Mobile app for skin mole analysis

Our client, a medical technology company, was building a product that lets users monitor their skin moles on their own, in order to catch changes that may indicate a risk of melanoma early. This called for three tightly synchronised layers: a backend storing the photo history and analysis results, a web application for consulting physicians, and a mobile app running an image analysis engine locally on the patient's device.

We built the complete platform: a Symfony backend managing accounts, photo history and communication with doctors; a web app for the medical team to review cases and issue consultations; and an Android app that uses OpenCV to detect the mole area, calibrate its size against a reference marker and automatically analyse its parameters according to the ABCDE rule.

At the heart of the app is a dedicated image analysis engine built on OpenCV, running entirely on the patient's device — no photos are ever sent to third-party ML services. The pipeline has four stages:

1. Size calibration — the patient photographs the mole next to a reference marker of known dimensions. The algorithm detects the marker, derives the px → mm scale and rejects shots taken from an angle that could distort the measurement.
2. Mole area detection — contour segmentation based on thresholding in the HSV colour space and morphological operations, filtering out artefacts such as hair, shadows and glare.
3. ABCDE parameter analysis — automatic calculation of the five parameters used in clinical dermatology: Asymmetry, Border (edge regularity), Colour (colour variation within the mole), Diameter (in mm) and Evolution (comparison against earlier measurements of the same mole over time).
4. Change timeline — every mole has its own measurement history; the app visualises the trend over time and raises an alert when parameters cross clinical thresholds.

The analysis serves as a monitoring and screening tool — the result does not replace a dermatological consultation, but it makes it easier for both the user and the consulting physician to track changes objectively.

The app is available on Google Play and has been under continuous development and maintenance since 2020. It offers real help in monitoring moles — detecting changes early is fundamental to melanoma prevention.