Ai ppt (1)

Fuzzy logic applied to a Patient Classification System

Presented by:

Dheeraj Mor (121416012)

Pallavi Patil (121416009)

Najuka Jagtap(121417015)

Anuja Agharkar(121417008)

CONTENTS

• INTRODUCTION• FUZZIFICATION• FUZZY CONCEPT USED.• RESULT • CONCLUSION• FUTURE SCOPE • REFERENCE

INTRODUCTION• Requirement of optimization of hospital management. - Correct number of nurses and healthcare workers

• Why patient classification required? - Nurse duty adjustment

- PCS

- Identification of PCS:

- Activity-based systems

- Dependency based systems

-Complexity level of each patient

-Economy point of view

• Why fuzzy logic? - Optimization

- Works best incase of uncertainity

FUZZIFICATION

• The process of transforming crisp values into fuzzy value.• The types fuzzification (Membership Function):– Trigonometric– Trapezoidal– Sigmoid– Gaussian etc.

• Example:

Temperature of a patient can be model into four categories :

-Normal temperature

-Hypothermia

-Hyperthermia

-Hypepyrexia

• Body temperature = {Low, Medium, High}

FUZZY CONCEPT USED• Grouping patients according to nursing care required.

• An algorithm MAP(Metodo Assistenziale Professionalizzante – Professionalizing Healthcare Method)

• MAP evaluates patients on his clinical condition and environment.

• Three dimensions are:

1)Clinical stability

2)responsiveness

3)Self sufficiency

• Dimensions along with environment is associated with set of characteristics for classification of patients.

• A list of variables is used to describe the possible patient conditions relative to a specific characteristic.

• This distribution enables minimum and recommended number of nurses

Model based on FUZZY• FIS (Fuzzy Inference System)• Model in terms of Membership Function (MF)• Example: Body Temperature Characteristic

-The four initial fixed variables :

-Hypothermia

-Normal Temperature

-Hyperthermia

-Hyperpyrexia

Figure 1. Example of the body temperature characteristic as defined in the original MAP version (left) and in the new version based on FL (right).

• Define the rules for each FIS (Fuzzy Inference System)• Implementation of FISs(Fuzzy Inference Systems) using

MatLab .

RESULT

Figure 2. Input (clinical stability, responsiveness, self-sufficiency,environment) and the output (complexity) MFs of the final FIS.

Figure 3. Example of rule activation and patient classification in the finalFIS. In this rule the four input variables are connected among them with the AND operator.

CONCLUSION• Patient classification based on fuzzy logic enable nurses to

evaluate single patient in more efficient way.

• Optimization of clinical resource is possible.

THANK YOU!!