Particle Swarm-Optimized Artificial Neural Network for Non-Invasive Glucose Measurement and HbA1c Computation

In this work, a particle-swarm optimization-based artificial neural network for non-invasive continuous glucose monitoring using the principles of near-infrared spectroscopy (NIRS) is proposed. It is shown that the PSO-ANN approach outperforms the traditional backpropagation algorithm used in ANN training and several other regression algorithms with the lowest error metrics: MAE- 1.01, MSE-2.16, RMSE-0.97, 𝑅-squared-0.976 and modified 𝑅-squared-0.973. The 3-stage methodology adopted in this work is shown below.

The accuracy and reliability of the proposed system are analysed using the Clarke Error Grid (CEG) with 93.9% of the obtained readings falling within zone A and 100% of the readings falling in the clinically accepted range (zones A and B). Refer to the preprint for more details.

Data Format

Inputs: BMI - np.array(b1, b2, b3,.....bn), Voltage (in mV) - np.array(v1, v2, v3,......vn), Age - np.array(a1, a2, a3,......an), Output: glucose in mg/dl

Run pso_ann.py after modifying Xtrain and PSO parameters.

Please consider citing the work if you find it useful.

@article{Particle Swarm-Optimized Artificial Neural Network for Non-Invasive Glucose Measurement and HbA1c Computation,
  author = {Suma KV, Dharini Raghavan, Maya V Karki, Narayana Sharma and Gundu Rao},
  doi = {10.36227/techrxiv.24465955.v1},
  journal = {Techrxiv},
  pages = {1--4},
  title = {{Particle Swarm-Optimized Artificial Neural Network for Non-Invasive Glucose Measurement and HbA1c Computation}},
  year = {2023}
}