Pneumonia Detection Using CNN

• Developed a CNN-based machine learning model that achieved 92% accuracy in detecting pneumonia from X-ray images, improving diagnostic speed by 30% compared to traditional methods.

• Trained the model on a dataset of 5,000 X-ray images using popular machine learning libraries, including TensorFlow, Keras, and Tflearn, reducing training time by 15%.

• A user-friendly frontend was created using Flask, enabling users to upload X-ray images for real-time prediction.

• Integrated multiple CNN layers, including convolutional layers, ReLU layers, and pooling layers, to extract critical features from medical images.

• Employed data preprocessing techniques, such as image segmentation and median filtering, to enhance image quality and minimize noise.

• Achieved significant sample size reduction using max pooling, while retaining essential texture information for accurate classification.

• Incorporated fully connected layers at the final stage of the CNN to predict pneumonia presence based on extracted features.