04450 - Medical Image Analysis
Medical image analysis is an important base technology
for developing diagnosis support systems and as such
it will be illustrated through lectures, case-studies,
and exercises. The course will also give an introduction
to central aspects of clinical testing, including
bio-statistical methods. Quality assurance is a central
issue in the development of a product specifically
in relation to regulatory approvals. ISO standards,
CE and FDA approvals will be discusses through examples.
- Link
to DTU course 04450
10445 - Bio-medicinsk optik
Interessen for anvendelser af kohærent lys og optiske
sensorer inden for de biologiske og medicinske områder
er stærkt stigende i disse år. Dette skyldes dels
udviklingen af nye avancerede lasersystemer, dels
udviklingen inden for integrerede optiske sensorer
og endelig de stærkt stigende forventninger om præcis
og effektiv diagnosticering og analyse af biologiske
og medicinske processer. Kurset er opdelt i fire hovedområder:
1. Optiske metoder til billeddannelse i humant væv,
2. Nye lasersystemer til medicinsk diagnostik og behandling,
3. Optiske biosensorer og non-invasiv spektroskopi
i øjet, 4. Biomedicinsk billedbehandling.
- Link
to DTU course 10445
04364 - Nonlinear Signal Processing
The lectures will be chosen among the following topics:
Neural Networks. An introduction to the theory of
neural networks is given. The theory will be illustrated
by applications in a number of areas, including the
biomedical area. Speech Coding. The point covers an
introduction to the principles of speech coding with
emphasis on coding at low bit rates (< 10 KBPS).
The theory will be illustrated by applications in
the areas of multimedia, telecommunication and aids
for the handicapped. Speech Recognition. The point
includes methods for recognition of speech, especially
methods based on Hidden Markov Models and/or neural
networks. Applications are equivalent to the topics
mentioned above, multimedia, telecommunication and
aids for the handicapped. Vector Quantization. Vector
quantization is a general discipline for effective
coding of signals, including speech signals and biomedical
signals. Pattern Recognition. Bayesian decision theory
and mathematical modelling of pattern recognition
systems
- Link
to DTU course 04364
04365 - Advanced Digital Signal Processing
The course starts with lectures on current research
subjects, e.g., prediction of time signals, neural networks,
hidden Markov models, wavelets, parametric spectral
estimation, vector quantization, speech coding, noise
reduction in speech signals, linear adaptive filters,
filter banks, implementation of signal processing algorithm
on fast signal microprocessor. The participants carry
out project within the research subjects presented.
- Link
to DTU course 04365 |