Mahdi Pirayesh Shirazi Nejad

Motivated PhD candidate in Biomedical Engineering with 3 years of experience in ML-driven biomedical signal processing, particularly in denoising severely corrupted CPR signals in collaboration with Defibtech. Strong expertise in deep learning, real-time system development, and biosignal analysis. Published researcher with contributions to ECG classification, artifact removal, and generative modeling. Seeking an internship in AI for healthcare, biomedical data science, or ML research to apply expertise in signal processing and AI-based classification models.

Defibtech, Thesis Project, 2021 - Present, Developed ML algorithms for denoising severely corrupted CPR signals, improving automated defibrillator decision-making., Optimized real-time signal processing algorithms for enhanced cardiac event detection., Applied deep learning techniques (CNNs, Autoencoders) for ECG signal enhancement., Collaborated with clinicians & engineers to enhance shock advisory algorithms using imbalanced CNNED models.

[Your LinkedIn Profile Here]

[Your GitHub Profile Here]

• $1,000 seed funding at Innovate Health Pitch Fest for a glucose monitoring device startup (HealthStream).
• $1,000 grant from UConn Get Seeded Program for a glucose monitoring startup (HealthStream).
• Ranked 342nd among 200,000+ participants in Iran's Nationwide University Entrance Exam (Konkour, 2015).
• Member of the National Organization for Development of Exceptional Talents (NODET).

• Enhancing the Accuracy of Shock Advisory Algorithms in AEDs Using CNNs – Computers in Biology and Medicine.
• ECG Classification via Adaptive Beat Segmentation & Deep Learning – Computers in Biology and Medicine.
• Noise Reduction in Photoplethysmography Signals Using Autoencoders – IEEE Transactions on Biomedical Engineering.
• Towards Continuous Skin Sympathetic Nerve Activity Monitoring – Body Sensor Networks.
• Development of an Electrode for Electrodermal Activity Underwater – Body Sensor Networks.

Cardiopulmonary Resuscitation, Treatment, and Analysis (US Patent App #18659762).

• Graduate Researcher, University of Connecticut, Developed Generative Adversarial Networks (GANs) for ECG Signal Generation & Denoising, enhancing cardiac monitoring through synthetic data and noise reduction., Designed ECG classification models integrating adaptive beat segmentation & relative heart rate analysis with deep learning networks., Conducted ECG classification & artifact removal for continuous electrodermal activity monitoring., Designed and validated an underwater electrode for electrodermal activity recording.
• Undergraduate Researcher, University of Tehran, Developed a real-time fall detection system using multimodal sensor fusion and machine learning., Created a non-cuff-based blood pressure measurement system utilizing pulse wave velocity & peak detection algorithms.