|
from smolagents import CodeAgent, DuckDuckGoSearchTool, HfApiModel |
|
from Bio import Entrez |
|
from fpdf import FPDF |
|
import tempfile |
|
|
|
|
|
clinical_agent = CodeAgent(tools=[DuckDuckGoSearchTool()], model=HfApiModel()) |
|
|
|
|
|
Entrez.email = "Pub_Email" |
|
|
|
def generate_clinical_content(topic, audience): |
|
"""Generates medical content based on topic and audience.""" |
|
prompt = f"Write a detailed medical article on: {topic}.\nTarget Audience: {audience}.\nInclude latest medical research insights." |
|
return clinical_agent.run(prompt) |
|
|
|
def generate_summary(content): |
|
"""Summarizes a given clinical document or research paper.""" |
|
summary_prompt = f"Summarize the following medical research:\n{content}" |
|
return clinical_agent.run(summary_prompt) |
|
|
|
def generate_soap_note(symptoms, history): |
|
"""Generates a SOAP Note for clinicians based on symptoms and patient history.""" |
|
soap_prompt = ( |
|
f"Create a structured SOAP Note for a patient with:\n" |
|
f"Symptoms: {symptoms}\n" |
|
f"History: {history}\n" |
|
f"Include Assessment & Plan." |
|
) |
|
return clinical_agent.run(soap_prompt) |
|
|
|
def fetch_pubmed_articles(query): |
|
"""Fetches latest medical research from PubMed based on a query.""" |
|
try: |
|
handle = Entrez.esearch(db="pubmed", term=query, retmax=5) |
|
record = Entrez.read(handle) |
|
handle.close() |
|
article_ids = record["IdList"] |
|
|
|
articles = [] |
|
for article_id in article_ids: |
|
handle = Entrez.efetch(db="pubmed", id=article_id, retmode="xml") |
|
article_record = Entrez.read(handle) |
|
handle.close() |
|
|
|
article_data = article_record["PubmedArticle"][0]["MedlineCitation"]["Article"] |
|
title = article_data.get("ArticleTitle", "No Title Available") |
|
abstract = article_data.get("Abstract", {}).get("AbstractText", ["No Abstract Available"])[0] |
|
authors = ", ".join([author["LastName"] for author in article_data.get("AuthorList", []) if "LastName" in author]) |
|
url = f"https://pubmed.ncbi.nlm.nih.gov/{article_id}/" |
|
|
|
articles.append({"title": title, "abstract": abstract, "authors": authors, "url": url}) |
|
|
|
return articles |
|
except Exception as e: |
|
return [{"title": "Error Fetching Articles", "abstract": str(e), "authors": "N/A", "url": "#"}] |
|
|
|
def save_as_text(content, filename): |
|
"""Saves AI-generated content as a TXT file.""" |
|
with tempfile.NamedTemporaryFile(delete=False, suffix=".txt") as tmp_file: |
|
tmp_file.write(content.encode()) |
|
return tmp_file.name, filename |
|
|
|
def save_as_pdf(content, filename): |
|
"""Saves AI-generated content as a PDF file.""" |
|
with tempfile.NamedTemporaryFile(delete=False, suffix=".pdf") as tmp_file: |
|
pdf = FPDF() |
|
pdf.add_page() |
|
pdf.set_font("Arial", size=12) |
|
pdf.multi_cell(0, 10, content) |
|
pdf.output(tmp_file.name) |
|
return tmp_file.name, filename |
|
|
