Spaces:
Running
Running
| import gradio as gr | |
| import torch | |
| from transformers import pipeline | |
| import tensorflow as tf | |
| import numpy as np | |
| from PIL import Image | |
| import random | |
| # Load AI models | |
| symptom_model = pipeline("text-classification", model="nlp-ai/symptom-disease") # Replace with trained model | |
| emotion_model = pipeline("sentiment-analysis") # Pre-trained NLP for emotions | |
| retina_model = tf.keras.models.load_model("retina_disease_model.h5") # Custom trained Retina AI | |
| stool_model = tf.keras.models.load_model("stool_microbiome_model.h5") # Gut microbiome detection AI | |
| # Function for AI-powered diagnosis | |
| def diagnose(symptoms): | |
| result = symptom_model(symptoms)[0] | |
| return f"Condition: {result['label']} (Confidence: {round(result['score'] * 100, 2)}%)" | |
| # Function for stool image microbiome analysis | |
| def analyze_stool(image): | |
| img = Image.open(image).resize((224, 224)) | |
| img_array = np.array(img) / 255.0 | |
| img_array = np.expand_dims(img_array, axis=0) | |
| prediction = stool_model.predict(img_array) | |
| class_names = ["Healthy Gut", "Possible Dysbiosis", "Severe Gut Imbalance"] | |
| return f"Result: {class_names[np.argmax(prediction)]}" | |
| # Retina disease detection function | |
| def analyze_retina(image): | |
| img = Image.open(image).resize((224, 224)) | |
| img_array = np.array(img) / 255.0 | |
| img_array = np.expand_dims(img_array, axis=0) | |
| prediction = retina_model.predict(img_array) | |
| class_names = ["Normal Retina", "Diabetic Retinopathy", "Glaucoma Detected"] | |
| return f"Retina Scan: {class_names[np.argmax(prediction)]}" | |
| # Emotion-to-disease analysis | |
| def emotion_to_disease(emotion_text): | |
| emotions = { | |
| "stress": "Long-term stress can cause high blood pressure, anxiety, and digestive issues.", | |
| "anger": "Frequent anger can increase heart disease risk.", | |
| "happiness": "A positive mindset improves overall well-being!" | |
| } | |
| detected_emotion = emotion_model(emotion_text)[0]['label'] | |
| return emotions.get(detected_emotion.lower(), "No specific health risks detected.") | |
| # Gradio UI | |
| with gr.Blocks(theme="soft") as demo: | |
| gr.Markdown("# 🏥 Diagnosify-AI: AI-powered Health & Disease Detection") | |
| # Symptom diagnosis | |
| with gr.Row(): | |
| symptom_input = gr.Textbox(label="Enter symptoms") | |
| symptom_output = gr.Textbox(label="Diagnosis Result", interactive=False) | |
| gr.Button("Diagnose").click(diagnose, inputs=symptom_input, outputs=symptom_output) | |
| # Gut microbiome (stool analysis) | |
| with gr.Row(): | |
| stool_input = gr.Image(label="Upload Stool Image") | |
| stool_output = gr.Textbox(label="Microbiome Analysis", interactive=False) | |
| gr.Button("Analyze Stool").click(analyze_stool, inputs=stool_input, outputs=stool_output) | |
| # Retina scan | |
| with gr.Row(): | |
| retina_input = gr.Image(label="Upload Retina Image") | |
| retina_output = gr.Textbox(label="Retina Disease Detection", interactive=False) | |
| gr.Button("Analyze Retina").click(analyze_retina, inputs=retina_input, outputs=retina_output) | |
| # Emotion-to-Disease | |
| with gr.Row(): | |
| emotion_input = gr.Textbox(label="Describe your emotions") | |
| emotion_output = gr.Textbox(label="Health Analysis", interactive=False) | |
| gr.Button("Analyze Emotion").click(emotion_to_disease, inputs=emotion_input, outputs=emotion_output) | |
| gr.Markdown("🔗 **Diagnosify-AI** - Bringing AI-powered diagnostics to your fingertips!") | |
| # Launch the app | |
| demo.launch() | |