1 / 44

Medical Imaging Instrumentation & Image Analysis

Medical Imaging Instrumentation & Image Analysis. Atam P. Dhawan, Ph.D. Dept. of Electrical & Computer Engineering Dept. of Biomedical Engineering New Jersey Institute of Technology Newark, NJ, 07102 Dhawan@adm.njit.edu. Imaging in Medical Sciences.

koto
Download Presentation

Medical Imaging Instrumentation & Image Analysis

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Medical Imaging Instrumentation & Image Analysis Atam P. Dhawan, Ph.D. Dept. of Electrical & Computer Engineering Dept. of Biomedical Engineering New Jersey Institute of Technology Newark, NJ, 07102 Dhawan@adm.njit.edu

  2. Imaging in Medical Sciences • Imaging is an essential aspect of medical sciences for visualization of anatomical structures and functional or metabolic information of the human body. • Structural and functional imaging of human body is important for understanding the human body anatomy, physiological processes, function of organs, and behavior of whole or a part of organ under the influence of abnormal physiological conditions or a disease.

  3. Medical Imaging • Radiological sciences in the last two decades have witnessed a revolutionary progress in medical imaging and computerized medical image processing. • Advances in multi-dimensional medical imaging modalities • X-ray Mammography • X-ray Computed Tomography (CT) • Single Photon Computed Tomography (SPECT) • Positron Emission Tomography (PET) • Ultrasound • Magnetic Resonance Imaging (MRI) • functional Magnetic Resonance Imaging (fMRI) • Important radiological tools in diagnosis and treatment evaluation and intervention of critical diseases for significant improvement in health care.

  4. A Multidisciplinary Paradigm

  5. Electromagnetic Radiation Spectrum

  6. Medical Imaging Information • Anatomical • X-Ray Radiography • X-Ray CT • MRI • Ultrasound • Optical • Functional/Metabolic • SPECT • PET • fMRI, pMRI • Ultrasound • Optical Fluorescence • Electrical Impedance

  7. Medical Imaging Modalities Medical Imaging Methods Using Energy Source Internal External Combination: Internal & External X-Ray Ultrasound Optical MRI Fluorescence EI SPECT PET

  8. Medical Imaging Thru Transmission Basic Principle: Radiation is attenuated when passed through the body.

  9. ROC: Performance Measure Ntp = Notp + Nofn and Ntn = Nofp + Notn

  10. Fractions True Positive Fraction (TPF): ratio of the number of positive observations to the number of positive true-condition cases. TPF = Notp/Ntp False Negative Fraction (FNF): ratio of the number of negative observations to the number of positive true-condition cases. FNF = Nofn/Ntp \ False Positive Fraction (FPF): ratio of the number of positive observations to the number of negative true-condition cases. FPF = Nofp/Ntn True Negative Fraction (TNF): ratio of the number of negative observations to the number of negative true-condition cases. TNF = Notn/Ntn TPF + FNF = 1 and TNF + FPF = 1

  11. Measures Sensitivity is TPF Specificity is TNF Accuracy = (TPF+TNF)/Ntotal

  12. ROC Curve FPF=1-TNF TPF a b c

  13. A Case Study Total number of patients = Ntot=100 Total number of patients with biopsy proven cancer (true condition of object present) = Ntp=10 Total number of patients with biopsy proven normal tissue (true condition of object NOT present) = Ntn=90 Out of the patients with cancer Ntp , the number of patients diagnosed by the physician as having cancer = Number of True Positive cases = Notp=8 Out of the patients with cancer Ntp, the number of patients diagnosed by the physician as normal = Number of False Negative cases = Nofn=2 Out of the normal patients Ntn, the number of patients rated by the physician as normal = Number of True Negative cases = Notn=85 Out of the normal patients Ntn, the number of patients rated by the physician as having cancer = Number of False Positive cases = Nofp=5

  14. Example True Positive Fraction (TPF) = 8/10 = 0.8 False Negative Fraction (FNF) = 2/10 = 0.2 False Positive Fraction (FPF) = 5/90 = 0.0556 True Negative Fraction (TNF) = 85/90 = 0.9444

  15. Linear System • A system is said to be linear if it follows two properties: scaling and superposition.

  16. Image Formation: Object f, Image g Non-negativity Superposition Linear Response Function Image Formation Linear Image Formation

  17. y b Radiating Object f(a,b,g) Image g(x,y,z) Image Formation System h g z Image Domain Object Domain x a Image Formation

  18. Simple Case: LSI Image Formation g=h**f

  19. b y Radiating Object Image Image Formation System h Selected Cross-Section g z Image Domain Object Domain x a Radiation Source Reconstructed Cross-Sectional Image Image Formation: External Source

  20. b y Image Radiating Object Image Formation System h Selected Cross-Section g z Image Domain Object Domain x a Reconstructed Cross-Sectional Image Image Formation: Internal Source

  21. Fourier Transform

  22. Properties of FT • Linearity: Fourier transform, FT, is a linear transform. FT {ag(x,y)+bh(x,y)}= aFT{g(x,y)+ bFT{h(x,y) Scaling: It provides a proportional scaling.

  23. FT Properties Translation Convolution Cross-Correlation

  24. FT Properties…. Auto-Correlation Parseval’s Theorem Separability

  25. y q p f(x,y) q x p P(p,q) q Line integral projection P(p,q) of the two-dimensional Radon transform. Radon Transform

  26. Projection p1 A Reconstruction Space B Projection p3 Projection p2 Radon Transform

  27. Backprojection Reconstruction Method where L is the total number of projections acquired during the imaging process at viewing angles

  28. Sampling Theorem • The sampling theorem provides the mathematical foundation of Nyquist criterion to determine the optimal sampling rate for discretization of an analog signal without the loss of any frequency information. • The Nyquist criterion states that to avoid any loss of information or aliasing artifact, an analog signal must be sampled with a sampling frequency that is at least twice the maximum frequency present in the original signal.

  29. Sampling The sampled version of the image, fd[x,y] is obtained from sampling the analog version as

  30. Sampling Effect • In Fourier domain the spectrum overlapping has to be avoided by proper sampling of the image in spatial domain. • Sampling in spatial domain produces a convolution in the frequency domain.

  31. (a) wy wymax Fa(wx, wy) wx wxmax -wxmax (c) -wymax (b) Nyquist (Optimal) Sampling

  32. Wavelet Transform • Fourier Transform only provides frequency information. • Windowed Fourier Transform can provide time-frequency localization limited by the window size. • Wavelet Transform is a method for complete time-frequency localization for signal analysis and characterization.

  33. Wavelet Transform.. • Wavelet Transform : works like a microscope focusing on finer time resolution as the scale becomes small to see how the impulse gets better localized at higher frequency permitting a local characterization • Provides Orthonormal bases while STFT does not. • Provides a multi-resolution signal analysis approach.

  34. Wavelet Transform… • Using scales and shifts of a prototype wavelet, a linear expansion of a signal is obtained. • Lower frequencies, where the bandwidth is narrow (corresponding to a longer basis function) are sampled with a large time step. • Higher frequencies corresponding to a short basis function are sampled with a smaller time step.

  35. Continuous Wavelet Transform • Shifting and scaling of a prototype wavelet function can provide both time and frequency localization. • Let us define a real bandpass filter with impulse response y(t) and zero mean: • This function now has changing time-frequency tiles because of scaling. • a<1: y(a,b) will be short and of high frequency • a>1: y(a,b) will be long and of low frequency

  36. Wavelet Decomposition

  37. Wavelet Coefficients • Using orthonormal property of the basis functions, wavelet coefficients of a signal f(x) can be computed as • The signal can be reconstructed from the coefficients as

  38. Wavelet Transform with Filters • The mother wavelet can be constructed using a scaling function f(x) which satisfies the two-scale equation • Coefficients h(k) have to meet several conditions for the set of basis functions to be unique, orthonormal and have a certain degree of regularity. • For filtering operations, h(k) and g(k) coefficients can be used as the impulse responses correspond to the low and high pass operations.

  39. H H 2 H 2 Data G 2 G 2 G 2 Decomposition

  40. Wavelet Decomposition Space

  41. s u b - s a m p l e h - h h s u b - s a m p l e h - g h g X h g - h g g - g g h o r i z o n t a l l y v e r t i c a l l y L e v e l 0 L e v e l 1 Image Decomposition Image

  42. Wavelet and Scaling Functions

  43. Image Processing and Enhancement

More Related