6CCYB020 Medical Imaging
Medical Imaging Coursework
Guidelines
This worksheet is assessed coursework. You should work individually and submit your own work. Please include and sign the following statement in your answers to acknowledge this:
I am agreeing to abide by the principles of the College’s Academic Honesty & Integrity Policy and the work I am submitting is my own.
You may refer to notes and online resources, but note that College rules for plagiarism apply and you should cite any sources in your answers.
You must submit your answers on KEATS by the deadline of 4pm on 3rd December 2025.
There are two questions on this coursework, relating to the practical sessions of Ultrasound and MRI. For both parts, you will need to have completed the practical by running and experimenting with the MATLAB simulations.
For each question, include at least one figure showing your final result from the practical, along with any other figures or diagrams that you find useful.
For each question, you should write no more than 1 page of A4 including diagrams, figures and references. Please do not overdo it.
This coursework is worth 30% of your overall module mark.
Coursework questions
Question 1 – Ultrasound B-mode image simulation
Describe each of the six parameters that you adjusted in the B-mode simulation to obtain a useful image. For each, explain its effect on the image and why you chose the final value.
- Centre frequency of pulse [4 marks]
- Bandwidth of the pulse [3 marks]
- ADC sampling rate [2 mark]
- Geometric focus of the array [2marks]
- Lateral dimension of the aperture [3 mark]
- Emitted pulse amplitude [1 mark]
[Total 15 marks]
Marks will be awarded for describing a working set of parameters and for demonstrating that you understand why these parameter values work.
Question 2 – Quantitative MRI task
The hidden function MRI_simulator.p can be run to generate simulated spin echo or inversion recovery spin echo images of a simulated phantom that consists of five separate sample tubes. The ‘ .p’ extension means that the code is protected and you can’t view it – you will have to treat this code as a ‘black box’ . An example image it produces is below:
Each sample has a different spin density p(r), T1 and T2 time. Images are of size 64x64
Calling MRI_simulator(TE,TR) returns a simulated spin echo image which has been calculated using the formula:
Calling MRI_simulator(TE,TR,TI) -- i.e. using three rather than two arguments – returns a simulated inversion recovery image calculated using:
The times are constrained as follows:
• 5ms
• TE
• 0
Calling MRI_simulator without any arguments prints out some help text
Notice that the images also contain some random noise.
Your task:
You are to ‘design an experiment’ to produce a T1 map for these samples. You can use the relations above to work out how to set the TE, TR and TI (where relevant) in order to make these measurements. Unlike the tutorial where you were given images, and then used fitting to determine T1 and T2 in this task you have to work out which images to generate first and then use fitting to estimate T1.
You can then use any code from the qMRI example (qMRI_tutorial.mlx) including the helper functions T1_fit_function.m and T2_fit_function.m as needed to fit these measurements to determine T1 and T2 of each sample.
Some important points to remember:
• The fitting functions T1_fit_function and T2_fit_function are not necessarily fitting the same formulae as above but may apply if you fix certain parameters, or set other parameters to be very large or very small
• You are free to use T1_fit_function and T2_fit_function or you can write your own fitting functions
• The simulated samples have different proton densities as well as different T1 and T2. We don’t need to estimate these accurately but you may need to take into account that they are not the same.
• The images include noise, and you need to choose enough samples to allow you to estimate T1 without it being too noisy
You will need to produce:
- a T1 map of the samples including colour bar for scale
- a list of the parameters used to generate your simulated images
- brief (one paragraph) justification for these parameter choices
Marks will be awarded for:
- Plausible T1 map (correct average and reasonable signal to noise ratio) [9 marks]
- Colour bar [1 mark]
- Justification for parameter choices [5 marks]