(SEM VIII) THEORY EXAMINATION 2024-25 BIO MEDICAL SIGNAL PROCESSING

BIO-MEDICAL SIGNAL PROCESSING (KOE082)

B.Tech – Semester VIII | Theory Examination (2024–25)

SECTION A

(Attempt all questions – brief but explanatory)

a) Basic Characteristics of Biomedical Signals

Biomedical signals are physiological signals generated by the human body as a result of biological processes. These signals are generally low in amplitude and frequency and are non-stationary in nature, meaning their statistical properties change with time. Biomedical signals are often nonlinear and highly susceptible to noise and artifacts arising from both internal and external sources. Due to their biological origin, these signals vary significantly from person to person and even within the same individual under different physiological conditions.

b) Electroencephalography (EEG) and Its Clinical Importance

Electroencephalography is a technique used to record the electrical activity of the brain by placing electrodes on the scalp. EEG signals represent the collective activity of neurons and are widely used in clinical diagnosis. EEG plays a crucial role in detecting neurological disorders such as epilepsy, sleep disorders, brain tumors, and coma assessment. It is also used in brain–computer interfaces and cognitive research due to its non-invasive nature.

c) Significance of Heart Sound (Phonocardiogram) Analysis

Heart sound analysis involves recording and studying the acoustic signals produced by the heart during its mechanical activity. A phonocardiogram helps in identifying abnormalities in heart valves, blood flow, and cardiac rhythm. It is particularly useful for diagnosing heart murmurs, valve defects, and congenital heart diseases. This technique provides valuable complementary information to electrocardiography.

d) Baseline Wander in ECG Signals

Baseline wander is a low-frequency noise component present in electrocardiogram signals that causes slow fluctuations in the ECG baseline. It is mainly caused by patient movement, respiration, and changes in electrode contact. Baseline wander can distort important ECG features such as ST segments, making accurate diagnosis difficult. Hence, its removal is essential for reliable ECG analysis.

e) Role of AZTEC Algorithm in Signal Compression

The AZTEC algorithm is a data compression technique used in biomedical signal processing, particularly for ECG signals. It approximates the signal using horizontal segments and slope segments, thereby reducing the amount of data required to represent the signal. The AZTEC algorithm helps in efficient storage and transmission of biomedical signals without significant loss of clinically important information.

f) Signal-to-Noise Ratio (SNR) in Biomedical Signals

Signal-to-noise ratio is a measure that compares the strength of a desired biomedical signal to the level of background noise present in it. A higher SNR indicates better signal quality and clearer physiological information. In biomedical applications, maintaining a high SNR is essential for accurate diagnosis, monitoring, and analysis of physiological conditions.

g) Evoked Potential (EP) Estimation

Evoked potential estimation refers to the extraction of small, stimulus-related signals from the background EEG activity. These potentials are generated in response to specific sensory stimuli such as visual, auditory, or somatosensory inputs. EP estimation is used to assess the functional integrity of neural pathways and is important in diagnosing neurological disorders.

h) Maximum Entropy Method (MEM) in Spectral Estimation

The Maximum Entropy Method is a spectral estimation technique that estimates power spectral density by maximizing entropy under known constraints. MEM provides high-resolution spectral estimates even with short data records. It is especially useful in biomedical signal processing where signals are often short and noisy.

i) Adaptive Filtering in EEG Signal Denoising

Adaptive filtering is used in EEG signal denoising to remove noise and artifacts that vary with time. The filter continuously adjusts its parameters based on the input signal characteristics. This approach is effective in eliminating interference such as eye blinks, muscle activity, and power line noise, thereby improving the accuracy of EEG analysis.

j) Applications of Wavelet Transform in Biomedical Signal Processing

Wavelet transform is widely used in biomedical signal processing due to its ability to analyze signals in both time and frequency domains. It is applied in ECG denoising, EEG analysis, feature extraction, compression, and detection of abnormalities. Wavelet transform is particularly effective for non-stationary biomedical signals.

SECTION B

(Attempt any three – long answers)

a) Types of Biomedical Signals and Their Acquisition Methods

Biomedical signals can be classified into electrical, mechanical, acoustic, chemical, and thermal signals based on their origin. Electrical signals such as ECG, EEG, and EMG are acquired using surface or invasive electrodes. Mechanical signals like blood pressure are measured using pressure transducers, while acoustic signals such as heart sounds are recorded using microphones or electronic stethoscopes. Proper acquisition techniques are essential to ensure signal accuracy and patient safety.

b) Noise and Artifact Removal in ECG Signals

ECG signals are affected by various types of noise such as power line interference, baseline wander, muscle noise, and electrode motion artifacts. Techniques such as digital filtering, adaptive filtering, wavelet-based denoising, and signal averaging are used to remove these disturbances. Effective noise removal enhances the visibility of ECG features and improves diagnostic reliability.

c) Fan and Turning Point Algorithms for Data Reduction

The Fan and Turning Point algorithms are data reduction techniques used to compress biomedical signals. The Fan algorithm approximates the signal using straight line segments within a tolerance limit, whereas the Turning Point algorithm retains only local maxima and minima. While both methods reduce data size, the Fan algorithm generally provides better accuracy, whereas the Turning Point algorithm is simpler and faster.

d) Importance of EEG Signal Processing in Neurological Disorders

EEG signal processing is essential for detecting and analyzing neurological disorders such as epilepsy, Alzheimer’s disease, and sleep disorders. Advanced processing techniques help in identifying abnormal patterns, frequency changes, and event-related potentials. EEG analysis supports early diagnosis, treatment planning, and monitoring of neurological conditions.

e) Adaptive Noise Canceling in Biomedical Signal Processing

Adaptive noise canceling involves using a reference noise signal and an adaptive filter to remove noise from biomedical signals. This technique is particularly useful when noise characteristics are unknown or time-varying. It improves signal quality without distorting the underlying physiological information.

SECTION C

Acquisition of Heart Sound Signals and Challenges

Heart sound signals are acquired using electronic stethoscopes or microphones placed on the chest wall. The acquisition process is affected by challenges such as ambient noise, motion artifacts, and variability in heart sounds. Proper sensor placement and signal processing techniques are required to obtain reliable 

phonocardiogram recordings.

Electrooculography (EOG) and Clinical Applications

Electrooculography measures the electrical potential generated by eye movements using electrodes placed around the eyes. EOG is used to diagnose eye movement disorders, monitor sleep stages, and support human–computer interaction systems. It is a simple and non-invasive technique with wide clinical applications.

Sources of Artifacts in EEG and Their Removal

Artifacts in EEG signals arise from eye blinks, muscle movements, electrode motion, and external electrical interference. Techniques such as filtering, independent component analysis, and adaptive filtering are used to remove these artifacts. Artifact removal is essential for accurate EEG interpretation.

Significance of Heart Rate Variability (HRV) Analysis

Heart Rate Variability analysis evaluates the variation in time intervals between consecutive heartbeats. HRV is an important indicator of autonomic nervous system function and cardiovascular health. It is used in stress analysis, cardiac risk assessment, and monitoring of chronic diseases.

Run-Length Coding in Biomedical Data Compression

Run-length coding is a lossless compression technique that represents consecutive identical values using a single value and its count. In biomedical signal processing, it is used to reduce data size for efficient storage and transmission, particularly in signals with repetitive patterns.

Huffman Coding for Biomedical Signals

Huffman coding is a variable-length encoding technique that assigns shorter codes to frequently occurring symbols. It is widely used in biomedical signal compression to reduce storage requirements while preserving signal integrity.

Maximum Likelihood Method for Spectral Estimation

The Maximum Likelihood Method estimates the power spectrum by maximizing the probability of observing the given signal. It provides accurate spectral estimates but involves high computational complexity. In biomedical applications, it is useful for analyzing EEG and other physiological signals.

Epilepsy Detection Using EEG

EEG-based epilepsy detection involves identifying abnormal rhythmic patterns and spikes associated with seizures. Signal processing techniques such as frequency analysis, wavelets, and machine learning help in early diagnosis and monitoring of epilepsy.

Least Squares Method for Adaptive Filtering

The Least Squares method minimizes the error between the desired signal and the filter output. It is used in adaptive filtering applications to remove noise from biomedical signals. This method provides accurate results and is widely used in ECG and EEG denoising.

Wavelet Transform in ECG Denoising

Wavelet transform is highly effective in ECG denoising because it can separate noise from useful signal components at different scales. Compared to traditional filters, wavelet-based methods preserve important ECG features while removing noise more efficiently.