Spaces:

rasmodev
/

Sepsis_Prediction_App_Streamlit

Running

App Files Files Community

Sepsis_Prediction_App_Streamlit / app.py

rasmodev

Update app.py

d0fd4bc verified 3 months ago

raw history blame contribute delete

No virus

6.84 kB

	import streamlit as st
	import pickle
	import pandas as pd
	import numpy as np

	# Page Title with Style
	st.markdown(
	f"""
	<div style="text-align: center;">
	<h1 style="color: #800000;">🩸 Sepsis Prediction App</h1>
	</div>
	""",
	unsafe_allow_html=True
	)

	# Welcome Message with Style (Centered)
	st.markdown(
	f"""
	<div style="text-align: center;">
	<p>👋 Welcome to the Sepsis Prediction App!</p>
	</div>
	""",
	unsafe_allow_html=True
	)

	# Sepsis Information
	st.markdown(
	"""
	Sepsis is a critical medical condition triggered by the body's extreme response to an infection. It can lead to organ failure and, if not detected early, poses a serious threat to life.
	"""
	)

	# Link to WHO Fact Sheet on Sepsis
	st.markdown("🔗 Learn more about sepsis from [World Health Organization (WHO)](https://www.who.int/news-room/fact-sheets/detail/sepsis#:~:text=Overview,problems%20are%20at%20higher%20risk.)")

	st.markdown("---")

	st.image("https://dinizululawgroup.com/wp-content/uploads/2020/07/news.jpg", width=700)

	# Additional Information for Sample Prediction
	st.write("📊 To make a sample prediction, you can refer to the training dataset information available in the sidebar.")
	st.write("💉 Enter the medical data in the input fields below, then click 'Predict Sepsis', and get the patient's Sepsis status prediction.")


	# About Section with Style
	st.sidebar.title("ℹ️ About")
	st.sidebar.info(
	"This app predicts sepsis onset using machine learning on medical data, aiding timely intervention by healthcare professionals. "
	"It utilizes a model trained on a sepsis dataset."
	)

	# Load The Train Dataset
	train_df = pd.read_csv("Train.csv", index_col=None)

	# Training Dataset Information in the sidebar
	st.sidebar.markdown("📊 Training Dataset Information:")
	st.sidebar.write("The model is trained on a sepsis dataset. Here's an overview of the dataset:")
	st.sidebar.write(train_df.head())

	# Auto-expand sidebar code
	st.markdown(
	"""
	<style>
	[data-testid="stSidebar"][aria-expanded="false"] > div:first-child {
	width: 100%;
	}
	</style>
	""",
	unsafe_allow_html=True
	)

	# Load the model and key components
	with open('model_and_key_components.pkl', 'rb') as file:
	loaded_components = pickle.load(file)

	loaded_model = loaded_components['model']
	loaded_scaler = loaded_components['scaler']

	# Data Fields
	data_fields = {
	"PRG": "Number of Pregnancies (applicable only to females)\n - The total number of pregnancies a female patient has experienced.",
	"PL": "Plasma Glucose Concentration (mg/dL)\n - The concentration of glucose in the patient's blood). It provides insights into the patient's blood sugar levels.",
	"PR": "Diastolic Blood Pressure (mm Hg)\n - The diastolic blood pressure, representing the pressure in the arteries when the heart is at rest between beats.",
	"SK": "Triceps Skinfold Thickness (mm)\n - The thickness of the skinfold on the triceps, measured in millimeters (mm). This measurement is often used to assess body fat percentage.",
	"TS": "2-hour Serum Insulin (mu U/ml)\n - The level of insulin in the patient's blood two hours after a meal, measured in micro international units per milliliter (mu U/ml).",
	"M11": "Body Mass Index (BMI) (weight in kg / {(height in m)}^2)\n - BMI provides a standardized measure that helps assess the degree of body fat and categorizes individuals into different weight status categories, such as underweight, normal weight, overweight, and obesity.",
	"BD2": "Diabetes pedigree function (mu U/ml)\n - The function provides information about the patient's family history of diabetes.",
	"Age": "Age of the Patient (years)\n - Age is an essential factor in medical assessments and can influence various health outcomes."
	}
	# Organize input fields into two columns
	col1, col2 = st.columns(2)

	# Initialize input_data dictionary
	input_data = {}

	# Function to preprocess input data
	def preprocess_input_data(input_data):
	numerical_cols = ['PRG', 'PL', 'PR', 'SK', 'TS', 'M11', 'BD2', 'Age']
	input_data_scaled = loaded_scaler.transform([list(input_data.values())])
	return pd.DataFrame(input_data_scaled, columns=numerical_cols)

	# Function to make predictions
	def make_predictions(input_data_scaled_df):
	y_pred = loaded_model.predict(input_data_scaled_df)
	sepsis_mapping = {0: 'Negative', 1: 'Positive'}
	return sepsis_mapping[y_pred[0]]

	# Input Data Fields in two columns
	with col1:
	input_data["PRG"] = st.slider("PRG: Number of Pregnancies", 0, 20, 0)
	input_data["PL"] = st.number_input("PL: Plasma Glucose Concentration (mg/dL)", value=0.0)
	input_data["PR"] = st.number_input("PR: Diastolic Blood Pressure (mm Hg)", value=0.0)
	input_data["SK"] = st.number_input("SK: Triceps Skinfold Thickness (mm)", value=0.0)

	with col2:
	input_data["TS"] = st.number_input("TS: 2-Hour Serum Insulin (mu U/ml)", value=0.0)
	input_data["M11"] = st.number_input("M11: Body Mass Index (BMI)", value=0.0)
	input_data["BD2"] = st.number_input("BD2: Diabetes Pedigree Function (mu U/ml)", value=0.0)
	input_data["Age"] = st.slider("Age: Age of the patient (years)", 0, 100, 0)

	# Predict Button with Style
	if st.button("🔮 Predict Sepsis"):
	try:
	input_data_scaled_df = preprocess_input_data(input_data)
	sepsis_status = make_predictions(input_data_scaled_df)

	# Display the sepsis prediction
	if sepsis_status == 'Negative':
	st.success(f"The predicted sepsis status is: {sepsis_status}")
	elif sepsis_status == 'Positive':
	st.error(f"The predicted sepsis status is: {sepsis_status}")

	# Add the sepsis prediction to the input DataFrame
	input_data['Sepsis'] = sepsis_status

	# Convert the input data to a pandas DataFrame
	input_df = pd.DataFrame([input_data])

	# Display the input DataFrame
	st.markdown(
	f"""
	<div style="text-align: center; font-size: 18px; font-weight: bold;">
	<p>Input Data with Sepsis Prediction</p>
	</div>
	""",
	unsafe_allow_html=True
	)
	st.table(input_df)

	except Exception as e:
	st.error(f"An error occurred: {e}")

	# Display Data Fields and Descriptions
	st.sidebar.title("🔍 Data Fields")
	for field, description in data_fields.items():
	st.sidebar.markdown(f"{field}: {description}")

	# Copyright statement at the bottom
	st.markdown(
	"""
	<div style="text-align:center">
	Developed by <a href="https://www.linkedin.com/in/rasmo-/" style="font-style:italic">Rasmo Wanyama</a>.
	</div>
	""",
	unsafe_allow_html=True
	)

	st.stop()