Download summary of medical information

Pertemuan 10
Title : 1. Clinical contents and eLearning. 2. XML Introduction. 3. DICOM Introduction
1. What is metadata ?
Metadata is data that describes other data, whether it is physical or electronic. The main
purpose of metadata is to facilitate in the discovery of relevant information. It helps organize
electronic resources, provide digital identification, and helps support archiving and preservation
of the resource.
2. Metadata to describe other clinical data :
-
XML tags for DICOM object
-
XDS document meta data
3. What is IHE solutions within the enterprise?
IHE (Integrating the Healthcare Enterprise) is an international organization involving healthcare
professionals and industry representatives working together to improve the way computer
systems in healthcare share information. IHE closes the gaps between disjoined systems, bridges
loosely connected departments and institutions, ensuring overall data consistency and
eliminating sources of redundant erroneous information.
4. What is XML?
XML is a set of rules for encoding documents in machine-processable form, but also is human
readable.
5. What the usages of XML in Medical?
-
As a clinical document standard (HL7 CDA)
-
Metadata to describe other clinical data.
-
Format messages for network communication or web services.
Pertemuan 11
Title : Acquisition of Medical Imaging
1. Image acquisition using X-Rays :
a. The intensity of beam is proportional to the number of x-ray photons in it.
Different tissues affect the beam depending on their thickness and the attenuation
coefficient (μ).
b. Contrast agents
They can be injected or swallowed to change the attenuation of soft tissues. Materials
with high atomic number (ex : iodine, barium) can be used to increase attenuation.
c. Image Types
-
Projection or planar x-ray radiography
-
Computed tomography (CT)
2. What is computed tomography?
-
A stack of 2D transverse planes, forming a 3D volume data.
-
With planar radiography, the superpositioning of overlapping organs complicates
their identification.
-
Tomographic imaging is a technique for producing transverse images.
-
Computer tomography is reconstruction of transverse image by manipulated the
gray values by using some techniques such as attenuation coefficients or Hounsfield
units.
3. How computed tomography works?
a. Translate-rotate
-
A tightly collimated beam minimizes scatter and radiation dose
-
The gantry rotates by a small angle and another scan performs.
b. A transverse slice of the body is schematically divided into many small volume (voxel).
c. The intensity of each voxel depends on the sum of attenuation coefficients for each
voxels in the path.
4. How CT reconstruction is used to solve for each μij in the 2D array to form an image?
a. Using matrix inversion to solve simultaneous equations.
b. Alternatives : backprojection, filtered backprojection, direct fouriour reconstruction.
5. Instead of using attenuation coefficients (μij) as the gray values directly, Hounsfield units are
used in CT. What the use of Hounsfield units?
To minimize the dependence on the energy of x-ray beam and to produce a unit-less pixel
values. HU is not as good as film-screen radiography, but free of the superpositioning of
structure.
6. How to calculate CT number by Hounsfield units?
μ−μH2O
CT number (or HU) = 1000 x μH2O
7. How to use CT by using Hounsfield units?
First, do segmentation (find the area we want to observe) based on CT number, rescale to
HU values to 0-255 scale. Then, use window level and window width to span the range of
interest in order to get the contrast color.
8. How HU in CT works?
The CT scanner uses a set of software algorithms to determine the amount of x-radiation
absorbed by every element in a plane of tissue. Each of these elements is represented by a
pixel on the video display, and the density (amount of x-radiation absorbed) is measured in
Hounsfield units.
9. How image manipulation works in CT Scan?
a. Radiodensity
The relative inability of electromagnetic radiation to pass through a particular material.
b. Windowing
Windowing is used to alter picture contrast. By this technique, the CT image grey scale
can be manipulated using the CT numbers that make up the image. The picture can be
changed to concentrate on soft tissues or dense structures, such as bone.
- Window Width : the range of CT numbers of each level of grey.
- Window Level : the centre CT value of the window width and determines which
Hounsfield numbers are displayed on the image.
c. Segmentation
To recognize regions within an image as distinct and belonging to different object. Gray
level is used to know the segment for heart, bones, liver, etc based on HU.
10. What is partial value effect / partial volume effect ?
The partial volume effect can be defined as the loss of apparent activity in small objects or
regions because of the limited resolution of the imaging system. It occurs in medical imaging
and more generally in biological imaging such as positron emission tomography (PET) and
single-photon emission computed tomography (SPECT). If the object or region to be imaged
is less than twice the full width at half maximum (FWHM) resolution in x-, y- and z-dimension
of the imaging system, the resultant activity in the object or region is underestimated.[1] A
higher resolution decreases this effect, as it better resolves the tissue.
Pertemuan 12
Chapter 16 : Patient – Care System
1. Concept of patient care
Spesific cognitive processes and therapeutic techniques very by disciplines, but all disciplines
share certain commonalities in the provision of care.
Concepts : collecting data, patient’s status, diagnostic labels, therapeutic goals – treatment,
adjustments, terminated.
2. Five perspectives from the system in hospitals :
a. Physician – diagnose, authorize care
b. Nurse – understand, teach and counsel, and help
c. Nutritionist – control diet
d. Physical therapist – teach excises
e. Occupational therapist – access abilities and limitation, teach adaptive technologies
3. Ambulatory care systems :
a. Previous
: paper based
b. Current
: laptop, handhold device, pen-based, voice-recognition, databases…
c. Old patient-care system’s drawbacks : lack the capacity to aggregate data across patients,
to query the data about subsets of patients, or to use data collected for clinical purposes
to meet informational needs of administrators or researchers.
4. Information process :
a. Data acquisition
b. Data storage
c. Data transformation or processing
d. Presentation
Pertemuan 13
Title : Essential Concepts for Biomedical Computing
1. What is CATER ?
Complete, Accurate, Timing, Economy, Relevant.
2. Knowledge -> Information -> Solution.
3. There are 4 red ball, 2 yellow ball, and 3 green ball. How to calculated to get the information?
Information equation : I(p)=-logb(p)
p = probability of the event happening
b = base
So, the answer is :
I(Red ball) = -log(4/9)
I(Yellow ball) = -log(2/9)
I(Green ball) = -log(3/9)
4. What is entropy ?
Entropy is simply the average (expected) amount of the information from the event.
Entropy equation : − ∑𝑛𝑖=1
pi logb (pi)
5. How was the entropy equation is derived ?
p = probability of the event happening
b = base
I = total information from N occurrences
N = number of occurrences
(N*Pi) = Approximated number that the
certain result will come out in N occurrence
The only thing that changed is the N is
moved to the right.
It means I/N is entropy.
6. There are 4 red ball, 2 yellow ball, and 3 green ball. How to calculated the entropy?
4
4
2
2
3
3
Entropy = -( 9 ) log ( 9 ) + -( 9 ) log ( 9 ) + -( 9 ) log ( 9 )
Therefore, you are expected to get xxx (entropy result) information each time you choose a
ball from the bin.
Pertemuan 14
Title : Biomedical Decision Making
1. How is the concept of probability useful for understanding test results and for making
medical decisions that involve uncertainty?
Because clinical data are imperfect and outcomes of treatment are uncertain, health
professionals often are faced with difficult choices. By using probability, health
professionals can have a medical reasoning to provide valuable insight between
symptoms and disease to evaluate a problem and then make medical decision (the best
possible outcomes).
2. How can we characterize the ability of a test to discriminate between disease and health?
a. A prior, prevalence (the frequency of an event in a population.)
Making an initial judgment about whether a patient is likely to have a disease. The
belief about the likelihood of disease is refined by a prior / pretest probability.
b. Sensitivity and specificity
Gathering more information, often by performing a diagnostic test, to reduce the
uncertainty. More tests will reduce more uncertainty, but it also cost more.
c. The posterior, predictive value
Update the initial probability estimate to have the posterior / post-test probability.
To calculate post-test probability, we must know the pretest probability, as well as
the sensitivity and specificity of the test.
3. What information do we need to interpret test results accurately?
The values of sensitivity and specificity. Accuracy = (TP+TN)/N. These values depend on
the cutoff value between normal and abnormal. The choice of cutoff depends on the
disease in question and on the purpose of testing.
Increase the cutoff value, the test becomes more specific but less sensitive. A lower
cutoff value, the test becomes more sensitive, but less specific. If the disease is serious
and if life-saving therapy is available, we minimize false-negative results (maximize
sensitivity as possible). If the disease is not serious and therapy is dangerous, we
minimize false-positive results (maximize specificity).
Higher the sensitivity and specificity are reducing uncertainty.
4. What is expected-value decision making? How can this methodology help us to
understand particular medical problems?
Expected value decision making is characterize each gamble by a number, and we use
that number to compare the gambles. The outcome of an individual illness is
unpredictable. Physician need to determine which of the two therapies is preferable. A
choice among therapies is a choice among gambles. It is to determine which therapies
are the best to give the best survival chance or longest length of life to patient.
5. What are utilities and how can we use them to represent patient’s preferences?
Utilities are typically expressed on a 0 to 1 scale, where 0 represents death and 1
represents ideal health. Several methods for assessing utilities : standard-gamble, timetradeoff, visual-analog scale.
6. What is sensitivity analysis? How can we use it to examine the robustness of a decision
and to identify the important variables in a decision?
Sensitivity analysis is a test of the validity of the conclusions of an analysis over a wide
range of assumptions about the probabilities and the values, or utilities.
7. What are influence diagrams? How do they differ from decision trees?
Influence diagrams and decision tress represent decision nodes as squares and chance
nodes as circles. In contrast to decision trees, the influence diagram also has arcs
between nodes and a diamond shaped value node.
8. Measures of test performance.
a. Measures of agreement/concordance : TP, TN.
TP : True Positive (sensitivity); TN : True Negative (specificity).
TPR : True Positive Rate (sensitivity); TNR : True Negative Rate (specificity).
b. Measures of disagreement/discordance : FP, FN
FP : False Positive; FN : False Negative .
c. True Positive Rate (TPR) = sensitivity.
P : Test is positive|patient has disease = P(T+|D+).
Ratio of diseased patients with a positive test : TP/(TP+FN).
d. True Negative Rate (TNR) = specificity.
P : Test is negative|patient has no disease = P(T-|D-).
Ratio of non-diseased patients with a negative test : TN/(TN+FP)
e. False Negative Rate
FNR = 1 – TPR
f.
False Positive Rate
FPR = 1 – TNR
9. What is ROC (Receiver-Operating Characteristic) Curve?
The range of values of sensitivity and specificity over all possible cutoffs (TPR vs FPR).
10. What kinds of posttest probability?
Bayes’ Theorem and Predictive Value.
Pertemuan 15
Title : Biomedical Decision Making
1. While you are planning a image processing system, what things you need to consideres?
-
New generation and future generations
-
Opportunities
-
Why it will make money?
-
Why do so many people like to use?
-
Sphere of influence
-
Creativity
2. What the use application of computer image processing system in healthcare?
-
Ultrasound image processing
-
Cardiac ultrasound imaging
-
Computer tomography (CT) imaging
-
Cardiovascular calcification detection
-
Mammography (Memmogram)
3. What is the major concept of computer image processing system?
Digital image processing enables the enhancement of visibility for detail in images using
algorithms that apply arithmetic and statistical procedures to stored pixel values, instead of
the classical darkroom manipulations for filtration of time-dependent voltages necessary for
analog images and video signals.
4. Medical imaging system consideration?
-
Integrate the existing system. (HIS consist of RIS+CIS+LIS)
-
Scheduled image exam workflow
-
Format of digital contents
-
Communication protocol
-
Structure of data element
-
The viewer
-
Server/repository/database
Pertemuan 17
Title : Bioinformatics
1. What is postgenomic database?
A postgenomic database bridges the gap between molecular biological databases with those
of clinical importance. Example : OMIM Database.
2. What is future challenges as bioinformatics and clinical informatics converge?
-
Completion of multiple human genome sequences
Challenge : collecting individual sequence data from patients who have diseases.
There are significant problems associated with obtaining, organizing, analyzing, and
using this information.
-
Linkage of molecular information with symptoms, signs, and patients
Challenge : creating the conceptual links among these databases to create an audit
trail from molecular-level information to macroscopic phenomena, as manifested in
disease.
-
Computational representations of the biomedical literature
Challenge : many experimental data sources not in standardized way. It is an
obstacle to build a knowledge base for storing information from biological
experiment.
-
Computational challenges with an increasing deluge of biomedical data
Challenge : storing, interpreting and integrating the massive amount of datasets the
biomedical community is generating.
3. What relation between bioinformatics and clinical informatics?
It moving sequence information in biological system to structural and function information.
They both focus on representing, storing, and analyzing biological or biomedical data. These
technologies include the creation and management of standard terminologies and data
representations, the integration of heterogeneous databases, the organization and
searching of the biomedical literature, the use of machine learning techniques to extract
new knowledge, the simulation of biological processes, and the creation of knowledge-based
systems to support advanced practitioners in the two fields.