app.py

import streamlit as st
from transformers import pipeline
from PIL import Image, ImageDraw

st.set_page_config(
    page_title="Fraktur Detektion",
    layout="wide",
    initial_sidebar_state="collapsed"
)

st.markdown("""
<style>
    .stApp {
        padding: 0 !important;
        height: 100vh !important;
        overflow: hidden !important;
    }

    .block-container {
        padding: 0.25rem !important;
        max-width: 100% !important;
    }

    .stImage > img {
        width: 70% !important;
        height: auto !important;
        max-height: 150px !important;
        object-fit: contain !important;
    }

    h2, h3 {
        font-size: 0.9rem !important;
    }

    .result-box {
        font-size: 0.8rem !important;
        margin: 0.2rem 0 !important;
    }
</style>
""", unsafe_allow_html=True)

@st.cache_resource
def load_models():
    return {
        "KnochenAuge": pipeline("object-detection", model="D3STRON/bone-fracture-detr"),
        "KnochenWächter": pipeline("image-classification", model="Heem2/bone-fracture-detection-using-xray"),
        "RöntgenMeister": pipeline("image-classification", 
            model="nandodeomkar/autotrain-fracture-detection-using-google-vit-base-patch-16-54382127388")
    }

def translate_label(label):
    translations = {
        "fracture": "Knochenbruch",
        "no fracture": "Kein Bruch",
        "normal": "Normal",
        "abnormal": "Auffällig"
    }
    return translations.get(label.lower(), label)

def draw_boxes(image, predictions):
    draw = ImageDraw.Draw(image)
    for pred in predictions:
        box = pred['box']
        label = f"{translate_label(pred['label'])} ({pred['score']:.2%})"
        color = "#2563eb" if pred['score'] > 0.7 else "#eab308"

        draw.rectangle(
            [(box['xmin'], box['ymin']), (box['xmax'], box['ymax'])],
            outline=color,
            width=2
        )

        # Ajouter des points de "chaleur" aux fractures détectées
        center_x = (box['xmin'] + box['xmax']) / 2
        center_y = (box['ymin'] + box['ymax']) / 2
        radius = 5
        draw.ellipse(
            [(center_x - radius, center_y - radius), (center_x + radius, center_y + radius)],
            fill=color
        )

        # Label plus compact
        text_bbox = draw.textbbox((box['xmin'], box['ymin']-15), label)
        draw.rectangle(text_bbox, fill=color)
        draw.text((box['xmin'], box['ymin']-15), label, fill="white")
    return image

def main():
    models = load_models()

    # Disposition en deux colonnes principales
    col1, col2 = st.columns([1, 2])

    with col1:
        st.markdown("### 📤 Röntgenbild Upload")
        uploaded_file = st.file_uploader("", type=['png', 'jpg', 'jpeg'])

        if uploaded_file:
            conf_threshold = st.slider(
                "Konfidenzschwelle",
                min_value=0.0, max_value=1.0,
                value=0.60, step=0.05
            )

    with col2:
        if uploaded_file:
            image = Image.open(uploaded_file)

            # Toujours afficher les résultats des autres modèles
            st.markdown("### 🎯 KI-Analyse")

            st.markdown("**🛡️ Der KnochenWächter**")
            predictions_wachter = models["KnochenWächter"](image)
            for pred in predictions_wachter:
                score_color = "#22c55e" if pred['score'] > 0.7 else "#eab308"
                st.markdown(f"""
                    <div class='result-box'>
                        <span style='color: {score_color}; font-weight: 500;'>
                            {pred['score']:.1%}
                        </span> - {translate_label(pred['label'])}
                    </div>
                """, unsafe_allow_html=True)

            st.markdown("**🎓 Der RöntgenMeister**")
            predictions_meister = models["RöntgenMeister"](image)
            for pred in predictions_meister:
                score_color = "#22c55e" if pred['score'] > 0.7 else "#eab308"
                st.markdown(f"""
                    <div class='result-box'>
                        <span style='color: {score_color}; font-weight: 500;'>
                            {pred['score']:.1%}
                        </span> - {translate_label(pred['label'])}
                    </div>
                """, unsafe_allow_html=True)

            # Analyse avec KnochenAuge (localisation)
            predictions_auge = models["KnochenAuge"](image)
            filtered_preds = [p for p in predictions_auge if p['score'] >= conf_threshold]

            if filtered_preds:
                st.markdown("#### 👁️ Das KnochenAuge - Lokalisation")
                result_image = image.copy()
                result_image = draw_boxes(result_image, filtered_preds)
                st.image(result_image, use_container_width=True)

        else:
            st.info("Bitte laden Sie ein Röntgenbild hoch (JPEG, PNG)")

if __name__ == "__main__":
    main()