<img height="1" width="1" style="display:none" src="https://www.facebook.com/tr?id=504731893395981&amp;ev=PageView&amp;noscript=1">

Firmware for a Custom Skin Cancer Screening Device

Detects malignant skin growth without a biopsy

  • C++
  • Linux
Solution FPGA-based LED-enabled device providing accurate assessment of suspect skin growth
Industry
  • Healthcare
  • Software and Technology
ENGAGEMENT MODEL

Fixed Price

METHODOLOGY Waterfall
Team
  • Firmware Developers
  • QA Lead
  • Business Analyst
  • QA Engineer
  • Project Manager
1
2
3
4
5
6

Customer

Case Highlights

Softeq participated in the development of a complex Wi-Fi-enabled system equipped with powerful functionality.  

  • Incorporates a FPGA-powered handheld device based on the Qualcomm Snapdragon SD410 processor
  • Is driven by embedded software
  • Includes hardware components such as an audio speaker, accelerometer, and display
  • Captures LED light signal variations and translates them into a visual diagram in real time
  • Monitors the vascular refilling rate and proceeds with a custom algorithmic data analysis

Problem

Veriskin decided to build a novel device that uses Machine Learning algorithms for non-invasive skin cancer detection.

They turned to Softeq to test the viability of their idea through a proof of concept (PoC) and develop firmware components powering the device.

Solution

Overview

The solution is a complex Wi-Fi-enabled system that incorporates:

  • A FPGA-powered handheld device based on the Qualcomm Snapdragon SD410 processor complete with LEDs
  • Embedded software
  • Hardware components: audio speaker, accelerometer, and display (among others)

Equipped with 2 photodetectors and a force sensor, the solution captures LED light signal variations and translates them into a visual diagram in real time. 

To diagnose the suspect lesion’s condition, the system monitors the vascular refilling rate and proceeds with a custom algorithmic data analysis complete with a plain report output on the UI.

Firmware and Software Functionality

  • Audiovisual guide with error messaging
  • Calibrated LED lights maximizing photodetector sensitivity
  • Memory capacity for 100+ test results downloadable via USB or BLE
  • Custom Linux BSP and drivers
  • Built-in flash memory-based software
  • Optimized Linux boot-time
  • Firmware updates via the USB
  • Secure login (SSH protocol)

Results

Project Outcomes

The resulting solution was expected to participate in rigorous clinical testing across US dermatological institutions. The system showed the capability of detecting malignant skin growth without extra biopsies or examinations. The device was meant to cut healthcare-associated costs by providing a low-cost and prompt cancer diagnostics method.