Seminar

SeminaronRecent Trends and Techniques in Medical Imaging : 

SeminaronRecent Trends and Techniques in Medical Imaging Presented By: Amit Kaul Lecturer EED

Slide 2: 

Why imaging? Diagnosis X-ray, MRI, Ultrasound, microscopic imaging (pathology and histology) … Visualization (invasive and noninvasive) 3-D, 4-D Functional analysis Functional MRI Phenotyping Microscopic imaging for different genotypes, molecular imaging Quantification Cell count, volume rendering, Ca2+ concentration …

Slide 3: 

Imaging modalities Wavelength Electron microscope X-ray UV Light Ultrasound MRI Fluorescence Multi-spectral Tomography Video

Development of Imaging Methods : 

Development of Imaging Methods The Imaging Methods ccan be divided into two categories Macroscopic – X- ray, Fast CT , Angiography, Tracer Methods (SPECT/PET) , MRI etc. Microscopic methods – Spectral imaging Scanning Electron microscope

MEDICAL IMAGING : 

MEDICAL IMAGING X-RAY IMAGE ULTRASOUND IMAGE

Positron Emission Tomography(PET) : 

Positron Emission Tomography(PET) BRAIN LUNG

3D ULTRASOUND IMAGE : 

3D ULTRASOUND IMAGE

Medical Images (Some Examples…) : 

Medical Images (Some Examples…) X-RAY IMAGE Positron Emission Tomography(PET) -Brain MRI IMAGE Sonic images

A Brief History of MRI : 

A Brief History of MRI In 1946, Felix Bloch and Edward Purcell discovered that when hydrogen atoms in a strong magnetic field were bombarded with radiowaves, the unpaired proton in the hydrogen nucleus produces a “nuclear magnetic resonance (NMR)” signal that could be measured. From 1950-1970, NMR was developed and used for molecular analysis. In 1971 Raymond Damadian showed that the nuclear magnetic resonance relaxation times of tissues and tumors differed, motivating scientists to consider magnetic resonance for the detection of disease. The advantage of NMR, or magnetic resonance imaging (MRI) as it has come to be known, is that unlike X-rays, MRI poses no health risks to the patient. Throughout the 1970s, researchers worked, on reducing the time it took to image the human body using NMR. By 1986, tissues of the human body could be imaged in only 5 seconds. As of 2003, there were approximately 10,000 MRI units worldwide, and approximately 75 million MRI scans per year performed.

Electron Microscope vs. Atomic Force Microscope : 

Electron Microscope vs. Atomic Force Microscope

Comparison Of SEM and AFM : 

Comparison Of SEM and AFM

Multimodal Imaging : 

Multimodal Imaging Multimodal Imaging is needed for three basic reasons To acquire complimentary which may be needed to reach a definitive diagnostic conclusion To provide added information and new images which are more informative than the individual source images To plan therapeutic procedures and monitor treatment

A System Approach : 

A System Approach Qualitatively vs. Quantitaively Mono Imaging vs. Fusion Imaging Static Imaging vs. Real Time Imaging with 3d & 4d Regional vs. Panoramic Macro vs. Spectral & Molecular

Slide 14: 

Areas of Image Processing and Analysis Image enhancement Color correction, noise removal, contrast enhancement … Feature extraction color, point, edge (line, curves), area cell, tissue type, organ, region Segmentation Registration 3-D reconstruction Visualization Quantization

Slide 15: 

Components of Imaging System Instrumentation : Electrical engineering, physics, histochemistry … Image generation Sensor technology (e.g., scanner), coloring agents … Image processing and enhancement Both software, hardware, or experimental (dynamic contrast) Image analysis at all levels Image processing, computer vision, machine learning Manual/interactive Image storage and retrieval Database/data warehouse

Image Processing in a Clinical Workflow : 

Image Processing in a Clinical Workflow

TYPICAL IMAGE PROCESSING SYSTEM FOR MEDICAL IMAGING : 

TYPICAL IMAGE PROCESSING SYSTEM FOR MEDICAL IMAGING Image Acquisition -Low Level- Pre-Processing (filtering) Enhancement (sharpen, smooth, …) Interpolate, reduce noise, crop Output -Intermediate Level- Segmentation Region (or Contour) Extraction, Labeling, Grouping -High Level- Recognition Region feature analysis (position, orientation, size…), Object Matching -High Level- Modeling Volume Rendering, Deformable Models, Mathematical Models

Distance Measurement : 

Distance Measurement Euclidean distance metric Mahanabolis distance metric Manhattan / city block distance metric Maximum value metric

: 

An edge is a set of connected pixels that lie on the boundary between two regions The pixels on an edge are called edge points Position & orientation of edge Gray level discontinuity across an edge WHAT IS AN EDGE?

TYPES OF EDGES : 

TYPES OF EDGES Gray level profile derivatives Step edge: Ramp edge: Peak edge: 2nd 1st 1st

DIFFERENT TYPES OF EDGES IN A CT IMAGE : 

DIFFERENT TYPES OF EDGES IN A CT IMAGE Peak edge Step edge Ramp edge

DEFINITION OF GRADIENT : 

DEFINITION OF GRADIENT A point is defined as an edge point if its 2-D first or second -order derivative is greater than a specified threshold. Gradient of digital image, , is defined by a vector:

GRADIENT CALCULATATION : 

GRADIENT CALCULATATION Gradient at an edge point, (x,y), can also be interpreted as a complex number with its magnitude determined by and the direction determined by

EDGE OPERATORS : 

EDGE OPERATORS The partial derivatives in x and y, , can be estimated using different ways: Roberts operator: Prewitt operator: Sobel operator:

GRADIENT MASKS : 

GRADIENT MASKS All operators can be performed by the convolution using different masks. Roberts operator masks:

GRADIENT OPERATORS : 

GRADIENT OPERATORS Prewitt operator masks: Sobel operator masks:

Slide 27: 

Common machine learning techniques Dimensionality reduction Principal component analysis (PCA, SVD, KLT) Linear discriminant analysis (LDA, Fisher’s discriminant)

Basic Learning Modes : 

Basic Learning Modes Supervised learning Training sets with labeled classes Classification Distance between the actual and the desired response output server as an error measure and is used to correct network parameters externally Unsupervised learning Use patterns that are typical redundant raw data without labels regarding their class memberships Network discovers for itself any possibly existing patterns Self-organization

Slide 29: 

Common machine learning techniques Supervised learning Learning algorithm Classifier ? Neural network, Support vector machine (SVM), MCMC, Bayesian network …

Slide 30: 

Common machine learning techniques Unsupervised learning K-means, K-subspaces, GPCA, hierarchical clustering, vector quantization, …

Thanks for your kind attention! : 

Thanks for your kind attention!