K.

RAMAKRISHNAN COLLEGE OF ENGINEERING

UAM1603-HEALTH CARE ANALYTICS
QUESTION BANK
UNIT-1
2 Marks
1.Define Electrogastrogram.
Measures the electrical activity in the stomach before, during and after food
2. What is Genomic Data Analysis?
Genetic in nature but nature of the causality between genetic markers and the diseases has not been
fully established. Eg:Diabetes
Full set of genetic markers that make an individual prone to diabetes are unknown. Genetic data
modeled into Sequence and networks
3. Distinguish between ECG and EGG.
ECG: Records the electrical signal from the heart to check for different heart conditions. Electrodes
are placed on the chest to record the heart’s electrical signals which causes the heart to beat. The signals
are shown as waves on an attached computer monitor or printer.
EGG: Test that measures electrical activity in the brain using small metal discs(electrodes) attached
to the scalp. Brain cells communicate via electrical impulses and are active all the time even during
asleep.
4. List out the categories of Supervised Learning methods for Predictions.
Statistical methods, Sophisticated methods and Survival models
5. What are the Challenges in Pervasive Health Analytics?
Knowledge Extraction and Real time processing
6. Define Electromyogram.
Measures muscle response (or) electrical activity in response to a nerve’s stimulation of the muscle
7. What is Clinical Text Mining?
Information about patients encoded in Clinical Notes
Stored in unstructured data format
Clinical Information
-Transcription of dictations
-Direct entry by providers
-Use of speech recognition
Converting clinical text to structured format will be more difficult
NLP and Entity Extraction are used
8. Distinguish between Pre-marketing and Post Marketing.
Pre-marketing: Focus on discovery activities such as finding signals that indicate relation b/w
drugs and targets, drugs and drugs, genes and diseases, Protein and diseases and finding biomarkers
(Objective measures that captures what is happening in a cell (or) Organism at a given moment)
Post-marketing: Find indications of adverse side effects for approved drugs. Provide list of
potential drug side effect associations
9. List out the major sources of temporal Data mining.
Electronic Health Records and Sensors
10. What are the particular fields in Clinical Decision Support Systems?
Pharmacy and Billing

16 Marks
1. Various sources of Healthcare Data
2. Advanced Data Analytics for Healthcare
3. Applications and Practical Systems for Healthcare
4. Components of Electronic Health Record
5. Challenges faced while the usage of Electronic Health Record
6. Coding Systems
7. Benefits and Barriers of EHR

UNIT-II
2 Marks
1.Define Hounsfield Units.
Measure of attenuation property which describes the difference between X-rays absorbed by body
and X-rays transmitted through the body
2.Distinguish Over Segmentation and Under Segmentation.
over segmentation-objects broken into many pieces
under segmentation-objects merged/missing
3.What are the different types of sensors?
1.Physiological sensors-ECG sensors, Temperature sensors
2.Contextual Sensors-RFID Sensors
3.Human Sensors-Care provider notes, entries in HER
4.Activity Sensors-Wearable devices

4.What are the two sets in watershed Transform?

Catchment basins and Watershed lines
5.What are the disadvantages of Thresholding?
-Objects must have similar appearance
-Does not take any spatial information into an account
-Results in holes in objects
-Works for simpler tasks
6.Define Biomedical Image.
Subset of images of biological specimen (human and Animal)
7.What is Intensive Care Data Mining?
Critically ill patients are often attached to large numbers of body sensors which are connected to
sophisticated monitoring devices producing the large volume of physiological data
8.What is Object Feature?
• O/p is a set of foreground regions that are separated from background
• Algorithm-connected components used to separate the foreground regions into individually
labeled regions where each unique label indicates a unique object in image
• Basic categories-shape, size, brightness and texture
9.Define PACS.
Medical Imaging Data stored and managed using Picture Archiving and Communication Systems
mainly for image modalities and electronic reports
10.List the types of transforms in Image Registration.
1.Rigid Body Transform
2.Similarity Transform
3.Affine Transform
4.B-Spline Deformable Transform
11.Define DICOM.
Digital Imaging and Communications in Medicine (DICOM, dicom.nema.org) is a widely used
standard that helps achieve this for the purposes of handling, storing, printing, and transmitting medical
imaging data. It defines a file format and a network communications protocol for these data types

16 Marks
1.Biomedical Imaging Modalities
2.Object Detection
3.Image Registration
4.Feature Extraction
5.Challenges in Health Care Data Analysis
6.Sensor Data Mining Applications
7.Non-Clinical Health Care Applications

UNIT-III

1.Define Action Potentials.

A nerve impulse, or an action potential, is a series of electrical responses that occur in the cell as
a consequence of mechanical contraction of a single cell, when stimulated by an electrical current.
Action potential is caused by the flow of certain ions such as sodium (Na+), potassium (K+), and
chloride (Cl-) along with other ions across the cell membrane.

2.What is EEG?
The electroencephalogram (EEG) represents the electrical activity occurring at the surface of the
brain.

3.Distinguish between Depolarization and Hyperpolarization.

This change in the membrane potential, known as depolarization, causes the voltage gated sodium
channels to open
The membrane potential becomes even more negative than the resting potential for a brief period; this
is called hyperpolarization

4.List out the various kind of noise which affects ECG.

(a)Power line interference
(b) Electrode contact noise
(c) Motion artifacts
(d) Muscle contraction (electromyographic, EMG)
(e) Baseline drift and ECG amplitude modulation with respiration
(f) Instrumentation noise generated during signal acquisition
(g) Electrosurgical noises, and many other less significant noises

5.What are the approaches in Denoising signals?

Statistical approach (principal component analysis), a nonstationary filtering technique (wavelet),
and an optimum filter in wavelet domain (wavelet-Wiener).
6.What are the stages in Pan-Tompkins Algorithm?
(i)Bandpass filter
(ii)Derivative
(iii)Squaring
(iv)Moving Window Integral
(v)Threshold Selection

7.What are the additional Biomedical Signals?

Carotid pulse (CP), the speech signals, signals recorded using catheter-tip sensors (signals such as
left ventricular pressure, right atrial pressure, aortic pressure), the vibromyogram (vibration signals that
accompanies EMG).

8.Define Wavelet Filtering.

Discrete Wavelet Transform (DWT) is a method, which has the ability to represent a large class of
well-behaved functions with a sparse representation in wavelet space. So, when DWT is applied to a
noisy ECG signal, the noise-free component of the signal (true signal) will be concentrated in a small
number of larger coefficients, while the noise will be distributed as smaller coefficients.

9.Define Spatial Smoothing.

Spatial smoothing is a preprocessing tool commonly applied to reduce the amount of noise in
functional magnetic resonance imaging (fMRI) data.

10. Define Functional Connectivity.

Functional connectivity is a measure of how regions of the brain interact with each other. Brain
activation can be mapped non-invasively with functional MRI (fMRI). There are various scales at
which one can measure functional connectivity.

16 Marks
1.Types of Biomedical Signals
2.ECG Signal Analysis
3.Denoising Signals
4.Multivariate Biomedical Signal Analysis
5.Cross-Correlation Analysis
6.Genomic Data Generation
7.Methods and Standards for Genomic Data Analysis
8.Types of Computational Genomic Studies

UNIT-IV
1.Define CDA.
Clinical data architecture (CDA), an XML standardized model for patient records further contribute to
substantial improvements in health care. The clinical information in EHR and CDA is the primary
source on medical history of patients that is normally expressed using natural language

2.Distinguish Overgeneration and Under generation.

Overgeneration = FP/(TP+FP) Under generation = FN/ (TP+FN)

3.What are the challenges occurring in clinical text processing?

Ungrammatical nature of short and telegraphic phrases, dictations, shorthand lexicons such as
abbreviations and acronyms, and often misspelled clinical terms and presence of alphabets.
4.What are the reasons for lag in processing clinical text data?
(1) limited access to clinical data as the patient’s medical history is protected confidentially
(2) challenges involved in creating large volume of shared data, tasks, annotation guidelines,
annotations, and evaluation techniques

5.What are the existing methodologies for coreference resolution?

1.Heuristics-based approaches based on linguistic theories and rules
2. Supervised machine learning approaches with binary classification of markable mention/entity
pairs or classification by ranking markable.
3. Unsupervised machine learning approaches, such as nonparametric Bayesian models or
expectation maximization clustering

6. What are the challenges in clinical report processing?

1.Domain Knowledge
2. Confidentiality of Clinical Text
3. Abbreviations
4. Diverse Formats
5. Expressiveness
6. Intra- and Interoperability
7. Interpreting Information

7.What is the use of NLP in Pharma?

NLP helps medical professionals identify eligible candidates for clinical trials from large quantities
of unstructured databases. This helps patients that need immediate medical assistance, saves lives, and
accelerates innovation.

8.What is NLP pipeline in healthcare?

The process flow of a medical natural language processing (NLP) pipeline, which transforms a
narrative sentence in a clinical note to structured outcome.

9.What are the applications of NLP?

Clinical Documentation, Clinical trial matching, Clinical Decision Making, Clinical decision
support, Medical Coding, Clinical research, Sentimental analysis, Chatbots, Electronic Health Record.

10.Formulas
16 Marks
1.NLP
2.Core Components of NLP
3.Mining Information from Clinical Text
4.Information Extraction
5.Current Methodologies
6.Challenges of Processing Clinical Reports
7.Clinical Applications

UNIT-V
1.What is the formula for Bayes Theorem?

2. Distinguish between Survival Function and Hazard Function.

The object of primary interest of survival analysis is the survival function, which is the probability
that the time to the event of interest
One other function commonly used in survival analysis is the hazard function (λ(t)), which is also
known as the force of mortality, the conditional failure rate, or the instantaneous death rate.
3.Define Learning Rate.
The learning rate, denoted by the symbol α, is a hyper-parameter used to govern the pace at which
an algorithm updates or learns the values of a parameter estimate.
4.What is Reinforcement Learning?
Reinforcement learning (RL) is a machine learning (ML) technique that trains software to make
decisions to achieve the most optimal results
5.What is Named Entity Recognition?
Named entities (NEs) are phrases or combinations of phrases that denote biomedical terms (e.g.,
proteins, genes, diseases, drugs, cells, chemicals). Automatically extracting them, a task known as
named entity recognition (NER), involves the demarcation of entity names of a specific semantic type,
e.g., proteins. It results in annotations corresponding to a name’s in-text locations as well as the
predefined semantic category it has been assigned to.
6.What is Coreference Resolution?
Introducing new information into the discourse by means of coreferring noun phrases (e.g.,
appositives) helps boost informativeness. Text written with these elements, however, can become
difficult to interpret for automated systems. This problem is addressed in a natural language processing
task known as reference resolution that aims to bring out the meaning of a mention (referring
expression) by determining the concept to which it refers (referent)
7.Define Discourse Relation Recognition.
The connections between textual spans, also known as discourse relations, make the text coherent
and cohesive, and their automatic discovery can lead to a better understanding of the conveyed
knowledge.
8.Define Bayesian Network.
A Bayesian network introduces a directed acyclic graph (DAG), which represent a set of random
variables by nodes and their dependency relationships by edges. Each node is associated with a
probability function that gives the probability of the current node conditional on its parent nodes’
probability. If the node does not have any parents, then the probability function will be the prior
probability of the current node.
9.Define Markov Random Field.
A Markov random field (MRF), Markov network or undirected graphical model is a set of random
variables having a Markov property described by an undirected graph.
10.What is Cost Sensitive Learning?
Cost-Sensitive Learning is a type of learning that takes the misclassification costs (and possibly
other types of cost) into consideration. The goal of this type of learning is to minimize the total cost.

16 Marks
1.Evaluation and Validation
2.Survival Trees
3.Cox Proportional Hazards Model
4.Non-Parametric Survival Analysis
5.Advanced Prediction Models
6.Survival Models
7.Logistic and Linear regression
8.Alternate Clinical Prediction Methods