15 Pages

IOP_Proposal

Course: BME 227, Fall 2009
School: Duke
Rating:
 
 
 
 
 

Word Count: 2738

Document Preview

Intraocular Home Pressure Measurement Device Team: Priya Karani, Shawn Tan, Lin Xiong, & Cen Zhang Clients: Richard Awdeh, MD and Felix Chau, MD Duke University Eye Center BME 227L Professor Aura Gimm February 10, 2006 Abstract Glaucoma is the second leading cause of blindness and affects approximately 66 million people worldwide. Although advancements in treatment have been made for patients with...

Register Now

Unformatted Document Excerpt

Coursehero >> North Carolina >> Duke >> BME 227

Course Hero has millions of student submitted documents similar to the one
below including study guides, practice problems, reference materials, practice exams, textbook help and tutor support.

Course Hero has millions of student submitted documents similar to the one below including study guides, practice problems, reference materials, practice exams, textbook help and tutor support.
Intraocular Home Pressure Measurement Device Team: Priya Karani, Shawn Tan, Lin Xiong, & Cen Zhang Clients: Richard Awdeh, MD and Felix Chau, MD Duke University Eye Center BME 227L Professor Aura Gimm February 10, 2006 Abstract Glaucoma is the second leading cause of blindness and affects approximately 66 million people worldwide. Although advancements in treatment have been made for patients with glaucoma, developments to accurately diagnose those with glaucoma lag behind. As a result, people become aware they have glaucoma only after they exhibit symptoms of irreversible damage. The disease is not yet fully understood, however elevations in intraocular pressure (IOP) have been identified as a major risk factor. To check their IOP, patients must go to the clinic to see their ophthalomologist. Recent studies show that IOP varies widely throughout the day and thus two to four measurements obtained periodically during the year are not indicative of a patients actual IOP. Thus, there is a dire need for a cost-effective home intraocular measurement device for early detection and management of glaucoma. The following project was an undertaking to design such a device for patients at risk for glaucoma. Research into the literature and previous designs led to electromechanical, chemical, and optical solutions. The reverse applanation tonometer (RAT) was selected as the most promising design after evaluation using a decision matrix. The design comprises of relatively simple and affordable components that could feasibly fit into a portable home device. However, there are still many modifications and improvements that need to be made before a functional prototype can be assembled from the design. Background: Glaucoma is a progressive, asymptomatic disease of the optic nerve that is often associated with optic disc cupping and irreversible visual field loss1. In the normal eye, aqueous humor is produced in the ciliary body and drained from the eye in the trabecular meshwork at a balanced rate (Figure 1). The rates at which aqueous humor is produced and drained determine the eyes intraocular pressure (IOP), which ranges from 12-21 mm Hg. Failure of aqueous humor to leave the eye due to obstruction at the drainage sites results in elevated IOP. Elevated IOP is not a diagnostic factor and many people with high pressures have normal vision. However, IOP is the only established risk factor for glaucoma. Statistics show that for every 1 mmHg increase in IOP, the risk for getting glaucoma increases by 10%. Figure 1. Production (ciliary body) and drainage (trabecular meshwork) sites of aqueous humor in the eye.2 Currently, the two most widely employed devices to measure IOP are the Goldmann tonometer and the Tono-Pen, which is a microprocessor-controlled, portable version of the Goldmann tonometer. Both devices utilize the applanation principle, which is based on the relationship between force, pressure, and area (figure 2).3 A known force 3 is applied to the cornea when it is in the convex state until it is flattened, or applanated. At this point, the IOP is equal to the force applied divided by the applanated area. Figure 2. Principle of applanation. IOP is measured by dividing the applied force by the applanated area when the cornea is completely flattened.4 Although the Goldmann method is considered the standard procedure for measuring IOP, it has a number of disadvantages. For the patient, the procedure is highly invasive and uncomfortable. First, the head of the tonometer must come into direct contact with the cornea (figure 3). Second, to prevent reflexes during measurement, anesthesia must be applied to numb the eye. A fluorescein strip must also be applied to the eye to detect the applanated state. All of these steps require a skilled technician or expertise of the ophthalmologist. Figure 3. Two common clinical tonometers: Goldmann5 (left) and the Tonopen6 (right). Another disadvantage of these devices is that despite the advanced techniques involved, they do not provide an accurate measure of the patients IOP. Since doctors only periodically check IOP, which has been known to fluctuate widely throughout the 4 day, the measurement obtained may significantly under- or overestimate the actual pressure in the patients eye. As a result, patients at risk for glaucoma may not receive the proper treatment or those with normal vision are treated unnecessarily. Thus, in order to gain an accurate representation of the patients IOP, patients must be able to monitor their IOP by themselves throughout the day. These facts point to a strong need for a simple, portable device that patients can use to measure their IOP on their own. Goal Statement and Functional Specifications: The teams overall goal is to build a device to allow patients with glaucoma to measure their intraocular pressure at home. The design requirements are as follows: (1) The person should be able to use the device by himself/herself without the assistance of another person. (2) The device should be user-friendly and not require complex instructions and additional training. (3) The device should be accurate (resolution of 5 mm Hg) so as to consistently give users measurements indicative of their actual IOP, and operate within the range of 0-100 mm Hg. While the normal IOP ranges from 10-21 mm Hg, it may be as high as 100 mm Hg for patients with late stage glaucoma. Hence, the design of the device should account for the entire range of pressures. (4) The device should not require anesthetic drops since these drops, which are highly toxic in excess, are not permitted by the Food and Drug Administration for home use. (5) The device should minimize patient discomfort. (6) The device is intended for patients to use at home and should be affordable for an average household. The cost should be approximately $100 per unit. 5 Proposed Designs: Design I Piezoelectric Device The first proposed design uses piezoelectric material, which has the ability to acquire a voltage in response to a mechanical stress or deform when subjected to an externally applied voltage7. Proposed Use of Device As figure 4 shows, the tonometer is applied over patients closed eyelid. The sensor contains a piezoelectric transducer that vibrates when an altering voltage is applied to it. The vibrational energy will be transmitted to the eyeball through the eyelid. Since the eyeball is stiff due to the pressure inside it, the vibrational amplitude of the device will change depending on the ocular pressure. This change in vibrational amplitude will result in a change in the current flowing through the transducer. Measurement of the current will provide an indirect measure of the measure IOP.8 Piezoelectric transducer Resistor Eyelid 6 Figure 4. Schematic diagram for piezoelectric design. Design II Pressure-sensitive Paints The second design was based on aerodynamic studies that use oxygen-sensitive, luminescent paints to determine the pressure at the surface of aircraft. A thin layer of paint (~5 um) is coated over the surface and a sample is exposed to light to excite the luminescent molecules. Figure 5. Oxygen quenching of luminescent molecules9 A photodetector is used to measure the intensity of the luminescence after the dyes have been quenched by oxygen. The intensity of the remaining molecules is related to the partial pressure of oxygen in contact with the dye through the Stern-Volmer equation10: Io/I = 1+kP (1) where P is the partial pressure of oxygen, Io is the maximum intensity in the absence of oxygen, I is the intensity of the molecules after oxygen quenching, and k is a constant. Proposed Use of Device Although aerodynamic pressures far exceed the range of IOP, the concept behind the method may be applied to determine IOP. Previous studies have used luminescence and non-toxic molecules in biological settings to relate the concentration of oxygen to the lifetime of the molecules11,12. Thus, the idea is to introduce luminescent molecules into 7 the eye, measure the consumption of these molecules by oxygen, and calculate the IOP from the intensity of the remaining dye. One way to administer the dye is to embed a strip of pressure-sensitive paint in soft contact lenses. The patient can then easily apply the contact to his/her eye and remove it for subsequent analysis by photodetection. Soft contact lens O2 O2 O2 Pressure-sensitive paint Oxygen diffusion Figure 6. Diagram of contact lenses loaded with pressure-sensitive paint 8 Design 3 Reverse Applanation Tonometer The Reverse Applanation Tonometer (RAT) as shown in figure 7 is a modification of the Goldmann Applanation Tonometer (GAT) that allows the user to determine the change in concavity of the cornea without assistance. Figure 7. Schematic representation of the reverse applanation Use tonometer Proposed of Device The user will place the handheld device on his/her eye and look through the viewing tube [1] onto the viewing surface at the end [2]. The user will be able to adjust the focal power of the myopic adjustment unit [3] using Knob 1 until the letters on the viewing surface are clear. The retractable o ring [4] should be used to align the cornea with the fluorescent red dot on the center of the view surface before initiating the IOP measurement. The user will then activate the control unit [5] to cause the solenoid plunger [6] to exert a known constant force via an air-stream onto the surface of the cornea. As the concavity of the cornea changes, the users vision of the viewing surface 9 becomes blurred. The user will have to adjust Knob 2, such that the focusing unit [7] changes to compensate for the loss of focal power in the eye, until the viewing surface is clear again. The loss of focal power in the eye is recorded via the amount that Knob 2 was turned. The intro-ocular pressure can be determined by relating the known force applied over a fixed area (diameter of 3.06 mm) to the change in curvature of the cornea. Working principles The device is based on the relationship between focal length, lens curvature and focal power of a lens as described in the following equations: P= 1 1 = ( n 1) f r1 1 r2 (2) where P is the power (diopters), f is the focal length (m), n is the refractive index, r1 and r2 are the radii of curvature for each side of the lens.13 Assuming that the cornea can be modeled as a single side of the lens (r1) and the curvature of the lens in the eye (r2) remains constant, the derivative of P with respect to r1 is: dP = ( n 1) dr1 1 r12 (3) The change in power can thus be related to the change in curvature: P = ( n 1) 1 r12 r1 (4) The exact relationship between the change in curvature of the cornea and the intra-ocular pressure has yet to be determined. 10 Design evaluations A decision matrix was constructed for the top three proposed designs (table 1). Each design was rated according to accuracy, ease of use, cost, and durability. Safety was considered to be an implicit requirement and was not included in the matrix. Table 1. Decision matrix for top three designs Criteria Accuracy Ease of use Cost Durability Total Weight Factor (1-100) 35 30 20 15 100 Piezo 4/140 3/90 1/20 3/45 295 Paint 3/105 2/60 1/20 2/30 215 RAT 3/105 3/90 5/100 4/60 355 The total scores in the design matrix suggest that the Reverse Applanation Tonometer (RAT) is the most promising design. It is essentially made up of relatively simple and affordable components that could easily fit into a handheld device. Nevertheless, there are many drawbacks of the design that require extensive modifications and improvements. For the device to function properly, there are two critical criteria that have to be fulfilled: the relationship between corneal curvature, force applied and intraocular pressure must be clearly defined; also, the user must have normal vision even under applied force to the cornea. In addition, the accuracy of the RAT relies on the vision and discretion of the user. This poses to be a potential problem since a large number of glaucoma suspects are 11 elderly with poor vision, or have a history of cataracts. Thus, the device needs to be improved to cater to as large a group of glaucoma suspects as possible. The piezoelectric tonometer has the next highest score, mainly due to its high accuracy and ease of use (can be used with eyelids closed). However, the exact relation between the vibrational resonance of the eye and the intraocular pressure has yet to be determined and is thus not as feasible either. The pressure-sensitive paint (PSP) design makes use of a novel concept with relatively high resolution (~1.125mmHg). The device scored lowest mainly due to its potentially high cost and difficulty of use, requiring additional devices to determine oxygen quenching. Although the design, which incorporated the PSP into a contact lens system, did not fare well in the ranking system, it might be feasible to incorporate the PSP in an alternative system. Conclusion and Future Direction: Detailed evaluations of the proposed designs suggest that the reverse applanation tonometer is the most promising design. However, to make it a functional prototype, several components will have to be modified. The team is currently looking into replacing the optics component of the design with an optical emitter-receiver component, which would eliminate the reliance on the vision of the user and also increase the accuracy of the device. In addition, the team is continuing to explore potential uses of the pressure-sensitive paint in determining either applanation or IOP. 12 Appendix: Project Design Specification for Intraocular Pressure Measurement Device Team: Priya Karani, Shawn Tan, Lin Xiong, and Cen Zhang Function: The purpose of the intraocular pressure (IOP) measurement device is to allow people with glaucoma to self-monitor IOP at home. Client Requirements: A person must be able to measure the pressure without another persons aid. The device should not be harmful to the eye. The device should not require anesthetic drops to numb the eye. The device should be able to measure pr...

Find millions of documents on Course Hero - Study Guides, Lecture Notes, Reference Materials, Practice Exams and more. Course Hero has millions of course specific materials providing students with the best way to expand their education.

Below is a small sample set of documents:

Duke - BME - 227
Design of an Aseptic Thymus Slicing DeviceCarolyn Eagan Richard Lee Marcus Reeslund Rachel WerginFebruary 9, 2006 BME 227: Design in BiotechnologyDesign of an Aseptic Thymus Slicing DeviceCarolyn Eagan, Richard Lee, Marcus Reeslund and Rachel
Duke - BME - 227
SupraStrain 3000 Turbo Plus: Mr. Artery Gold Edition - Device for Dynamic Pressurization of Arteries and Strain MeasurementJay Crowley, Jim Kragel, Phil Kragel, Ethan Puchaty, and Dr. Friedman Department of Biomedical Engineering, Duke University BM
Duke - BME - 227
SwelGelMeasuring Volume Transitions of HydrogelsTodd Elkin, Vy Hoang, Jon Weiss, Kirk Donahoe, and Jesse Longoria Department of Biomedical Engineering, Duke University, Durham, NC 27708 INTRODUCTIONHIV and other sexually transmitted diseases (STDs
Duke - ISIS - 250
ISIS 250. Critical Studies in New Media. Addresses key issues in the philosophy of new media. Central themes include the materiality of media; media configurations and the co-evolution of human being; computational media and recent discussions of pos
Kentucky - PA - 504
PHY 504Problem Set #4due 26 September 20081. Goldstein Problem 2.12. Can you generalize to a Lagrangian depending on the third time derivative of x? 2. Starting with velocity v0 at the top, a particle of mass m slides down a frictionless paraboli
Kentucky - PHY - 160
Force, Motion, and SoundPhysics 160, Spring 2006Galileo (1564-1642)Isaac Newton (1643-1727)Uniform Motionx= O t=1cm 1s2cm 2s3cm 3s4cm 4s5cm 5s6cm 6s7cm 7s8cm 8sThis picture shows the location of an object at successive time
Kentucky - PA - 600
PHY 600Problem Set #2due 20 February 20091. The motion of a free particle in a curved space with metric gij(x) is governed by the Lagrangian 2 (a) Obtain the equations of motion. (Fill in the details of the derivation given in class.) (b) Special
Kentucky - PHY - 151
PHY 151 - FALL 2008 TEST NO. 3 : NOVEMBER 19, 2008ANSWERS ARE IN BOLD AND/OR LARGE FONTSThe acceleration due to gravity is 9.8 m/s2 . You can take = 3.14 Density of water can be assumed to be 1000 Kg/m3 Question 1 A particle is moving along a cir
Kentucky - MA - 113
Calculus I Exam 2Common exam 7 March 2006Answer all of the following questions. Additional sheets are available if necessary. No books or notes may be used. You may use a calculator. You may not use a calculator which has symbolic manipulation ca
Kentucky - MA - 113
Kentucky - MA - 113
Kentucky - MATH - 113
Duke - CPS - 001
What is a Social Network The Other Kind of Networking: Social Networks on the WebDr. Jennifer Golbeck University of Maryland, College Park March 20, 20061 2/36 People and their connections to other peopleWeb-Based Social Networks (WBSNs) Socia
Duke - CPS - 001
Operating systemsMost important program that runs on a computer Every general-purpose (such as desktop) computer must have OS to run other programs Manages the hardware and software resources of the system Provides stable and consistent way for appl
Duke - CPS - 102
CPS102 DISCRETE MATHEMATICS Practice Final ExamIn contrast to the homework, no collaborations are allowed. You can use all your notes, calculator, and any books you think are useful. Write legibly and formulate each answer concisely, using only the
Duke - MATH - 290
The size of LONG is: 4 The size of UNSIGNED LONG is: 4 The size of DOUBLE is: 8 The default LONG is: 1073851552 The default UNSIGNED LONG is: 1073851552 The default DOUBLE is: 7.12893e-270 The smallest LONG is: -2147483648 The largest LONG
Duke - C - 724
G430: Chat 8 September 19, 2004 Welcome aboard! I may be a couple of minutes late this AM I am rushing here from a conference presentation that is supposed to end just before 11. See you soon!1 Questions?G430: Chat 8 September 19, 2004 J
Duke - CPS - 221
Reactive NUMA: A Design for Unifying SCOMA and CCNUMABabak Falsafi and David A. Wood University of Wisconsin, Madison, 1997 Presented by: Jie Xiao Feb 6, 2008 Outline: Introduction CCNUMA, SCOMA, RNUMA Theoretical Results Simulat
Kentucky - ECO - 707
Duke - PHY - 115
Four-Day OutlineBrent PerdueMy objective for these four sessions is to introduce the student to nuclear radiation detection. Over the span of the four classes we will be discussing common skills and procedures utilized frequently by an experimenter
Duke - PHY - 305
PHY305: Lecture Acknowledgement Brad Filippone Introduction UCNA production The UCNA Experiment at LANLPrecision neutron decay measurements Neutron lifetime essential in Big-Bang Nucleosynthesis Calculations < 0.3% measurements can be sensiti
Kentucky - EE - 572
EE572 HW #7 Due Wednesday, February 11^^1. a)If x k+1 = A x k + B w k , find the eigenvalues and eigenvectors then use the similarity transformation, xk = Pzk, to determine which of the eigenvalues are controllable (you may use Matlab if you
Duke - STAT - 214
Statistics 214: Probability and Statistical Models1Course Information- Prof: Scott Schmidler - Email: (put Sta 214 in subject line) schmidler@stat.duke.edu - Oce: 223D Old Chem Building - Phone: 684-8064 - Oce hours: Mon 4-5pm or by appt. - Cour
Duke - STA - 101
Statistics 101:Answers to Practice Problems for Final Exam1. False. The mean is roughly half way between the median and 25 percentile, so that about 37.5 percent of people have taxed greater than the mean. 2. 1000 3. True. 4. False: 5. Around 16%
Duke - STA - 101
ANSWERS TO PRACTICE PROBLEMS FOR STAT 101 MIDTERM 1. This is longer than the actual midterm, but it gives you an idea of the types of questions that will be asked. Topics not covered in these problems still may be covered on exams.Name (Please print
Duke - STA - 244
1561557603515445.3214502557.0260162657204693549.6415681553.7860173457906194546.42631-3354.14100184457903255552.7554-2864.76250195657003734148.0220832352.521201121657203445154.32
Duke - PHY - 042
Physics 42 Lecture 16Announcements Assignment 5 due on Monday noon D.C. circuit analysis Properties of electric circuit elements Kirchhoffs rules for circuit analysis Circuits with emfs and resistors Circuits with emfs resistors and capacitorsPhys
Kentucky - FCS - 4411
FCS4-411Healthy Homes: Indoor Air QualityManaging Asthma Triggers in the HomeLinda Adler, Family and Consumer SciencesAsthma, a serious lung disease, is a leading cause of longterm illness in children. It affects 232,870 adults and 50,883 ch
Duke - CPS - 130
10/06/2002CPS130, Lecture 11-12; Another Recursion Equation; Order Statistics; Lower Bound Sorting; Radix Sorting1. Introduction We have seen that the max or min of a set on n elements can be computed with n-1 comparisons, and that, in L5-6 (p 9)
Duke - LECTURE - 130
10/06/2002CPS130, Lecture 11-12; Another Recursion Equation; Order Statistics; Lower Bound Sorting; Radix Sorting1. Introduction We have seen that the max or min of a set on n elements can be computed with n-1 comparisons, and that, in L5-6 (p 9)
Duke - PHY - 217
5. Paper The first draft should be submitted within one week after the experiment is completed. The paper will be returned with comments for the final paper, which should be submitted within two weeks after it is returned. Although an experiment can
Duke - CM - 171
OIKOS 109: 555 /566, 2005Mid-domain models as predictors of species diversity patterns: bathymetric diversity gradients in the deep seaCraig R. McClain and Ron J. EtterMcClain, C. R. and Etter, R. J. 2005. Mid-domain models as predictors of spec
Duke - CPS - 212
Continuous Consistency and AvailabilityHaifeng YuCPS 212 Fall 2002Consistency in Replicationq qReplication comes with consistency cost: Reasons for replication: Better performance and availability Replication transforms client-server communica
Duke - CHEM - 8304
Restrictions on Chemical TechnologiesChemical Restrictions Health/Environmental RestrictionsCurrent Technologies to be Discussed in This Course Food Pharmaceuticals Materials - Clothing and Shelter Energy and Transportation Radiation Technol
Kentucky - EE - 522
Effective apertureIn receiving mode, the maximum power received in a receiving antenna is PAm . Consider this power to be that intercepted from the incoming wave by a maximum effective area Aem .AemIf the power density of the incoming wave is S,
Duke - STA - 103
Statistical testing vs. interval estimationChp. 8 focused on estimating a parameter (, p, etc.) and placing some bounds on the error of estimation. Can describe the population more accurately - knowing the mean income of a population gives us an id
Duke - STA - 215
Statistical InferenceRobert L. Wolpert Institute of Statistics and Decision Sciences Duke University, Durham, NC, USAWeek 11. Testing Statistical Hypotheses1.Statistical HypothosesUp to now we've been looking exclusively at the problem of es
Duke - STA - 215
Duke - STA - 110
Two Sample Problems*two.tex *It is often of interest to test a hypothesis that the means of two populations are the same. Imagine two populations P1 and P2 of items we are interested in buying. Assume that we can observe some quality feature lengt
Duke - STA - 293
Classication Problem training data {(xi, yi), i = 1, . . . , n} i.i.d P (x, y), where yi = 1. classier a multi-variate function g : Rd {1} 0-1 loss L(g(x), y) = I{g(x)=y} Risk R[g] = EX,Y L(g(X), Y ) = P (g(X) = Y ). The best classier g = argmi
Duke - STA - 345
Stat 345Homework #2 Due Friday, March 10Spring 2006EX1. Golub et al (1999) were interested in identifying genes that are dierentially expressed in patients with two type leukemia, acute ly,phoblastic leukemia (ALL) and acute myeloid leukemia (A
Duke - CPS - 216
Outline Online Aggregation[Hellerstein, Haas, Wang]Introduction and motivation Related work Usability and performance goals Building a systemrandom access to data GROUP BY and DISTINCT index stridingMark Fashing and Stacy Presidentto be follo
SUNY Potsdam - CS - 203
Collection ClassesBagsA Collection class is a data type that is capable of holding a group of items. In Java, Collection classes can be implemented as a class, along with methods to add, remove, and examine items.For the first example, thin
Kentucky - ECO - 401
ECO 401-002 Spring 2006 Problem Set #6 Due: Thursday, April 13, 2006 1. Problem 9.1, p. 347, in the textbook. 2. True or false, and explain with a diagram: To maximize profit, a firm should produce that output where the difference between price and a
Duke - CPS - 296
CPS 296.2 - Computational Game Theory and Mechanism DesignHomework 3 (due 11/21)Note the rules for assignments on the course web page. Show all your work, but circle your final answer. Contact Vince (conitzer@cs.duke.edu) with any questions. 1. (S
Duke - CPS - 111
1279Algebraic Linear SystemsAn algebraic linear system is a set of m equations in n unknown scalars, which appear linearly. Without loss of generality, an algebraic linear system can be written as follows: Ax = b (54)where A is an m n matrix
Duke - CPS - 258
Iteration Method for the Solution of the Eigenvalue Problem of Linear Differential and Integral OperatorsIn 1946, the Office of Naval Research asked NBS to help found a national center for mathematical computation. This center was to have four parts
Duke - CPS - 296
Chapter 12 The Winner Determination ProblemDaniel Lehmann Rudolf Mller u Tuomas Sandholm1IntroductionThis part of the book gives a comprehensive overview of the computational challenges in solving the winner determination problem (WDP): given
Duke - CPS - 110
Real Time Schedulers Real-time schedulers must support regular, periodic execution of tasks (e.g., continuous media). CPU Reservations "I need to execute for X out of every Y units." Scheduler exercises admission control at reservation time: appl
Duke - CPS - 296
Software Development Infrastructure for Sensor Networks Operating systems (TinyOS) Resource (device) management Basic primitives Protocols (MAC, routing) {covered in many previous lectures} Programming languages and runtime environments Low le
Duke - CPS - 221
Executing a Program on the MIT Tagged-Token Dataflow Architecture*(This is true black magic) Presented by: Michael BauerECE 259/CPS 221 Spring Semester 2008 Dr. Lebeck* Based on "Executing a Program on the MIT Tagged-Token Dataflow Architecture" i
Duke - OCPS - 214
Overview Congestion and the Role of RoutersJeff Chase Duke University Problem is "Bullies, Mobs, and Crooks" [Floyd] AQM / RED / REM ECN Robust Congestion Signaling XCP PushbackStoica Following slides are from Ion Stoica at Berkeley, with s
Duke - UPE - 301
Duke - CPS - 216
Query OptimizationCPS 216 Advanced Database SystemsWavelets Mathematical tool for hierarchical decomposition of functions and signals Haar wavelets: recursive pair-wise averaging and differencing at different resolutions Simplest wavelet basis,
Duke - ECE - 254
OutlineECE 254 / CPS 225 Fault Tolerant and Testable Computing Systems Concepts in Fault ToleranceCopyright 2006 Daniel J. Sorin Duke University Basic Concepts Forward Error Recovery Backward Error Recovery Redundancy in its Many Forms Ph
Duke - ECE - 254
OutlineECE 254 / CPS 225 Fault Tolerant and Testable Computing Systems Concepts in Fault ToleranceCopyright 2008 Daniel J. Sorin Duke University Basic Concepts Forward Error Recovery Backward Error Recovery Redundancy in its Many Forms Ph
Duke - CPS - 221
STARFIRE:Extending the SMP EnvelopeAlan Charlesworth Presented By Bob KoutsoyannisThe Nature of StarfireAcomplex Snoopy-Bus-Based UniformMemory-Access System. 1997 UltraSparc-II (250MHz) More to comeOutline of Key Points CCSMPDesign Ch
Duke - CPS - 296
Lecture notes 2: ApplicationsVincent ConitzerIn this set of notes, we will discuss a number of problems of interest to computer scientists where linear/integer programming can be fruitfully applied.1Maximum flowIn the maximum flow problem, we
Duke - CPS - 110
Starvation and DeadlockAnother EventBarrier ExampleEventBarrier channel;void OutputThread { while (TRUE) { ComputeDataToSend(); channel.Wait(); SendData(); channel.Complete(); } }Invariants: 1. Output thread never blocks in Wait() if the chann
Duke - CPS - 004
Inheritance & InterfacesInheritance & InterfacesInheritance & Interfaces The PlanMotivate Inheritance Motivate Interfaces Examples Continue Practice on Observer/Observable Inheritance MotivationInheritance in Java is achieved throug
Duke - CPS - 004
Recursion & the Minimax AlgorithmKey to Acing Computer Science If you understand everyting, ace your computer science course. Otherwise, study & program something you don't understand, then see Key to Understanding Computer Science.Winning Tic-ta