Upload 2 files

requirements.txt

@@ -0,0 +1,6 @@
+
+
+streamlit
+openai
+pydub
+requests

task.py

@@ -0,0 +1,167 @@
+import streamlit as st
+import openai
+from pydub import AudioSegment
+import io
+import time
+
+# Set your OpenAI API key
+openai.api_key = "YOUR_API_KEY"
+
+# Function to summarize the transcript into medical sections
+def medical_summary(transcript):
+    prompt = f"""
+    Organize the following medical transcript into the predefined sections:
+
+    Sections:
+    1. Medical Specialty
+    2. CHIEF COMPLAINT
+    3. Purpose of visit
+    4. HISTORY and Physical
+       - PAST MEDICAL HISTORY
+       - PAST SURGICAL HISTORY
+       - ALLERGIES History
+       - Social History
+       - REVIEW OF SYSTEMS
+    5. PHYSICAL EXAMINATION
+       - GENERAL
+       - Vitals
+       - ENT
+       - Head
+       - Neck
+       - Chest
+       - Heart
+       - Abdomen
+       - Pelvic
+       - Extremities
+
+    Transcript:
+    {transcript}
+
+    Provide a structured summary in the above format.
+    """
+    try:
+        response = openai.ChatCompletion.create(
+            model="gpt-4",  # Use gpt-3.5-turbo if GPT-4 is unavailable
+            messages=[
+                {"role": "system", "content": "You are a helpful medical assistant."},
+                {"role": "user", "content": prompt}
+            ],
+            temperature=0.5,
+            max_tokens=3000  # Adjust max_tokens based on the expected size of the summary
+        )
+        structured_summary = response['choices'][0]['message']['content'].strip()
+        return structured_summary
+    except Exception as e:
+        return f"An error occurred: {e}"
+
+# Function to convert audio to text using OpenAI's Whisper or any suitable service
+def transcribe_audio(audio_file):
+    audio = AudioSegment.from_file(audio_file)
+    audio = audio.set_channels(1).set_frame_rate(16000)  # Convert to mono and set appropriate frame rate
+
+    buffer = io.BytesIO()
+    audio.export(buffer, format="wav")
+    buffer.seek(0)
+
+    # Here, you would normally call the transcription API or process the file
+    # Example (assuming transcription API integration):
+    # response = openai.Audio.transcribe(model="whisper-1", file=buffer)
+
+    return "Patient complains of chest pain, shortness of breath, and dizziness. He has a history of hypertension."
+
+# Streamlit UI
+def main():
+    # Custom CSS to improve UI aesthetics
+    st.markdown("""
+        <style>
+            body {
+                font-family: 'Arial', sans-serif;
+                background-color: #f4f4f9;
+            }
+            .title {
+                font-size: 36px;
+                color: #1E2A47;
+                font-weight: 600;
+                text-align: center;
+                padding-top: 30px;
+                padding-bottom: 10px;
+            }
+            .subheader {
+                font-size: 24px;
+                color: #2F4F9A;
+                text-align: center;
+                margin-bottom: 20px;
+            }
+            .card {
+                background-color: #ffffff;
+                border-radius: 8px;
+                box-shadow: 0 4px 8px rgba(0, 0, 0, 0.1);
+                padding: 20px;
+                margin-bottom: 30px;
+            }
+            .button {
+                background-color: #4CAF50;
+                color: white;
+                padding: 12px 20px;
+                font-size: 16px;
+                border-radius: 8px;
+                border: none;
+                cursor: pointer;
+                transition: background-color 0.3s ease;
+            }
+            .button:hover {
+                background-color: #45a049;
+            }
+            .error {
+                color: red;
+                font-weight: bold;
+                text-align: center;
+            }
+            .footer {
+                text-align: center;
+                margin-top: 40px;
+                color: #555;
+            }
+        </style>
+    """, unsafe_allow_html=True)
+
+    # Title and subtitle
+    st.markdown('<div class="title">Medical Transcription Dashboard</div>', unsafe_allow_html=True)
+    st.markdown('<div class="subheader">Upload an audio file or enter the transcript to generate a medical summary</div>', unsafe_allow_html=True)
+
+    # Layout: Input Area in Card
+    with st.container():
+        st.markdown('<div class="card">', unsafe_allow_html=True)
+
+        # Audio upload and text input options
+        audio_file = st.file_uploader("Upload an audio file (wav, mp3, etc.)", type=["wav", "mp3"], label_visibility="collapsed")
+        transcript_text = st.text_area("Or, enter the transcript manually:", height=200)
+
+        # Show "Generate Medical Summary" button
+        if st.button("Generate Medical Summary", key="generate", use_container_width=True):
+            if audio_file:
+                # Show loading spinner while processing
+                with st.spinner('Transcribing audio...'):
+                    transcript = transcribe_audio(audio_file)
+                    time.sleep(2)  # Simulate delay
+                    st.write(f"Transcript: {transcript}")
+            elif transcript_text:
+                transcript = transcript_text
+            else:
+                st.markdown('<div class="error">Please upload an audio file or enter a transcript.</div>', unsafe_allow_html=True)
+
+            if transcript:
+                # Show loading spinner while generating summary
+                with st.spinner('Generating Medical Summary...'):
+                    time.sleep(2)  # Simulate processing time
+                    summary = medical_summary(transcript)
+                    st.markdown("### Medical Summary:")
+                    st.write(summary)
+
+        st.markdown('</div>', unsafe_allow_html=True)
+
+    # Footer
+    st.markdown('<div class="footer">Developed with ❤️ by Your Company</div>', unsafe_allow_html=True)
+
+if __name__ == "__main__":
+    main()