Mobile application prototyping using Artificial Intelligence to support childhood tuberculosis diagnosis

Area of Science:

• Medical Informatics
• Artificial Intelligence in Healthcare
• Pediatric Infectious Diseases

Background:

• Pulmonary tuberculosis (TB) diagnosis in children presents unique challenges.
• There is a need for innovative tools to support early and accurate diagnosis.
• Existing diagnostic software for pediatric TB is limited.

Purpose of the Study:

• To develop a mobile application prototype named TB Kids.
• To leverage Artificial Intelligence (AI) for predicting and supporting the diagnosis of pulmonary tuberculosis in children.
• To address the gap in AI-driven diagnostic software for pediatric TB.

Main Methods:

• Technological development research utilizing a prototyping approach.
• Adherence to the Rational Unified Process (RUP) model with four distinct stages: conception, elaboration, construction, and transition.
• Development timeline: November 2022 to July 2023.

Main Results:

• The TB Kids prototype integrates features for comprehensive patient assessment, including risk and nutritional evaluation.
• Includes functionalities for tuberculin skin testing, antibiotic therapy, and contact tracing.
• Features AI-driven interpretation of chest X-rays with risk graphing and decision support, alongside clinical guidance and recording.

Conclusions:

• The high-fidelity TB Kids mobile application prototype offers an innovative solution for diagnosing pediatric tuberculosis.
• The application demonstrates a consistent and creative interface, aligning with Sustainable Development Goal 3.
• Addresses the critical need for AI-based prediction software in the diagnosis of children at risk for tuberculosis.