Borel GOUDJOU

As a biomedical engineer specializing in AI for medical imaging, I designed and trained deep learning models (U‑Net, diffusion, attention‑based architectures) to denoise clinical images and enhance diagnostic reliability. I built a full training pipeline—from DICOM extraction and normalization to augmentation and preprocessing—ensuring data integrity and reproducibility. I developed a C++ inference pipeline using LibTorch, integrated into an existing medical application for real‑time GPU execution and optimized memory management. The system was validated on clinical datasets, with performance evaluation and regulatory documentation focused on traceability and reproducibility.

As an intern, I explored how cutting-edge technologies can enhance athletic performance and prevent injuries. Using the Kdeltas platform by Kinvent, I measured forces and assessed balance to evaluate both static and dynamic postures. To gain deeper insights, I integrated pose estimation and computer vision techniques, allowing me to analyze athletes’ movement patterns and postural stability in detail. By combining these approaches, I identified key biomechanical factors that influence performance and highlighted potential injury risks, contributing to smarter, data-driven strategies for athlete training and optimization.

As an Application Engineer, I designed and implemented complex medical systems, including dialysis centers, oxygen supply networks, RX scanners, and remotely controlled radiology units, ensuring each installation was perfectly sized for operational efficiency. I managed system maintenance using CMMS software, keeping equipment reliable and compliant. I also developed and delivered training programs for clinical staff, helping them use the devices safely and effectively. By optimizing logistics and consumables for dialysis services, I improved service continuity and reduced operational costs, collaborating closely with multidisciplinary teams to integrate technology seamlessly into clinical workflows.

As an Embedded Systems Engineer at CACES, I designed and implemented solutions in C/C++ for ARM Cortex-M processors (STM32, RENESAS) using IAR and VS Code. I programmed low-level protocols like UART, SPI, and I2C, and developed libraries for GPS and GSM systems, enabling reliable communication and device control. I worked with real-time operating systems (FreeRTOS, TI-RTOS, CrossWorks) to ensure precise and efficient embedded performance. By following Agile and DevOps practices and emphasizing thorough documentation with Doxygen, I delivered robust, maintainable systems while strengthening my expertise in hardware-software integration and real-time system optimization.

As a Research & Development Intern at TonkaIn Laboratory, I contributed to innovative biomedical projects by combining signal processing, embedded programming, and electronics design. I applied signal conditioning and digital extraction techniques to oscillometric signals using MATLAB and Excel, and developed C++ algorithms for blood pressure and pulse calculation on a CC2650 microcontroller. I validated results through rigorous statistical analysis and benchmarking against existing databases, while also supporting circuit design and mentoring fellow interns. Throughout this work, I strictly followed ISO 14971 for risk management and IEC 62304 for medical device software development, ensuring compliance, safety, and robustness. This experience allowed me to bridge biomedical engineering and software development, enhancing my expertise in biomedical signal processing, embedded systems, and regulated medical device development.