The data is obtained from Physical Characteristics of Urines With and Without Crystals , a chapter from Springer Series in Statistics. The 79 urine specimens were analyzed in an effort to determine if certain physical characteristics of the urine might be related to the formation of calcium oxalate crystals. The six physical characteristics of the urine are: (1) specific gravity, the density of the urine relative to water; (2) pH, the negative logarithm of the hydrogen ion; (3) osmolarity (mOsm), a unit used in biology and medicine but not in physical chemistry; osmolarity is proportional to the concentration of molecules in solution; (4) conductivity (milli-Mho), conductivity is proportional to the concentration of charged ions in solution; (5) urea concentration in millimoles per litre; and (6) calcium concentration in millimoles per litre.
Kidney stones are a common health problem that affects millions of people worldwide. Early detection and timely intervention can help prevent complications and improve treatment outcomes. However, traditional methods for diagnosing kidney stones can be invasive and time-consuming. Therefore, the aim of this project is to develop a machine learning and deep learning model that can predict the presence of kidney stones in patients based on their medical history, lab reports, and other relevant factors. The model has been trained on a large dataset of patient data using advanced techniques such as Convolutional Neural Networks (CNNs) and Long Short-Term Memory (LSTM) networks to achieve high accuracy. The model has been deployed on the web using Streamlit Cloud, a powerful platform for building and deploying data applications. This will allow patients and healthcare providers to easily access the model and get quick and accurate predictions about the presence of kidney stones in patients. The ultimate goal of this project is to improve the speed and accuracy of kidney stone diagnosis, leading to early precautions and diagnosis and reduced healthcare costs.
- Data cleansing and exploratory data analysis (EDA)
- Correlation heatmap to derive correlation amongst the features
- Using algorithms such as Logistic Regression, Decision Tree, Random Forest, Support Vector Classifier, Extreme Gradient Boosting Classifier
- Using Convolutional Neural Networks (CNN) and Bi-Long Short Time Memory (Bi-LSTM) models for deep learning based prediction
- Evaluating the models based on the performace metrics such as accuracy, precision value, recall value, f1-score
- Deploying the model on Streamlit Cloud
** Click here to run the application live on server