Using AI to Differentiate Between Iron Deficiency Anemia and Aplastic Anemia
Author Information
Author(s): Darshan B. S. Dhruva, Sampathila Niranjana, Bairy G. Muralidhar, Prabhu Srikanth, Belurkar Sushma, Chadaga Krishnaraj, Nandish S.
Primary Institution: Manipal Academy of Higher Education, Manipal
Hypothesis
Can machine learning and explainable AI effectively differentiate between iron deficiency anemia and aplastic anemia using blood test attributes?
Conclusion
The study successfully developed a machine learning model that accurately distinguishes between iron deficiency anemia and aplastic anemia, achieving an accuracy of 96%.
Supporting Evidence
- The model achieved an accuracy of 96% using a stacked ensemble approach.
- Five explainable AI techniques were employed to enhance model interpretability.
- The dataset was ethically sourced from Kasturba Medical College, Manipal.
Takeaway
This study used computer programs to help doctors tell the difference between two types of anemia by looking at blood tests.
Methodology
The study utilized machine learning algorithms and five explainable AI techniques to analyze a dataset of blood test attributes from 500 patients.
Potential Biases
Potential biases in the dataset and the need for careful interpretation of model predictions.
Limitations
The model's performance is dependent on the quality of the dataset, and extensive testing is needed before clinical implementation.
Participant Demographics
The dataset included 500 patients, with 266 diagnosed with iron deficiency anemia and 234 with aplastic anemia.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website