Oral candidiasis or oral thrush is a disease of the oral cavity caused by infection with the fungus Candida albicans. Under normal conditions, this fungus is part of the mouth’s natural flora. However, when the microflora is imbalanced or the immune system is weakened, it can become pathogenic, leading to inflammation and white lesions on the oral mucosa. The disease is often mistaken for a common mouth ulcer, though it can indicate systemic disorders such as diabetes mellitus or a weakened immune system.
Addressing the issues of low public awareness and the lack of early diagnostic tools, a team of students from Universitas Gadjah Mada (UGM) has developed Canditect, an early-detection device for oral candidiasis based on the Internet of Things (IoT) and machine learning.
The innovation was created by the Student Creativity Program for Intellectual Creation (PKM-KC) team led by Naila Evelyna Difani Putri, with members Maria Benita Amarantha Elani (Dentistry), Irfan Satrio Wibowo (Internet Engineering Technology), Attahya Dzikra Hidayat (Internet Engineering Technology), and Aziz Nofikri (Mechanical Engineering).
Naila Evelyna Difani Putri explained that early detection of oral candidiasis is crucial, as many people are unaware that this fungal infection is more than a simple mouth sore.
Oral candidiasis can serve as an indicator of systemic diseases such as diabetes or HIV/AIDS.
“With a device like Canditect, people can conduct non-invasive self-checks using saliva, which helps increase awareness of oral health,” said Putri on Saturday, Oct. 18, 2025.
Putri further explained that the Canditect prototype uses three main parameters, pH, glucose levels, and electrochemical activity, to detect Candida albicans in saliva.
The device integrates a Screen-Printed Carbon Electrode (SPCE) sensor with pH and glucose sensors, and all data are processed through an IoT system and a random forest machine learning algorithm for automatic analysis.
Unlike laboratory methods such as microbiological culture or molecular testing, which are time-consuming and costly, Canditect offers a faster, more practical, and efficient alternative suitable for both clinical and field use.
The device is designed to detect early signs of infection in real time while providing recommendations for medical professionals to take further action.
During development, the team encountered several challenges, particularly in integrating sensitive biological detection systems with IoT technology.
“We initially focused our research on the enzyme Secreted Aspartyl Proteinase (SAP) because of its important role in Candida virulence. However, after further study, we found that the enzyme’s detection sensitivity was too low, so we shifted to more stable parameters, pH, glucose, and SPCE electrode response,” Putri added.
The Canditect team continues to refine the prototype and detection algorithm.
Saliva sampling tests are scheduled for mid-October 2025 at Korpagama UGM, focusing on diabetic respondents who are at higher risk of oral candidiasis.
For Putri and her team, interdisciplinary collaboration has been the key to the success of this research.
The development of Canditect required not only biomedical knowledge but also expertise in electronics and programming.
“We’ve learned that health innovation isn’t just about the result, but about understanding people’s needs and turning them into real solutions,” she said.
In the future, the team hopes Canditect can be further developed into an accessible, accurate, and affordable early-detection device for use in healthcare facilities and dental clinics.
“We envision Canditect as the first step toward an IoT-based oral disease monitoring system in the future,” concluded Putri.
Author: Kezia Dwina Nathania
Editor: Gusti Grehenson
Post-editor: Lintang Andwyna
Photographs: Canditect Team