About this previous year paper
Subject Code: BEC056
Maximum Marks: 70
Time: 3 Hours
Paper ID: 310335
Question Paper Overview
SECTION A (2 × 7 = 14 Marks)
(Short conceptual questions — basic principles and signal analysis)
a. Define bio-medical signals.
b. List the difficulties faced during acquisition of bio-medical signals.
c. Define portable arrhythmia monitors.
d. Explain ST segment analysis and its clinical relevance.
e. Examine the main objective of data reduction in bio-medical signal processing.
f. Describe linear prediction theory in the context of EEG analysis.
g. List the general structures of adaptive filters.
SECTION B (Attempt any three × 7 = 21 Marks)
a. Explain the classification of bio-medical signals and their acquisition methods.
b. Explain the methods used for QRS detection in ECG signals.
c. Explore the Turning Point algorithms for data reduction.
d. Explain the maximum entropy and autoregressive (AR) methods for EEG analysis.
e. Explain signal averaging and its role in evoked potential estimation.
SECTION C (Attempt one part from each question × 7 = 35 Marks)
Q3
(a) Describe the challenges in acquiring bio-medical signals.
OR
(b) Explain the basics of ECG (Electrocardiography) and EEG (Electroencephalography).
Q4
(a) Illustrate the techniques to remove baseline wander in ECG signals.
OR
(b) Explore the working of portable arrhythmia monitors and their clinical importance.
Q5
(a) Explain the Fan algorithm and its application in bio-medical signal processing.
OR
(b) Explain Huffman and modified Huffman coding techniques.
Q6
(a) Illustrate the dynamics of sleep/wake transition in EEG signals.
OR
(b) Explain the characteristics of brain waves and their relevance in epilepsy detection.
Q7
(a) Explain the concept of adaptive wavelet detection and its advantages.
OR
(b) Illustrate the matched filtering technique for detecting overlapping wavelets.
Key Topics for Revision
1. Bio-Medical Signals
Represent physiological processes of the human body.
Examples:
ECG (cardiac activity) EEG (brain activity)
EMG (muscle activity) EOG (eye movement)
GSR (skin response)
Types:
Electrical: ECG, EEG, EMG Non-electrical: Blood pressure, respiration rate
2. Signal Acquisition Challenges
Motion artifacts, electrode placement errors. Power line interference, baseline drift, and noise.
Physiological variability among subjects.
Solutions: High-quality sensors, proper grounding, and adaptive filtering.
3. Portable Arrhythmia Monitors
Used to record irregular heartbeats (arrhythmias) over extended periods.
Examples: Holter monitors, event recorders, smart ECG patches.
Importance: Detects transient abnormalities that normal ECG misses.
4. ST Segment Analysis
Evaluates ischemic heart conditions. Elevation → Myocardial infarction.
Depression → Ischemia or ventricular strain.
Clinical relevance: Detects cardiac events in ECG monitoring.
5. Data Reduction in Biomedical Signal Processing
Objective: Compress data while preserving diagnostic information.
Techniques:
Turning Point algorithm Fan algorithm
Huffman coding Principal Component Analysis (PCA)
6. Linear Prediction in EEG Analysis
Models EEG signal as a linear combination of past values.
Predicts future samples using autoregressive (AR) modeling.
Helps in feature extraction and abnormality detection (e.g., seizure).
7. Adaptive Filters
Structure adjusts filter coefficients to minimize error dynamically.
Used for noise cancellation and baseline drift removal.
Algorithms: LMS (Least Mean Square), RLS (Recursive Least Square).
8. QRS Detection Techniques
| Method | Description |
|---|---|
| Derivative & Thresholding | Detects slopes of QRS peaks |
| Pan–Tompkins Algorithm | Widely used, robust against noise |
| Wavelet Transform | Multi-resolution analysis for ECG peaks |
9. EEG Signal Analysis
| Method | Description |
|---|---|
| Maximum Entropy | Spectral estimation using probability models |
| AR Method | Time-domain signal modeled via past samples |
| Applications: Sleep studies, epilepsy detection, brain-computer interfaces (BCIs). |
10. Baseline Wander Removal in ECG
Causes: Respiration, electrode movement, poor contact.
Techniques: High-pass filtering
Polynomial fitting Wavelet decomposition
Adaptive filtering (LMS-based)
11. Fan Algorithm
Used for data compression and feature extraction.
Identifies major turning points, reducing redundant samples.
Maintains important morphology for diagnosis.
12. Huffman & Modified Huffman Coding
Lossless data compression techniques. Assigns shorter codes to frequent symbols.
Modified Huffman: Adapts dynamically for changing data patterns.
13. EEG and Sleep Analysis
Sleep/Wake Dynamics: Characterized by EEG frequency bands.
Delta (0.5–4 Hz): Deep sleep Theta (4–8 Hz): Light sleep
Alpha (8–13 Hz): Relaxed wakefulness Beta (13–30 Hz): Active thinking
Transitions show shifts in dominant frequency bands.
14. Brain Waves & Epilepsy Detection
| Wave | Frequency | Clinical Importance |
|---|---|---|
| Delta | 0.5–4 Hz | Deep sleep |
| Theta | 4–8 Hz | Emotional stress |
| Alpha | 8–13 Hz | Relaxed state |
| Beta | 13–30 Hz | Alertness |
| Gamma | >30 Hz | Cognitive processing |
Epileptic seizures often show spike-and-wave patterns in EEG.
15. Adaptive Wavelet Detection
Wavelets: Time–frequency localized functions.
Adaptive wavelet adjusts its parameters based on the signal’s features.
Used in QRS complex detection, EEG event identification, and noise suppression.
16. Matched Filtering
Used to detect known waveform patterns (like QRS or spikes).
Maximizes signal-to-noise ratio (SNR) for identifying overlapping wavelets.
Exam Strategy Tips
Prepare flow diagrams for algorithms (Pan–Tompkins, Fan, Huffman).
Revise formulas for filtering and AR modeling.
Use labeled EEG/ECG waveform diagrams.
For long answers, focus on clinical significance and practical applications.
Review compression and data reduction examples with brief numerical illustrations.