Computer pattern recognition of

Slides:



Advertisements
Similar presentations
AIME03, Oct 21, 2003 Classification of Ovarian Tumors Using Bayesian Least Squares Support Vector Machines C. Lu 1, T. Van Gestel 1, J. A. K. Suykens.
Advertisements

Course Syllabus 1.Color 2.Camera models, camera calibration 3.Advanced image pre-processing Line detection Corner detection Maximally stable extremal regions.
Chapter 9: Morphological Image Processing
Course Syllabus 1.Color 2.Camera models, camera calibration 3.Advanced image pre-processing Line detection Corner detection Maximally stable extremal regions.
DIGITAL IMAGE PROCESSING
Chapter 1: Introduction to Pattern Recognition
ISDH Lab TB Testing Update Lixia Liu PhD MP(ASCP) September 16, 2010.
Automatic Face Recognition Using Color Based Segmentation and Intelligent Energy Detection Michael Padilla and Zihong Fan Group 16 EE368, Spring
Darlene Goldstein 29 January 2003 Receiver Operating Characteristic Methodology.
Feature Screening Concept: A greedy feature selection method. Rank features and discard those whose ranking criterions are below the threshold. Problem:
Copyright © 2012 Elsevier Inc. All rights reserved.. Chapter 9 Binary Shape Analysis.
Object Detection Procedure CAMERA SOFTWARE LABVIEW IMAGE PROCESSING ALGORITHMS MOTOR CONTROLLERS TCP/IP
E.G.M. PetrakisBinary Image Processing1 Binary Image Analysis Segmentation produces homogenous regions –each region has uniform gray-level –each region.
Pattern Classification All materials in these slides were taken from Pattern Classification (2nd ed) by R. O. Duda, P. E. Hart and D. G. Stork, John Wiley.
GUIDED BY: C.VENKATESH PRESENTED BY: S.FAHIMUDDIN C.VAMSI KRISHNA ASST.PROFESSOR M.V.KRISHNA REDDY (DEPT.ECE)
Attention Deficit Hyperactivity Disorder (ADHD) Student Classification Using Genetic Algorithm and Artificial Neural Network S. Yenaeng 1, S. Saelee 2.
An Interactive Segmentation Approach Using Color Pre- processing Marisol Martinez Escobar Ph.D Candidate Major Professor: Eliot Winer Department of Mechanical.
July 11, 2001Daniel Whiteson Support Vector Machines: Get more Higgs out of your data Daniel Whiteson UC Berkeley.
Biomechanics and biology: bridging the gap Sam Evans School of Engineering
Participation in the NIPS 2003 Challenge Theodor Mader ETH Zurich, Five Datasets were provided for experiments: ARCENE: cancer diagnosis.
Mathematical Morphology Set-theoretic representation for binary shapes
An efficient method of license plate location Pattern Recognition Letters 26 (2005) Journal of Electronic Imaging 11(4), (October 2002)
Implementation of Thin Layer Agar for Mycobacterium culture in rural Kenya Médecins Sans Frontières.
Course Syllabus 1.Color 2.Camera models, camera calibration 3.Advanced image pre-processing Line detection Corner detection Maximally stable extremal regions.
Detection of Labeling Markers on Synthetic DNA molecules Background Deoxyribonucleic acid (DNA) is the “code of life” that provides the recipe of genetic.
CSE554SkeletonsSlide 1 CSE 554 Lecture 2: Shape Analysis (Part I) Fall 2015.
Orientable Textures for Image- Based Pen-And-Ink Illustration Michael P. Salisbury Michael T. Wong John F. Hughes David A. Salesin SIGGRAPH 1997 Andrea.
Low level Computer Vision 1. Thresholding 2. Convolution 3. Morphological Operations 4. Connected Component Extraction 5. Feature Extraction 1.
References Books: Chapter 11, Image Processing, Analysis, and Machine Vision, Sonka et al Chapter 9, Digital Image Processing, Gonzalez & Woods.
Morphological Image Processing Robotics. 2/22/2016Introduction to Machine Vision Remember from Lecture 12: GRAY LEVEL THRESHOLDING Objects Set threshold.
ECE472/572 - Lecture 14 Morphological Image Processing 11/17/11.
Mathematical Morphology A Geometric Approach to Image Processing and Analysis John Goutsias Department of Electrical and Computer Engineering Image Analysis.
Che-An Wu Background substitution. Background Substitution AlphaMa p Trimap Depth Map Extract the foreground object and put into another background Objective.
INSTITUTO DE INFECTOLOGIA EMÍLIO RIBAS Identification of Mycobacterium tuberculosis complex in clinical specimens of HIV-infected patients at Instituto.
A NEW RAPID RESAZURIN-BASED MICRODILUTION ASSAY FOR ANTIMICROBIAL SUSCEPTIBILITY TESTING OF NEISSERIA GONORRHOEAE Sunniva Förster 1,3,4, Valentino Desilvestro.
1 Better Off Dead: Suicidal Thoughts in Cancer Patients Jane Walker, Rachel A. Waters, Gordon Murray, Helen Swanson, Carina J. Hibberd, Robert W. Rush,Dawn.
Basic Image Analysis GUI for Cell Counting and Segmentation Tri Phan, Ravi Shrivastav, Rubina Narang.
PROS AND CONS OF LYME DISEASE TESTS:
Identification of Leaves by Interior Shape and Texture
7. Performance Measurement
ACCELERATING PROGRESS TOWARDS THE FIRST 90 AMONG MEN: A TRIAL OF THE PEER-BASED DISTRIBUTION OF HIV SELF-TEST KITS IN BULISA, UGANDA M. Nanfuka 1, A. Choko.
CSE 554 Lecture 2: Shape Analysis (Part I)
Comprised of Blood Agar and CHROMagarTM Orientation using WASPLabTM
Facial Imaging in FASD and Related Disabilities
Digital Image Processing CP-7008 Lecture # 09 Morphological Image Processing Fall 2011.
Field Testing of OMNI-gene TB Sputum Optimizer in Malawi
In Search of the Optimal Set of Indicators when Classifying Histopathological Images Catalin Stoean University of Craiova, Romania
Results for all features Results for the reduced set of features
Detection of Labeling Markers on Synthetic DNA molecules
CJT 765: Structural Equation Modeling
39 DEVELOPED HCC by EASL criteria
Mycobacterium. Mycobacterium Important Human Pathogens Mycobacterium tuberculosis Mycobacterium leprae (uncommon) Mycobacterium avium-intracellulaire.
Brain Hemorrhage Detection and Classification Steps
Students: Meiling He Advisor: Prof. Brain Armstrong
A New Approach to Track Multiple Vehicles With the Combination of Robust Detection and Two Classifiers Weidong Min , Mengdan Fan, Xiaoguang Guo, and Qing.
Dingding Liu* Yingen Xiong† Linda Shapiro* Kari Pulli†
The Basics of Microarray Image Processing
An Improved Neural Network Algorithm for Classifying the Transmission Line Faults Slavko Vasilic Dr Mladen Kezunovic Texas A&M University.
Tools of the Laboratory Power Point #1: Culturing Microorganisms
Department of Electrical Engineering
Models of education in medicine, public health, and engineering
2 types of scale factor problems
Midterm Exam Closed book, notes, computer Similar to test 1 in format:
Deciphering TB Lab Reports
ECE 692 – Advanced Topics in Computer Vision
Examination of specimens for mycobacteria in clinical laboratories in 21 countries: a 10- year review of the UK National Quality Assessment Scheme for.
Treatment for PTSD and SUD:
Aim: How can we do color adjustment in ImageJ?
Synchronization of a basic diagnostic algorithm (linked diagnostic tests) with different levels of diagnostic services. Synchronization of a basic diagnostic.
Basic diagnostic algorithm to link the molecular line probe assay with solid culture- and liquid culture-based growth detection and susceptibility testing.
Presentation transcript:

Computer pattern recognition of Mycobacterium tuberculosis in MODS culture Alicia Katherine Alva Mantari1, Mirko Zimic1, David A. J. Moore2, Robert H. Gilman3, Mark F. Brady4 1 Universidad Peruana Cayetano Heredia, Lima, Peru; 2 Wellcome Centre for Clinical Tropical Medicine, Imperial College London, London, United Kingdom; 3 Department of International Health, Bloomberg School of Public Health, Johns Hopkins University, Baltimore, USA; 4 Warren Alpert Medical School of Brown University, Providence, USA Background Results Conclusions Image Processing The Microscopic-Observation Drug-Susceptibility assay (MODS) is a non-proprietary, low-technology, low-cost liquid broth TB culture tool MODS is a direct observation culture method that simultaneously yields drug susceptibility and is an improvement on current diagnostics in terms of accuracy, speed, and cost1 Scaling-up this test internationally is difficult because well-trained laboratory technicians are necessary to read the test results Access to MODS might be significantly improved if a computer could be trained to accurately identify Mycobacterium tuberculosis (MTB) in MODS cultures using only free computer programs2, 3 This first attempt at a pattern recognition algorithm to identify MTB in MODS culture yielded both a high sensitivity and specificity The algorithm was able to differentiate well between MTB and atypical mycobacteria, even without additional adjusting The performance of this algorithm lends quantitative support to laboratory technician claims that MTB in MODS has a unique growth pattern Border Original Grayscale Threshold Filter Median Exclude Border Dilation Background & Border Erosion Filter Fill holes Filter 2 Median Filter 2 Fill holes Filter Area Color the background Skeleton Extraction of image variables Future directions While 10 days incubation was chosen because characteristic cords have formed by this time but conglomerations have not yet formed, it is possible that the evolution of image variables over time could allow for higher precision or faster identification of MTB This algorithm may benefit from adjustment with machine learning optimization A clinically applicable experiment will use the entire image as the measure of success rather than the image, leading to another degree of complexity Larger image libraries will allow for further evolution of this algorithm 7 image variables were selected out of >40 according to significance in a logistic regression based on the learning set of hand-picked typical MTB forms: Length, width, width variability, birefringence, the ratio of the length to the perimeter, and another composite variable Methods MODS cultures at 10 days of incubation were digitally photographed and put through a series of image processing procedures to extract image variables Logistic regression was used on a set of typical MTB cording forms to make an algorithm to identify specific morphological characteristics and geometrical relationships typical to MTB This algorithm was then challenged to identify a series of 2875 MTB and non-MTB forms Because of the concern that atypical mycobacteria might resemble MTB, other mycobacteria were included to ensure that the algorithm could differentiate between mycobacteria: M. kansasiis, M. avium, and M. chelonae Algorithm cross-reactivity Typical cords vs. detritus Bacteria Specificity M. avium 97.06 M. chelonae 99.14 M. kansasii 93.75 Detritus 98.25 Irregularly shaped MTB 97.89 ROC References Sensitivity= 98.91 Specificity= 98.38 Moore DA, Evans CA, Gilman RH, et al. Microscopic-observation drug-susceptibility assay for the diagnosis of TB. N Engl J Med. 2006 Oct 12;355(15):1539-50. Scilab Toolbox available at http://www.scilab.org/ ImageJ available at http://rsbweb.nih.gov/ij/ Youden index = 0.9594 Visit modsperu.org for more information