Approaching Explainable Artificial Intelligence Methods in the Diagnosis of Iron Deficiency Anemia Using Blood Parameters

Anemia is a global health disorder diagnosed by observing blood parameters. It is a tedious and time-consuming method for healthcare workers to analyze the data manually and may also lead to mistakes. This paper proposes a novel method to understand the impact of blood parameters in diagnosing anemia. Machine learning methods have been used to classify the data, and the impact of the attributes was explained using explainable AI tools to bring transparency and trust to the architectures. XAI helps in ensuring fairness, accountability, and transparency. The models show a high accuracy of 80-100
^%
• The beeswarm plot explained the impact of the various attributes present in a complete blood count in the diagnosis of iron deficiency anemia. The methods introduced help in the quick diagnosis of anemia and save time for healthcare professionals. Improvement in the current technology in collaboration with healthcare workers will lead the medical domain to new heights.