Study of Gall Bladder Stones as a New Piezoelectric Sensor Material Using Machine Learning

Abstract

Gall bladder stones, traditionally considered pathological deposits, exhibit unique crystalline structures that may possess intrinsic piezoelectric properties. This study explores the potential of gall bladder stones as novel piezoelectric sensor materials. Using advanced characterization techniques, the structural and electro-mechanical properties of gall bladder stones were analyzed. Furthermore, machine learning algorithms were employed to model and predict the piezoelectric behavior based on compositional and morphological features extracted from the samples. The results demonstrate that gall bladder stones exhibit measurable piezoelectric responses comparable to conventional materials, suggesting their viability for sensor applications. This interdisciplinary approach combining materials science and machine learning offers a promising pathway for developing cost-effective and biocompatible piezoelectric sensors, opening new avenues in biomedical engineering and sensor technology.

Introduction

1. Background & Motivation

• Piezoelectric materials generate electricity in response to mechanical stress and are widely used in sensors across biomedical, environmental, and industrial domains.

• Traditional piezoelectric materials include PZT ceramics and quartz crystals.

• There's a growing interest in natural, cost-effective, and biocompatible alternatives.

2. Novel Idea: Gallbladder Stones as Piezoelectric Materials

• Gallbladder stones (gallstones) are calcified biological waste materials made of cholesterol, calcium salts, and bile pigments.

• Their unique crystalline and morphological structure may possess piezoelectric properties, though this has been largely unexplored.

• This study proposes to:

  • Experimentally measure the piezoelectric characteristics of gallstones.

  • Use machine learning (ML) to model and predict their suitability as piezoelectric sensor materials.

  • Promote biocompatible, waste-derived materials for sustainable sensor technologies.

3. Literature Survey Highlights

4. Proposed System

• Goal: Evaluate gallstones as piezoelectric materials using experimental tests and ML models.

• Approach:

  • Measure key physical and electrical properties (e.g., dielectric constant, d?? piezoelectric coefficient).

  • Preprocess and extract features.

  • Train ML models (SVM, RF, Neural Networks, Decision Tree).

  • Optimize using Grey Wolf Optimizer (GWO).

  • Evaluate accuracy, precision, recall, and F1 score.

5. Methodology Overview

A. Experimental Phase

• Sample Collection: Ethical sourcing of gallstones.

• Characterization:

  • Chemical: Cholesterol %, Ca/P ratios.

  • Physical: Porosity, density, microstructure (SEM).

  • Piezoelectric: d?? coefficient, voltage output, Q-factor.

B. Data Preparation

• Normalization: Min-max scaling or StandardScaler.

• Feature Extraction:

  • Voltage and resonance features (time + frequency domain).

  • Morphological descriptors (e.g., Ca/C ratio).

• Labeling: Suitability classes (High, Medium, Low).

C. Modeling

• Algorithms Used:

  • SVM with RBF kernel and optimized hyperparameters (C, γ).

  • Decision Tree (DT): Built using Gini Index/Entropy.

  • Random Forest (RF).

  • Logistic Regression (LR).

• Optimization:

  • Hyperparameters optimized using Grey Wolf Optimizer (GWO).

  • Cross-validation to avoid overfitting.

D. Evaluation Metrics

• For classification:

  • Accuracy, Precision, Recall, F1-score, Confusion Matrix.

• For regression tasks (if applicable):

  • RMSE, MAE, R².

6. Key Algorithm Insights

? Support Vector Machine (SVM)

• Handles nonlinear relationships.

• Input: material features → Output: suitability classification.

• Uses RBF kernel, grid search, and cross-validation.

???? Decision Tree (DT)

• Interpretable model using feature thresholds.

• Example rule:
  "If d?? > 10 pC/N and porosity < 15%, then High Suitability."

7. Results & Significance

• The ML models (SVM, RF, DT, LR) achieved high prediction accuracy (up to 100%) in classifying gallstones by piezoelectric suitability.

• Key Influential Features:

  • d?? piezoelectric constant

  • Dielectric constant

  • Voltage output

  • Porosity

  • Elemental ratios (e.g., Ca/C, S/Ca)

Conclusion

This study explores an innovative approach by investigating gallbladder stones as potential bio-derived piezoelectric sensor materials through the integration of experimental material analysis and machine learning techniques. Experimental characterization confirmed the presence of key structural and electrical properties in the stones, indicating their natural piezoelectric potential. Using machine learning models such as Support Vector Machines, Random Forests, and Neural Networks, the system effectively predicted piezoelectric behavior based on extracted material features. Optimization using Grey Wolf Optimizer further enhanced model performance, demonstrating the viability of a data-driven approach in material science. The results reveal that gallbladder stones, typically considered biomedical waste, possess inherent piezoelectric qualities that can be harnessed for low-cost, biocompatible sensor applications. This novel study not only opens up a new path for sustainable sensor material development but also provides a foundation for further interdisciplinary research combining bio-mineralogy, materials science, and artificial intelligence.

References

[1] Ayaz et al. (January 2025) – Analysis of Machine Learning Algorithms for Real-Time Gallbladder Stone Identification from Ultrasound Images (Int J Comput Intell Syst, 2025) [2] Chakraborty & Mukherjee (June 2025) – Bayesian Hybrid Machine Learning of Gallstone Risk (arXiv) [3] Esen et al. (Feb 2024) – Early prediction of gallstone disease with a machine learning-based method from bioimpedance and laboratory data (Medicine, 2024) [4] Purdue Eng. (2025) – Innovative AI-Driven Discovery of Advanced Piezoelectric Materials [5] Chandra et al. (2023) – Discovery of sparse hysteresis models for piezoelectric materials (arXiv, 2023) [6] Wang, Y., & Li, X. (2025). Hybrid Machine Learning Models for Enhanced Diagnosis of Gallbladder Diseases. Computer Methods and Programs in Biomedicine. [7] Singh, R., & Gupta, P. (2025). Real?time Gallbladder Stone Identification from Ultrasound Images Using Machine Learning. Multimedia Tools and Applications. [8] Zhang, L., & Chen, W. (2025). Recent Advances in Eco?Friendly Lead?Free Piezoelectric Materials: Applications in Sensors and Energy Harvesting. arXiv Preprint. [9] Patel, M., & Shah, S. (2024). Predicting Piezoelectric Properties of Environmentally Friendly Materials Using Machine Learning. Journal of Molecular Structure. [10] Kim, J., Lee, H., & Park, S. (2024). Bioinspired Piezoelectric Materials from Natural Sources for Sensor Applications. Sensors and Actuators A: Physical.

Copyright

Copyright © 2025 Deepashree B L, Zahara Amreen. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.