Abstract
In this paper, we present an Internet of things (IoT)-based data collection system for the risk assessment of urinary stone formation, or urolithiasis, by the measurement and storage of four parameters in urine: pH, concentrations of ionized calcium (Ca2+), uric acid and total dissolved solids. The measurements collected by the system from patients and healthy individuals grouped by age and gender will be stored in a cloud database. These will be used in the training phase of an artificial intelligence (AI) machine learning process utilizing the logistics regression model. The trained model provides a binary risk assessment, indicating if the end user is either a stone-former or not. For system validation, standard chemical solutions were used. Preliminary results indicated a sufficient measurement range, falling within the physiological range, and resolution for pH (2.0–10.0, +/−0.1), Ca2+(0.1–3.0 mmol/l, +/−0.05), uric acid (20–500 ppm, +/−1) and conductivity (1.0–40.0 mS/cm, +/−0.1), exhibiting high correlation with standard instruments. We intend to deploy this system in few hospitals in Taiwan to collect the data of patients’ urine, with analysis aided by urologist assessments for the risk of urolithiasis. The modularized design allows future modification and expansion to accommodate other sensing analytes.
Highlights
IntroductionThe occurrence rate of urolithiasis (i.e., the formation of urinary stones) is increasing in most countries worldwide [1]
In recent decades, the occurrence rate of urolithiasis is increasing in most countries worldwide [1]
Kidney stone formation remains asymptomatic until the final stages, when there is the presence of blood in the urine, painful urination or pain in the back or lower abdomen [6]
Summary
The occurrence rate of urolithiasis (i.e., the formation of urinary stones) is increasing in most countries worldwide [1]. Details of the prognosis and factors for urolithiasis progression, as well as comparisons between different methods of stone formation risk assessment based on urine composition, are presented in [1]. The presented system in this paper provides three essential roles—data collection, data storage, and binary risk assessment—organized in a centralized database and performed in the cloud Such data will be used in a machine learning training process to create the model for assessing urinary stone formation risk. The IoT-enabled design of the system allows for distributing multiple devices to different locations, collecting measurement results from multiple samples simultaneously and storing them in a central database This configuration makes the access to a large amount of data in a short period of time possible. This feature allows for the re-evaluation and improvement of the initial model in the future with a larger data set
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
