M.Sc. in Medical Imaging Nuclear Medicine 2014-15 Supplemental Examination 11:00 – 13:00 Friday 14th August 2015 Physics, Instrumentation and Computer Technology A total of 3 Questions MUST be answered All parts of a chosen question must be answered Each Question carries equal marks Please use a separate answer book for each question. Print the number of the question on each answer book. Question 1 Image Processing can be used to alter “raw” image data in such a way as to highlight image details that may not be easily visible in the original image. Discuss in detail the following Image Processing techniques: a] Histogram Equalisation b] Maximum Intensity Projection [MIP] [10 marks] [10 marks] Sample Answer and Marking Scheme a] Histogram Equalisation • Histogram Equalisation is a form of Contrast Enhancement [in fact it is Optimum Contrast Enhancement. [1 mark] • Definition of a image Histogram – “a graph showing the number of occurrences of each pixel value in the image”. [1 mark] • Figure showing a typical Histogram with proper labels [Frequency/No. Pixels] vs. [Pixel Value]. [2 marks] • In Histogram Equalisation Pixel Values are re-mapped to give a completely flat histogram. [1 mark] • Figure showing the flat Histogram of a processed image with proper labels. [2 marks] • A major disadvantage is the computation time – a lot of calculation – but is often combined with Region Of Interest [ROI] analysis to a specific area of the image. [2 marks] • In practice the resulting Histogram is never completely flat because of discrete nature of the image data, but it is always flatter than the original image, giving optimal contrast enhancement. [1 mark] b] Maximum Intensity Projection [MIP] • “Rendering” is a process of generating an image from a model i.e. by computer programming. [1 mark] • MIP is used in architecture, video games, simulators, movies, and MRI [an equivalent technique called Minimum Intensity Projection is used in CT]. [1 mark] • MIP is used to display 3D information on a 2D device i.e. computer monitor [1 mark] • MIP is used in Magnetic Resonance Angiography [MRA] [for example] where there usually is a stack of ‘slices’ representing a 3D volume in the body. [1 mark] • With MIP, the computer simulates rays through the volume [slice by slice] for a certain projection, and selects the highest pixel value on each ray for display. [1 mark] Page 2 of 9 Abdomen Slice Simulated Rays for this Projection [2 marks] • • • The resulting image resembles a conventional catheter angiographic image. [1 mark] The projection is slightly changed, and the process repeated. [1 mark] The resulting set of projection images are combined into a cine loop, giving the observer a good impression of the 3D structure. [1 mark] Page 3 of 9 Question 2 a) Radioactive decay is described by the following exponential decay equation: Nt = N0 e-λt where: N0 is the initial number of nuclei, Nt is the number of nuclei remaining after a time t, and λ is the decay constant. Derive this equation from first principles. [10 marks] b) At 9am a 99mTc source has an activity of 1200 MBq. What is its activity at 5 pm ?, [Take the half life of Technetium to be 6 hours] [10 marks] Answer Plan a] dN/dt α N dN/dt = - λN [- means N is decreasing with time] ∫ dN/N = ∫ [-λ dt] [2 marks] [2 marks] ln N = -λt + constant at time t=0, initial number of nuclei is N0 hence constant = ln N0 ln N = -λt + ln N0 [2 marks] ln N - ln N0 = -λt N = N0 e-λt [2 marks] Graph of exponential decay with labels [2 marks] Page 4 of 9 b] N = N0 e-λt or, Activity[t] = Activity[0] e-λt t1/2 = [ln [2] ] / λ [2 marks] or λ = [ln [2] ] / t1/2 λ = 0.693 / [6 hrs] = 0.116 hrs-1 [3 marks] time interval 9am to 5pm => t = 8 hours Activity[t=0] = 1200 MBq [2 marks] using Activity[t] = Activity[0] e-λt Activity[t = 8 hours] = [1200MBq] exp[ - [0.116 hrs-1] [8 hrs]] Activity[t = 8 hours] = [1200MBq] exp[- 0.928 ] Activity[t = 8 hours] = [1200MBq] [0.395] Activity after 8 hours is ~474 MBq Page 5 of 9 [3 marks] Question 3 a) In Gamma Camera Imaging, list the factors which may cause a reduction of image quality and also detail the resultant effect on the image. [7 marks] b) For the energy spectrum of photons incident on a gamma camera, outline some of the mechanisms which cause the energy spectrum to be ‘smeared’ or spread out. [5 marks] c) Outline how additional extra peaks are formed in the energy spectrum. [3 marks] d) Briefly describe how detected scattered photons can be rejected using pulse height analysis. [5 marks] Answer Plan a) Factors which cause reduction of image quality in imaging [1 mark for each correct, extra 0.5 mark for 3]: • Attenuation of radiation in the patient • Scatter of radiation in the patient • Detection efficiency • Spatial resolution Resulting in… [1 mark for each correct, extra 0.5 mark for 3]: • High image noise • Poor resolution • Low contrast • Reconstruction artifacts and distortions b) [1.5 mark for each, extra 0.5 for 3 correct, total of 5 marks]: • Random variations in the number of light photons per kV of gamma ray absorption • Random variations in the number of light photons converted to photoelectrons • Random variations in the number of secondary electrons emitted at each dynode • Variations in the PMT kV • Electronic Noise c) [1 for each of the below, total of 3 marks] • Backscatter from surrounding shielding material • Iodine escape • Lead Characteristic radiation Page 6 of 9 d) [Total of 5 marks] • Scattered photons have less energy than the original gamma ray • Scattered photons will give less fluorescent light when they interact with the PMT crystal • The total sum of the electronic pulses from the PMTs will be less for scattered photons • It’s possible to reject scattered photons by rejecting pulses which do not have the expected known isotope energy • An appropriate energy window is set, only pulses which fall within that energy window are accepted Page 7 of 9 Question 4 a) In nuclear medicine imaging, explain in detail what is meant by the term “Aliasing”. Why does it arise? What effect does it have on the image acquired? What steps can we take to minimise its appearance in the image? [8 marks] b) If we acquire an image with a field of view of 400mm, and a matrix of 128 x 128, what is the minimum size object which can be resolved in the image? [4 marks] c) Explain what an image histogram is and how it may be used to enhance the contrast in an image with inherently low contrast. Illustrate your answer with an example histogram of an image with low contrast and how this would be altered after enhancing the contrast. [8 marks] Answer Plan a) Aliasing – due to using too low a sampling frequency. Relates to the spatial frequencies present in the object which we wish to image using the gamma camera. Should mention the Nyquist sampling theorem – state what this is, i.e. related to sampling frequency used to acquire the image – must sample at a frequency at least twice that of the highest frequency present in the objet to be imaged. Appears as spurious low frequency features in the image – high spatial frequencies will not be reproduced accurately in the image, i.e. results in a blurred image. Related to the spatial resolution – increasing the resolution will decrease its appearance in the image. b) 400mm FOV, 128 matrix à pixels are 3.125mm. Nyquist states that the sampling frequency must be at least twice the highest frequency present. Hence, the smallest object which can be resolved in the image will be 3.125mm x 2 = 6.25mm c) A histogram is a graph showing the distribution of pixel values in an image It plots the number of pixels with different grey levels in the image. Should show an example histogram of an image with low contrast [e.g. all pixels bunched towards the centre]. [4 marks] Enhance the contrast using the technique of windowing: apply a function like the following, should explain how the function changes the histogram: Page 8 of 9 [4 marks] Page 9 of 9