lung-cancer-cnn-model

Lung Cancer CNN Model

A TensorFlow CNN model to classify chest CT images as Normal or Lung Cancer (~96% accuracy).

Dataset

• IQ-OTH/NCCD Lung Cancer Dataset
• Input: 256x256 grayscale images
• Classes: Normal, Lung Cancer

Usage

• Model: SavedModel format (lung_cancer_cnn_saved_model.zip)
• Inference: Use handler.py for API endpoints
• Accuracy: ~96%

Local Inference

import tensorflow as tf
import numpy as np
import cv2

model = tf.saved_model.load("lung_cancer_cnn_saved_model")
infer = model.signatures['serving_default']
class_names = ['Normal', 'Lung Cancer']

def predict(image_path):
    img = cv2.imread(image_path, cv2.IMREAD_GRAYSCALE)
    img = cv2.resize(img, (256, 256)) / 255.0
    img = img.reshape(1, 256, 256, 1).astype(np.float32)
    input_key = list(infer.structured_input_signature[1].keys())[0]
    outputs = infer(**{input_key: tf.convert_to_tensor(img)})
    output_key = list(outputs.keys())[0]
    pred = outputs[output_key].numpy()[0][0]
    class_id = 1 if pred > 0.5 else 0
    confidence = pred if class_id == 1 else 1 - pred
    return class_names[class_id], confidence

result, confidence = predict("path/to/ct.jpg")
print(f"Prediction: {result}, Confidence: {confidence:.4f}")

Disclaimer

For educational use only. Consult healthcare professionals.