Full-Cycle Software Development for Smart Bassinette: Intelligent Crying Recognition
With skin cancer being the most common form across the US, the customer decided to build a novel device capable of non-invasive disease detection early on, with no biopsy and costly treatment at later stages. In 2016, the client’s startup won a Small Business Innovation Research and Development grant awarded by the US National Cancer Institute.
The customer was looking to obtain a viable PoC solution within a short timeline. They opted to contract several independent Europe- and US-based teams so as to streamline the engineering process.
Boasting a solid full-cycle engineering expertise, Softeq stood out for its extensive portfolio of relevant healthcare projects, including a complete IoT ecosystem for a smart cradle and a doctor-patient communication mobile and web app. During the selection process, Softeq demonstrated exceptional teamwork providing instant response beyond regular hours. Complete with reasonable rates, those were the decisive advantages for the customer.
Designed for the US medical institutions, the solution is intended to help general practice clinicians and nurse practitioners conduct non-invasive early-stage skin cancer detection and direct patients for biopsies or escalation of care.
The solution is a complex Wi-Fi-enabled system capable of detecting cancerous skin growth in 2 minutes by taking a simple scan. The system comprises an FPGA-powered handheld device based on the Qualcomm Snapdragon SD410 processor complete with LEDs, embedded software, and daemon modules — audio speaker, accelerometer, and button handler, among others.
Equipped with 2 photo detectors and a force sensor, the solution captures the LED light signal variations and translates them into a visual diagram in real time. To diagnose the suspect lesion’s condition, the system monitors vascular refilling rate and proceeds to the custom algorithmic data analysis complete with a plain report output on the UI.
Softeq developed the core firmware components and integrated them into the modular ecosystem on the customer side.
The solution’s highlights:
The SSH protocol utilities enable an encrypted and secure access to the solution.
Delivered exactly within the budget, the solution met the customer’s expectations. Currently, the system is under rigorous clinical testing across the US dermatological institutions.
The studies demonstrate that the device proves efficient in determining cancerous lesions. The solution is capable to definitively tell benign and malignant skin growth apart, detecting the disease in 96.7% of suspect cases with no extra biopsies or examinations. Allowing for a low-cost and prompt cancer diagnostics method, the device is meant to cut health care associated spendings.