An Efficient, Cost-effective and Reliable Non-invasive Anaemia Detection Method by Analysing Palm Pallor
Sumana Naskar, Abhishek Kesarwani, Sunanda Das, Mamata Dalui · 2024
Anaemia, resulting due to paucity of haemoglobin concentration in blood, has become a severe public health issue, with higher prevalence in developing countries like India. Traditional screening test for anaemia is invasive which involves expert technicians for collecting venous blood from the patient and a well-equipped diagnostic laboratory for performing the tests thereby involving cost-prohibitive affairs. This essentially demands a cost-effective, non-invasive, and reliable solution for anaemia detection which can easily be administered in mass-scale. Hence, this work proposes a smartphone-based robust, non-invasive haemoglobin estimation method using a deep Convolutional Neural Network (CNN) model. The proposed approach predicts haemoglobin accurately by analysing the video of palm colour changes. The proposed solution beats the state-of-the-art schemes in performance by ensuring a mean RMSE value of 0.56 g/dL and classification accuracy of 88.64%.