MULTIPLE DISEASE PREDICTION 

ABSTRACT

In a world where advanced medical devices coexist with significant time constraints in healthcare, we are launching a machine learning project to identify patients with serious diseases (heart disease, diabetes, pneumonia and brain tumor) at an early stage to facilitate timeliness and efficiency. treatment . Through careful data collection, analysis and cleaning of various datasets, we determined the most suitable algorithms and achieved significant accuracy. Relying on cutting-edge technologies ensures robust and accurate predictions for all four models, while the user-friendly Flask website simplifies the process, making early disease detection easier and potentially life-saving for patients and healthcare professionals. This project combines machine learning, data analytics and data science to bridge the gap between the latest diagnostic tools and time-sensitive healthcare services.

INTRODUCTION

   In our information world, where information is a valuable asset, healthcare is no exception to the growth of information production. Patient information is the lifeblood of healthcare, encompassing a wide range of patient information and history. However, traditional disease prediction models tend to be tailored to individual diseases, which has led to a piecemeal approach to health analytics. Each disease - be it heart disease, diabetes, pneumonia or a brain tumor - was answered with separate predictive models. The lack of a unified system for evaluating multiple diseases simultaneously has created a significant gap in the field. In response to this challenge, we present an innovative architecture that rapidly and accurately predicts these critical diseases based on user input symptoms. Our system, built on the solid foundation of streamlit, will initially focus on analyzing these four main diseases. However, our design is highly scalable, making room for new diseases in the future. The main advantage of this approach is that users do not have to navigate multiple platforms in search of disease predictions. This versatile system uses machine learning algorithms with Streamlit and offers an intuitive user interface. When users use our system, they enter both disease-specific parameters and the name of the disease being evaluated. In turn, the system applies the corresponding model to quickly assess the patient's condition

PROBLEM DEFINITION AND OBJECTIVES

A. Problem Definition

Modern healthcare systems are fulfilled with latest and effective diagnostic systems then also a lot of patients suffer from death due to lack of treatment on time. One thing that healthcare systems lack is time so they cannot determine which patient should be treated first. As a result, a needy patient could not get treatment on time which may cost his life too.

B. Objectives

1. Support healthcare systems
2. Reduce workload of healthcare systems and save their time
3. Identify patients having high risk to particular disease at an early stage
4. Build Relationships
5. Effectively predict if patient have chances of developing particular disease

CONCLUSION

We are successfully building a system that predicts more than one disease with high accuracy. It is also easy to use. We achieve design accuracy  as follows: 

        Brain Tumor Prediction:

1. Accuracy:Approximately 96%

2. Algorithm Used: Support Vector Machine (SVC)

3. Significance: The model successfully predicts brain tumors with a high accuracy rate, offering reliable diagnostic support for this critical medical condition.

Heart Disease Prediction:

1. Accuracy: 92%

2. Algorithm Used: Logistic Regression

3. Significance: The heart disease prediction model demonstrates a strong performance, providing accurate assessments of heart-related conditions for better patient care.

       Diabetes Prediction:

1. Accuracy: 99%

2. Algorithm Used: Decision Trees

3. Significance: The diabetes prediction model excels with an impressive 99% accuracy, enabling early detection and effective management of diabetes, a prevalent and potentially severe disease.

Pneumonia Prediction:

1. Accuracy: 97%

2. Algorithm Used: The ensemble ML models performed better in this study.

3. Significance: The pneumonia prediction model aims to provide accurate assessments of pneumonia, which is crucial for timely treatment and management of respiratory conditions.