.com/)

Tipps[Bearbeiten | Quelltext bearbeiten]

Start early with the project & homework assignments

The vtk/mevislab documentation pages can be pretty challenging to work with, but some topics have good tutorials on youtube that you can follow 1:1 to solve the homework

Highlights / Lob[Bearbeiten | Quelltext bearbeiten]

The project topics were interesting & engaging

The invited lectures were also interesting

Verbesserungsvorschläge / Kritik[Bearbeiten | Quelltext bearbeiten]

Skip the presentations of the project, as one already has to write a paper.

The grading of the presentation was a bit harsh (you could easily lose 50% of the points for presenting just a bit too long)