Development of Altera NIOS II Soft-core system to predict total Hemoglobin using Multivariate Analysis

In Today’s world, Hemoglobin is measured using an invasive method. This method leads to delayed diagnosis, painful experiences for patients, and a lot of biomedical waste. To overcome these problems, an Altera NIOS II soft-core based system was built to monitor hemoglobin non-invasively. The heart o...

Full description

Saved in:
Bibliographic Details
Published in:Journal of physics. Conference series 2021-05, Vol.1921 (1), p.12039
Main Authors: Pinto, C F, Parab, J S, Sequeira, M D, Naik, G M
Format: Article
Language:English
Subjects:
Accuracy
Anemia
Embedded system
Empirical analysis
FPGA
Hemoglobin
Microprocessors
Model accuracy
Multivariate
Multivariate analysis
NIOS II
Physics
PLSR
Root-mean-square errors
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In Today’s world, Hemoglobin is measured using an invasive method. This method leads to delayed diagnosis, painful experiences for patients, and a lot of biomedical waste. To overcome these problems, an Altera NIOS II soft-core based system was built to monitor hemoglobin non-invasively. The heart of the system is NIOS II soft-core processor which was configured on the DE0 Nano FPGA board having Cyclone IV EP4CE22F17C6N. This system also has a finger probe which consists of five LED sources (670 nm, 770 nm, 810nm, 850nm and 950nm) and a photodetector (OPT101) to acquire the signal using photoplethysmography (PPG). The incoming real-time PPG signal is recorded at five different wavelengths for fifteen individual subjects. Before applying Multivariate Partial Least Square Regression (PLSR), mathematical empirical formulas was used to predict hemoglobin which gave Root Mean Square Error (RMSE) of 0.442 g/dL and the prediction accuracy of 97.05%. To further improve the system accuracy, the PLSR model was implemented on the NIOS II soft-core system. With this, the hemoglobin was predicted with an accuracy of 98.98% and a RMSE of 0.179 g/dL. The designed system was validated with Bland-Altman analysis which shows good agreement between predicted and reference hemoglobin
ISSN:1742-6588
1742-6596
DOI:10.1088/1742-6596/1921/1/012039