EE60094: Biomedical Signal Processing

EE60094
Course name	BIOMEDICAL SIGNAL PROCESSING
Offered by	Electrical Engineering
Credits	4
L-T-P	3-1-0
Previous Year Grade Distribution
{{{grades}}}
Semester	{{{semester}}}

Syllabus

Syllabus mentioned in ERP

Overview:Biomedical Signal Processing is an emerging discipline that has been defined as the study, invention and implementation of structures and algorithms to improve the communication, understanding and management of information in the form of digital signal. This paradigm emphasizes on the integration of data, knowledge, and tools necessary for efficient knowledge discovery in the decision-making process associated with information extraction and computational intelligence. The aim of this course is to present an overview of different methods used in biomedical signal processing. Signals with bioelectric origin are given special attention and their properties and clinical significance are reviewed. In many cases, the methods used for processing and analyzing biomedical signals are derived from a modeling perspective based on statistical signal descriptions. The purpose of the signal processing methods ranges from reduction of noise and artifacts to extraction of clinically significant features. The course gives each participant the opportunity to study the performance of a method on real, biomedical signals. It covers principles and algorithms for processing both deterministic and random signals. Topics include data acquisition, imaging, filtering, coding, feature extraction, and modeling.Objective: â¢To develop an in-depth understanding with the basic theoretical knowledge and practical experience from the area of biomedical signal processing.â¢To learn the fundamentals of design of biomedical signal processing system and its challenges to design the sensors of physiological variables in the wired and wireless paradigm.â¢To know the various statistical techniques for processing and visualization of physilological information on the digital hardware platform.Course ContentCourse Overview: Review of Probability, Stochastic Process, and Signal ProcessingPhysiology and characteristics of biomedical signalsNon-linear Signal Processing, Morphological Signal processingBiomedical signal processing by random fractal theory: Long-range correlation and the Hurst parameter, Structure-functionâbased multifractal formulation, Cascade multifractal formulationBiomedical signal processing by chaos theory: Feature identification using Lyapunov exponent, Feature identification using fractal dimension, Feature identification using entropy formalism (including Kolmogorov entropy, approximate entropy, sample entropy, permutation entropy)Analysis of Nonstationary signals: Non-stationary second-order statistics, Case-studies with Heart sound, murmurs, EEG rhythms and waves, knee joint vibration signals and breath sound signals, Fixed segmentation (short-time Fourier transform), Adaptive segmentation-Spectral Error Measure, ACF distance, Generalized likelihood ratio, ACF, SEM and GLR methods, RLS filter and RLS lattice filterHeart-rate variability (HRV): Relationship with blood pressure, Diabetis, respiration, renal failure, duugs, smoking alcohol, sleep, fatigue, age and gender with HRV. Nonlinear methods of analysis like capacity dimension, correlation dimension, Lyapunov exponent, Hurst exponent, Detrended fluctuation

�analysis, entropies, fractal dimension, recurrence plot Pattern Classification and Diagnostic Decision: Examples with ECG, EEG, Murmur, Supervised pattern classification, discriminant, decision and distance functions, Nearest neighbor rule, Unsupervised classification and clustering, probabilistic models, likelihood functions, Leave-One-Out Cross Validation(LOOCV), Measures of diagnostic accuracy, ROC, McNemarâs test, Reliability of classifiers and decisions, Neural networks, SVM Matched and Wiener filter â filtering and detection of objects in ultrasound signals,Multiresolution analysis of Heart and lung sound, Multivariate analysis-PCA, ICAGenomic and proteomic Signal Processing: Gene localization by 1/3 periodicity, DNA spectrogram, 1/f behavior, Frequency domain Protein hotspot localization, Role HMM, profile HMM, identification of non-coding genes, and secondary structures by filtering techniques, DNA microarrayConclusions, Feedback, Expectations