MDN Projects
← Back to Projects
Healthcare & Artificial Intelligence / Skin Burn Check - AI Burn Severity Assessment
Case Study • Healthcare & Artificial Intelligence

Skin Burn Check - AI Burn Severity Assessment

AI-assisted Android application for preliminary burn severity assessment from images.

My Role
Android Developer / AI Mobile Developer
Domain
Healthcare & Artificial Intelligence
Status
● Production Ready
Contracting
MDN Projects
Inquire for Similar Project
CNN-based
AI Model
Android
Platform
Burn Images
Input
AI-assisted Preliminary Classification
Assessment
Skin Burn Check - AI Burn Severity Assessment preview 1

01 / Overview

Skin Burn Check is an Android-based application designed to help identify the severity of burn injuries through image analysis. The application uses Artificial Intelligence with a Convolutional Neural Network (CNN) model to analyze burn images and provide an initial assessment of the condition. Developed natively using Kotlin, the application combines mobile technology and AI to provide an accessible and practical burn assessment tool. Skin Burn Check is intended to support early identification and provide users with useful preliminary information about burn conditions directly through an Android device.

02 / Problem & Solution

The Challenge

Initial burn assessment can be difficult for users without medical expertise, particularly when visual differences between burn conditions are not easy to interpret. The project aimed to explore how image-based AI could provide an accessible preliminary assessment tool while presenting the result in a practical mobile experience.

The Engineering Solution

Developed a native Android application using Kotlin that analyzes burn images with a Convolutional Neural Network model. Implemented an image-based assessment workflow that allows users to provide a burn image and receive an AI-generated preliminary classification. Designed the application to present the analysis result in an accessible format while positioning it as an initial assessment aid rather than a replacement for professional medical evaluation.

03 / Architecture & Implementation

System Architecture Design:
Native Android application built with Kotlin, integrating an image-processing pipeline and Convolutional Neural Network model for burn image classification.

04 / My Responsibilities

  • ▹ Developed the native Android application using Kotlin
  • ▹ Implemented burn image capture and image input workflows
  • ▹ Integrated the CNN-based image classification model
  • ▹ Implemented image preprocessing and inference flow
  • ▹ Developed the burn assessment result interface
  • ▹ Handled Android UI implementation, testing, debugging, and application optimization.

05 / Challenges & Business Outcomes

Technical Obstacles

Processing user-provided burn images consistently across different image conditions; Integrating CNN inference into a mobile application workflow; Designing an understandable result interface for non-technical users; Managing model performance and mobile resource constraints; Presenting AI results responsibly as preliminary information rather than a clinical diagnosis.

Measurable Impact

Delivered a practical Android prototype for image-based burn assessment using AI. The application demonstrates how CNN-based image classification can be integrated into a mobile workflow to support early identification and provide accessible preliminary information.