Impact of AI in Healthcare Services: Analysis Using Medical Synthetic Data

Artificial Intelligence has effectively spread in the field of healthcare in recent years. It is leading to substantial changes and transforming the traditional healthcare system into much more advanced technology. For more improvement in AI healthcare systems, comprehensive data analysis is required. This paper focuses on deeply analyzing a synthetic AI healthcare dataset utilizing statistical approaches. Statistical approaches are significantly used to predict the efficiency of different models, accuracy of AI diagnosis, patient satisfaction, and other correlations between relevant variables of an AI healthcare system. This paper provides an in-depth analysis of an AI-driven healthcare dataset utilizing pearson's correlation coefficient, multiple linear regression plot, and ANOVA testing. Using different techniques, it can be stated that there is no relation between AI diagnosis with patient's treatment duration and recovery time as the p-value (0.771963) was greater than the common significance level (0.05). The hypothesis, where it was assumed that there was not a statistically major difference between AI diagnostic confidence and human diagnosis, has been proved as the p-value (0.248596) was greater than the common significance level (0.05). To conclude, it can be said that the assumptions were correct as the null hypotheses are accepted by testing them in several statistical approaches.