[ Home ] [ Up ] [ Exam 1 Prep ]

 

 

BIEN 289/157: Topics in Intelligent Biosystems

 

Instructor:  Jack Winters, Ph.D.  (office hours: Tu-Th 2-4,  W-F 9-10 or by appnt.)

Time:  Tu-Th 4:20 - 5:35 PM

Prereq:  BIEN 155 (Biomedical Signals & Systems) or consent of instructor  

Catalog Description:  Principles and performance of intelligent biosystems, with emphasis on the use of simulation as a tool to understand adaptive bioprocesses and clinical decision-making.  Survey of intelligent "soft computing" tools (adaptive neural networks, fuzzy systems, evolutionary computing), with special focus on recurrent neurocontrollers for physiologic systems and on fuzzy expert systems for clinical diagnosis (including integration of biomonitored signals).  Includes self-selected student project using tools from one or both of these two areas.   (3)

Textbook:  There is no required book.  Materials from the following will be used:

Neuro-Fuzzy and Soft Computing, J-S.R. Jang et al, 1996

Neural Fuzzy Systems, C.-T. Lin & G. Lee, Prentice Hall, 1996

Computational Intelligence PC Tools, Eberhart et al, AP Professional, 1996

Fuzzy and Neural Approaches in Engineering, Tsoukalas & Uhrig, Wiley, 1997 

Biomechanics and Neural Control of Posture and Movement, J.M. Winters & P.E. Crago, eds., Springer-Verlag, 2000

Selected short writings by famous scientists on mind-body phenomena, biocybernetics, complexity and living tissues (Descartes, De Vinci, James, Einstein, Wiener, Eccles, Arbib, Fung, ...), e.g. from The Brain Project, S. Jones (web-based) and Chapter 2 (history) in Eberhart.

Three manuals will also be of value:

Online Matlab Manual (Fuzzy Logic Toolbox, Neural Network Toolbox, Simulink)

Manual for the General Fuzzy Inference Engine (GFIE) Package, A. O'Brien

Manual for the Musculoskeletal Modeling in Simulink (MMS) Package, G. Loeb et al

Course Objectives: 

to help awaken your own mind to the timeless mystery of trying to understand its own function  

to understand the basic principles and terminology of feedback and feedforward control, adaptive biosystems, optimization, learning and memory processes, and distributed tissue remodeling

to gain experience using the Matlab-Simulink environment, as well as parts of the Neural Network and Fuzzy Systems Toolboxes.  

to understand the core principles and tools of Soft Computing (dynamic neural networks, fuzzy systems, and genetic/optimization algorithms).

to use modeling/simulation to develop a stronger foundation in two selected subsets:

adaptive neuromuscular biocontrol processes: a) neuromotor learning for goal-directed movement tasks (using adaptive recurrent neurocontrollers along with simple musculoskeletal models); and b) rehabilitative tissue adaptation with exercise,

clinical decision-support fuzzy expert systems: simple rule-based systems that integrate expert medical reasoning and objective sensor physiologic data (e.g., ECG).

to successfully implement an intelligent neuro/fuzzy system of your choice (fuzzy expert system or adaptive neurocontroller), perhaps related to your professional interests, and write a short paper on this project (in a previous course, over half of these projects were subsequently published and presented at national meetings)

Modules:

  1. Overview of Adaptive Intelligent Systems Tools for Control, Learning and Intelligent Reasoning.  Basic Structure and Function of Neuromuscular and Neuroregulatory Systems.  [2-3 weeks]

  2. Overview of Fuzzy Logic, with focus on Fuzzy Control and Fuzzy Expert Systems.  [4-6 weeks]

  3. Overview of Neural Networks, with focus on Recurrent Networks for Adaptive Neurocontrol.  Principles and Structures for Neurofuzzy/Evolutionary Optimization.  [4-6 weeks]

  4. Final Project and Presentations  [2 weeks]

Grading#: 

  • BIEN 157:

    • 50% Exams (2 @ 25%)

    • 25% Homeworks (4 @ 5%) and participation in e-discussions (5%)

    • 25% Final Project (10% deliverable, 10% paper, 5% presentation)

  • BIEN 289:

    • 40% Exams (2 @ 20%)

    • 30% Homeworks (4 @ 5%), special presentation (5%), and participation in e-discussions (5%)

    • 30% Final Project (15% deliverable, 10% paper, 5% presentation)

# BIEN 289 includes all in BIEN 157, plus one added presentation and a more extensive project