Teaching

A decade in industry taught me what textbooks cannot: how to ship. Today, with AI making raw information instantly accessible, the classroom must shift from delivering answers to building system-level thinking, the ability to frame a problem, define requirements, and reason through a solution from first principles. The teacher's role becomes more essential, not less: guiding that reasoning, providing context AI cannot replicate, and turning surface-level answers into deep understanding. My teaching brings field experience into this environment, where students learn to think like engineers, not search like users.

Proposed Courses

Graduate

Biomedical Signal Processing

Processing and interpretation of physiological signals (PPG, ECG, EEG) with emphasis on filtering, feature extraction, and real-world sensor data.

Prerequisites: Signals and Systems, basic programming (Python)

Biomedical Image and Video Processing

Image enhancement, segmentation, and video-based analysis for biomedical applications including camera-based physiological measurement.

Prerequisites: Digital Signal Processing, basic image processing concepts

Computational Imaging for Physiological Sensing

Using standard cameras as computational sensors to extract invisible physiological information, rPPG, pulse waveform analysis, and remote vital sign estimation.

Prerequisites: Biomedical Signal Processing or equivalent

Embedded Systems & IoT for Wearable Devices

Designing and prototyping wearable health monitors: sensor integration, embedded firmware, wireless data pipelines, and power management.

Prerequisites: Basic embedded systems or microcontroller experience

Undergraduate

Signals and Systems

Foundations of continuous and discrete-time signal analysis: Fourier, Laplace, convolution, filtering, and system response.

Prerequisites: Calculus, linear algebra

Introduction to Digital Signal Processing

Sampling, quantization, DFT, digital filter design, and hands-on implementation of signal processing algorithms.

Prerequisites: Signals and Systems

Introduction to Biosensors

Principles of physiological sensing: optical, electrical, and mechanical transducers for biomedical measurement, from theory to signal acquisition.

Prerequisites: Basic physics, introductory electronics

Internships & Graduation Projects

I supervise research internships and graduation projects (B.Sc., M.Sc.) in signal processing, applied AI for healthcare, and full-stack development for biosensing platforms. If you're interested, get in touch. You can also propose your own topic, I'm open to ideas intersecting signal processing, biosensing, or applied AI for healthcare.

Project Directions

Signal Processing & Wearable Biosensing

  • Real-time PPG/rPPG signal quality assessment and artifact detection under motion
  • Adaptive filtering pipelines for wearable ECG/PPG denoising (LMS, RLS, Kalman variants)
  • Edge-deployable heart rate variability (HRV) feature extraction on Raspberry Pi / Jetson
  • Multi-wavelength optical sensing for SpO₂ estimation from reflectance-mode signals

Full-Stack & Biosensing Platforms

  • Building a real-time biosignal streaming API (FastAPI + WebSocket) with dashboard visualization (React/Vue)
  • Designing a laboratory information system for biosignal experiment management (subject, session, device, annotation)
  • Containerized deployment pipeline (Docker + CI/CD) for biosignal processing microservices

Agentic AI for Biomonitoring

  • LLM-powered RAG system for biomedical literature synthesis and hypothesis generation from physiological datasets
  • Multi-agent framework for automated physiological report drafting from raw biosignal recordings
  • Conversational agent for patient health data exploration and insight generation

Exploratory Research

  • Investigating PPG waveform morphology restoration using deep generative models (VAEs, GANs, diffusion)
  • Cross-domain transfer learning for physiological signal representation across optical, electrical, and mechanical sensors
  • Federated learning approaches for privacy-preserving physiological model training across institutions
  • Explainable AI (XAI) for cardiovascular risk prediction from wearable photoplethysmographic data