AGGREGATION TECHNIQUES FOR MEDICAL DIAGNOSIS
AGGREGATION TECHNIQUES FOR MEDICAL DIAGNOSIS
Information is combined using aggregation operators, which are statistical models. It can be used to combine N different data points into a single data point. Many disciplines are concerned with the problem of aggregating criterion functions to generate ultimate decision operations. The connections between the criteria involved are an important component in determining the form of such aggregation processes. The most popular aggregation operators are the ordered weighted average aggregation operators (OWA) & weighted OWA (WOWA).
Aggregation Techniques
OWA operators give a generalized family of mean type aggregation operators in fuzzy logic. R. Yager was the one who introduced them. This class includes many well-known mean operators like maximum (max), arithmetic mean (AM), minimum (min), & median. The reordering step is a crucial feature of the OWA operator; an aggregate ai is not linked to a specific weight wi, a weight is linked to a specific rearranged circumstance.
Where, σ is a permutation of that arranges them from largest (a1) to smallest (aN).
Weighted OWA (WOWA) operator satisfied the mean of weight. WOWA considered two weighting vectors (p & w) of dimension (n).
Here wi derived as follows,
Where w* represent the monotonic functions.
The Electroencephalogram (EEG) is a measurement of the total number of neurons that fire in different areas of the brain. It comprises data gathered from a collection of recording electrodes on changes in the brain’s electrical potential. Metabolic, biochemical, hormonal, neuro-electric, & behavioral factors have all been found to influence EEG patterns. Initially, the Encephalographer could essentially discern normal EEG activity from localized or widespread problems included within quite extensive EEG records. The epilepsy is the most significant activity that may be observed from of the EEG data. Epilepsy is characterized by uncontrollable extreme activity as well as potential discharge of the central nervous system in part or in whole. Several EEG waveform patterns define different types of epileptic seizures.
Epilepsy Risk Level Optimization using Aggregation Operators
In our example, the variables x1, x2, x3 are generated from clinical characteristics, while y1, y2, y3, y4 are obtained from epileptic EEG signals. Over the range of 0 to 1, these variables are normalized nonlinearly using Sigmoid & Secant functions. Consider the f(x) basic aggregating operator that we utilized in our process.
Where, x1, x2, x3 represents convolution index, the timing of the seizure & total fatigue of body. The variables y1, y2, y3, y4 indicated energy, peaks, events & sharp wave’s features of EEG. In well-randomized controlled trials (RCT), the factors are far less important & y is high significant. α1 represents the less important weight & α2 represents the great important weight. Consider the circumstance where α2 = 1- α1 & α1 is set several values, & the equations that result are as follows.
Even though in this example employing a recognized patient, the x factors in the given equations are assumed to be constants. In the aforementioned equations, the y variables are replaced with suitable recorded EEG modelling signal points.
Applications of Aggregation Techniques
Fuzzy logic-based automated systems are widely utilized in control systems, medical diagnosis, cluster analysis, home appliances, forecasting, decision-making systems, wireless sensor networks, manufacturing grid & the vehicle sectors. Membership Function fuzzification, rules, defuzzification, linguistic variables, & other terms are used in fuzzy logic.