OmniSalus

OmniSalus bridges the gap between health questions and medical care through AI-powered risk assessments, easy-to-understand health insights, and personalized doctor and hospital recommendations.

Comment

OmniSalus — Understand More. Care Smarter.

Your AI-powered companion for understanding health risks and making more informed healthcare decisions.

💡 Inspiration

Every day, millions of people experience health concerns but don't know where to begin. Should they visit a doctor? What could their symptoms or health indicators mean? Which specialist should they consult?

Searching online often leads to overwhelming medical information, confusing terminology, and unnecessary anxiety. Meanwhile, healthcare professionals frequently encounter patients who struggle to explain their concerns or understand their health indicators.

We asked ourselves a simple question:

What if everyone had an intelligent health companion that could help them understand their health before stepping into a doctor's office?

That's how OmniSalus was born.

Our vision is to bridge the gap between health uncertainty and professional medical care by making AI-powered health insights more accessible, understandable, and actionable.

We aren't trying to replace doctors. We're helping people ask better questions before they meet them.

🚀 What It Does

OmniSalus is an AI-powered health assessment platform combining machine learning with an interactive AI companion named Omni.

The platform brings together 7 health assessment suites and 9 predictive models within one unified experience.

Our assessment categories include:

  • Heart Disease: Multiple prediction pipelines evaluating cardiovascular health indicators.
  • Diabetes: Risk assessment using 21 health and lifestyle indicators.
  • Stroke: Risk classification based on relevant medical and lifestyle factors.
  • Liver Health: Classification of liver-related health conditions using laboratory indicators.
  • Chronic Kidney Disease: Machine learning classification using clinical measurements.
  • Breast Cancer: Classification models analyzing diagnostic cell measurements.
  • Obesity: Weight-category classification using lifestyle and physical indicators. 🤖 Meet Omni — Your AI Health Companion

What makes OmniSalus different is that users aren't left alone with complicated predictions.

Omni is an interactive AI assistant designed to help users:

  • Understand assessment questions and medical terminology.
  • Explore the health indicators used in each prediction model.
  • Interpret model outputs in clear, conversational language.
  • Prepare more informed questions for healthcare professionals.

Each assessment has its own context-aware assistant, creating a more relevant and focused experience.

Our long-term goal is to connect these insights with appropriate healthcare specialists and nearby hospitals, helping users move from understanding their concerns to finding professional care.

🛠️ How We Built It

We designed OmniSalus as a unified health intelligence platform, bringing multiple machine learning approaches into a single, user-friendly web experience.

  1. Machine Learning Integration

We integrated predictive models using algorithms including:

  • XGBoost
  • LightGBM
  • Random Forest
  • Logistic Regression
  • HistGradientBoosting

Different health conditions require different features, inputs, and prediction approaches. Our goal was to make these diverse models accessible through one consistent interface.

  1. Interactive Health Assessments

We developed condition-specific assessment forms that accept relevant clinical measurements, demographic information, and lifestyle indicators.

Users can review their information, run an assessment, and explore the resulting model output.

  1. Context-Aware AI Assistant

We incorporated an AI-powered conversational experience that helps explain each assessment's purpose, input fields, and prediction terminology.

Rather than offering a generic chatbot, Omni focuses on the health assessment the user is currently exploring.

  1. User-Centered Interface

We created a responsive, dark-themed interface with intuitive navigation, structured assessment cards, and accessible health information.

Our guiding principle was simple:

Complex intelligence behind the scenes. Clarity and simplicity for the user.

🧩 Challenges We Faced

One of the biggest challenges was combining multiple machine learning models with different input requirements into a consistent user experience.

A diabetes assessment may rely heavily on lifestyle information, while kidney and liver assessments require specific clinical measurements. Designing interfaces that respected these differences without overwhelming users was a major consideration.

Another challenge was translating technical model predictions into language that everyday users could understand.

In healthcare, responsible communication matters just as much as technical performance. We needed to distinguish statistical predictions from actual medical diagnoses and encourage users to seek professional medical guidance.

Balancing sophisticated AI functionality with a simple, approachable design was another important challenge.

📚 What We Learned

Building OmniSalus taught us that successful healthcare AI is about much more than prediction accuracy.

We learned that:

  • Explainability matters: Users need to understand what a result means, not just see a number.
  • Human-centered design matters: Powerful models are only useful when people can interact with them confidently.
  • Medical responsibility matters: AI-generated insights must communicate uncertainty and should never be presented as confirmed diagnoses.
  • Integration matters: Combining specialized models can create a more useful experience than presenting each model as an isolated tool.

Most importantly, we learned that AI's real potential in healthcare lies in supporting better-informed human decisions.

🏆 Accomplishments We're Proud Of

We're proud to have created a unified prototype featuring:

  • 7 health assessment suites.
  • 9 machine learning models.
  • A context-aware AI health companion.
  • Condition-specific interactive assessments.
  • A consistent, responsive health dashboard.
  • A foundation for connecting AI-driven health education with real-world healthcare services.

Instead of focusing on a single health condition, we created a platform designed to bring multiple health perspectives together.

🔮 What's Next for OmniSalus?

Our vision goes beyond health assessments.

We plan to expand OmniSalus with:

  1. Doctor and Hospital Recommendations: Help users discover relevant specialists and healthcare facilities based on their location and health concerns.
  2. Smarter Health Conversations: Improve Omni's ability to explain results and help users prepare for medical appointments.
  3. Personal Health Dashboard: Let users track their health indicators and assessment history over time.
  4. Multilingual Accessibility: Introduce Urdu and other regional languages to make health information more accessible.
  5. Responsible AI Improvements: Strengthen model validation, privacy safeguards, transparency, and clinical review before any real-world medical deployment.

Our ultimate goal is to transform OmniSalus from an educational health assessment prototype into a trusted companion that helps people navigate their healthcare journey.

❤️ Our Mission

We believe understanding your health shouldn't require a medical degree.

OmniSalus isn't here to replace healthcare professionals. It's here to make the journey toward professional care clearer, smarter, and more informed.

Understand More. Care Smarter.

Disclaimer: OmniSalus is an educational prototype. Machine learning predictions and AI-generated explanations are not medical diagnoses or substitutes for professional healthcare advice.

Share this project:

Updates