Zobraziť v EDS

Adaptive neuro fuzzy inference system approach for medical decision making.

Uložené v:
Podrobná bibliografia
Názov: Adaptive neuro fuzzy inference system approach for medical decision making.
Autori: Birudu, Sujatha1 (AUTHOR) birudusujatha@gmail.com, Madhu, Shrija1 (AUTHOR) shrijamadhu.76@gmail.com, Nagu, Chandra Sekhar Reddy2 (AUTHOR) naguchinni@gmail.com, Ramachandran, Tamilkodi1 (AUTHOR) rtamilkodi1978@gmail.com, Mishra, Nirmith Kumar3 (AUTHOR) nirmithmishra@gmail.com
Zdroj: AIP Conference Proceedings. 2025, Vol. 3342 Issue 1, p1-7. 7p.
Predmety: *MEDICAL decision making, *GESTATIONAL diabetes, *MEDICAL records, *PREDICTION models, *FUZZY logic, *SENSITIVITY & specificity (Statistics), *EARLY diagnosis, *ARTIFICIAL intelligence
Abstrakt: The employment of computers for diagnosis has been made possible by developments in the fields of artificial intelligence and medicine. It is now crucial to lower the death rate by early disease detection. One of the main areas of information mining from clinical datasets is the development of a decision–making model utilizing artificial intelligence approaches. Using adaptive neurofuzzy inference system from clinical datasets, a medical decision–making model has been constructed in this study work. Five clinical datasets were acquired in order to evaluate the proposed model. It has been established in this contribution to diagnose gestational diabetes mellitus in the presence or absence. The gestation period refers to the time a woman spends pregnant. One kind of diabetes that affects pregnant women in the first trimester of their pregnancy is called gestational diabetes mellitus. The amount of fuzzified input characteristics is used to estimate the number of neurons in the layer, which was developed using an adaptive neurofuzzy inference method. The Gaussian membership function is used to fuzzified the input features. The nodes are these input features that have been fuzzified. The precise interpolation algorithm is used to estimate the weights between the hidden and output layers. The fuzzy rules are pruned using the weights that were acquired. The knowledge base contains these regulations. The testing set's samples are categorized by the adaptive neural fuzzy inference system using the knowledge base's rules. With 78% sensitivity and 95% specificity, the established model for gestational diabetes mellitus achieves an overall accuracy of 89%. The sensitivity, specificity and overall accuracy have improved. [ABSTRACT FROM AUTHOR]
Databáza: Academic Search Index
Popis
Abstrakt:The employment of computers for diagnosis has been made possible by developments in the fields of artificial intelligence and medicine. It is now crucial to lower the death rate by early disease detection. One of the main areas of information mining from clinical datasets is the development of a decision–making model utilizing artificial intelligence approaches. Using adaptive neurofuzzy inference system from clinical datasets, a medical decision–making model has been constructed in this study work. Five clinical datasets were acquired in order to evaluate the proposed model. It has been established in this contribution to diagnose gestational diabetes mellitus in the presence or absence. The gestation period refers to the time a woman spends pregnant. One kind of diabetes that affects pregnant women in the first trimester of their pregnancy is called gestational diabetes mellitus. The amount of fuzzified input characteristics is used to estimate the number of neurons in the layer, which was developed using an adaptive neurofuzzy inference method. The Gaussian membership function is used to fuzzified the input features. The nodes are these input features that have been fuzzified. The precise interpolation algorithm is used to estimate the weights between the hidden and output layers. The fuzzy rules are pruned using the weights that were acquired. The knowledge base contains these regulations. The testing set's samples are categorized by the adaptive neural fuzzy inference system using the knowledge base's rules. With 78% sensitivity and 95% specificity, the established model for gestational diabetes mellitus achieves an overall accuracy of 89%. The sensitivity, specificity and overall accuracy have improved. [ABSTRACT FROM AUTHOR]
ISSN:0094243X
DOI:10.1063/5.0296264