# -----------------------------
# Install pydicom if not present
# -----------------------------
import subprocess
import sys

try:
    import pydicom
except ImportError:
    subprocess.check_call([sys.executable, "-m", "pip", "install", "pydicom"])
    import pydicom


import streamlit as st
import numpy as np
import requests
import cv2
import tempfile

# -----------------------------
# Backend API URL
# -----------------------------
API_URL = "https://viveksardey-pneumoniadetectionpredictionbackend.hf.space/v1/predict"

IMG_SIZE = 224

st.title("🩺 Pneumonia Detection from Chest X-ray (DICOM)")

st.write("""
Upload a **Chest X-ray DICOM (.dcm)** file.
The image will be converted to **NumPy (.npy)** format and sent to the backend model.
""")

# -----------------------------
# Upload file
# -----------------------------
uploaded_file = st.file_uploader(
    "Upload DICOM Chest X-ray",
    type=["dcm"]
)

# -----------------------------
# Convert DICOM → Image
# -----------------------------
def dicom_to_image(dicom_file):

    dicom = pydicom.dcmread(dicom_file)
    img = dicom.pixel_array

    return img


# -----------------------------
# Preprocess image
# -----------------------------
def preprocess_image(img):

    # Resize image
    img = cv2.resize(img, (IMG_SIZE, IMG_SIZE))

    # Normalize
    img = img.astype("float32")

    if img.max() > 0:
        img = img / img.max()

    return img


# -----------------------------
# If file uploaded
# -----------------------------
if uploaded_file is not None:

    try:

        img = dicom_to_image(uploaded_file)

        st.subheader("X-ray Preview")
        st.image(img, caption="Uploaded DICOM Image", use_column_width=True)

        processed_img = preprocess_image(img)

        npy_array = np.array(processed_img)

        st.success("DICOM successfully converted to NumPy format")

    except Exception as e:

        st.error(f"Error processing DICOM file: {e}")


# -----------------------------
# Predict button
# -----------------------------
if st.button("Predict Pneumonia"):

    if uploaded_file is None:

        st.warning("Please upload a DICOM image first")

    else:

        try:

            with st.spinner("Running Pneumonia Detection..."):

                # Save temporary npy file
                with tempfile.NamedTemporaryFile(suffix=".npy", delete=False) as tmp:

                    np.save(tmp.name, npy_array)
                    tmp_path = tmp.name

                # Send request to backend API
                with open(tmp_path, "rb") as f:

                    response = requests.post(
                        API_URL,
                        files={"file": f}
                    )

                if response.status_code == 200:

                    result = response.json()

                    predicted_class = result["Predicted_Class"]
                    probability = result["Probability_Pneumonia"]

                    st.subheader("Prediction Result")

                    if predicted_class == "Pneumonia":
                        st.error("⚠️ Pneumonia Detected")
                    else:
                        st.success("✅ Normal Lungs")

                    st.write(f"Probability of Pneumonia: **{probability:.4f}**")

                else:

                    st.error(f"Prediction failed. Status Code: {response.status_code}")

        except Exception as e:

            st.error(f"Error during prediction: {e}")