AI-Powered COVID-19 Detection via Chest X-Rays

🔹 Problem

During the COVID-19 pandemic, hospitals faced an overwhelming surge of patients needing diagnosis. Traditional testing methods such as PCR were accurate but often slow or limited in availability. Radiologists began exploring chest X-ray imaging as a rapid screening tool, but manual review was time-consuming and error-prone.

The client required an AI system that could quickly analyze chest X-rays to detect signs of COVID-19 infection and assist doctors in triaging patients faster.

🔹 Solution

I built a TensorFlow/Keras deep learning model with OpenCV preprocessing to classify chest X-rays as COVID-positive or negative.

Applied augmentation & normalization to improve accuracy on limited medical data.

Fine-tuned a pre-trained CNN (transfer learning).

Achieved ~92% accuracy, with 100% sensitivity (no positive cases missed).

🔹 Impact

✅ Rapid triage — results in seconds, not minutes. ✅ Supported doctors during peak pandemic workloads. ✅ Scalable framework for detecting other lung diseases.

🔹 Tech Stack

TensorFlow · Keras · OpenCV

👉 Result: Delivered an AI screening assistant that helped healthcare teams diagnose COVID-19 faster and more accurately.