Sistercheck

Inspiration

Our inspiration came from seeing how many women suffer in silence with ovarian cysts confused by medical jargon, fearful of unknown outcomes, and often overlooked in busy clinical systems. We were motivated by the stories of women who waited too long for diagnosis, felt dismissed during consultations, or had no idea what their scan results meant. At the same time, we spoke to clinicians who wanted to help but lacked the tools or time to offer truly personalized, proactive care.

What it does

Our solution uses AI to predict ovarian cyst growth and deliver clear, personalized insights to both patients and clinicians enabling earlier decisions, reduced anxiety, and better outcomes.

How we built it

We built our solution by combining machine learning, user-centered design, and medical insight to deliver a functional MVP that predicts cyst growth, supports patients, and aids clinicians.

AI Model for Risk Prediction:

We trained a machine learning model on anonymized ovarian cyst ultrasound features and simulated patient histories to output a growth risk score (Low/Medium/High). Built using Python with plans for clinical tuning during pilot phases.

Flutter Frontend

We used Flutter to build a clean, responsive interface for both patients and doctors — enabling cross-platform access on mobile and web.

Doctor Dashboard

A backend dashboard (Node.js ) shows patient risk assessment, cost estimation module, inventory module and an AI chabot.

Patient Portal

A backend dashboard that also allows a patient to see cyst risk assessment, cost estimation module, find a clinic nearby through geolocation and an AI Chatbot

AI Chatbot using Mistral AI

We integrated Mistral AI to create a chatbot that responds with empathy, explains results in simple terms, and guides users in managing their condition.

Anxiety-Aware UX

Designed with input from clinicians and patients to ensure calm, reassuring language and visual layouts that reduce fear especially during uncertain diagnoses.

Challenges we ran into

1. Serving Two Different User Types

   • Doctors want fast, clinical insights.
     
   • Patients want clarity and reassurance.
     
   • How we handled it: Built a dual-interface system — a doctor dashboard with clinical suggestions and a patient portal with simplified, friendly summaries.

2. New Team Collaboration

   • We were a newly formed team, meeting for the first time during this hackathon.
     
   • **How we handled it: Open communication and collaboration helped us work effectively and unify around a shared mission.

Accomplishments that we're proud of

Developed a Functional MVP:

We built a working prototype that includes: AI-powered cyst growth risk prediction (Low/Medium/High) Patient-friendly summary reports Clinician dashboard for monitoring and decision support Mistral-powered chatbot for patient education and support

What we learned

1. Communication Matters in Healthtech

   • It’s not just about giving answers it’s about how those answers are delivered.
     
   • Patients need clarity, not confusion. Comfort, not fear.

2. Balancing Two User Groups

   • Designing for both patients and clinicians required layered thinking:
     
     • Empathy for patients
       
     • Efficiency for clinicians

What's next for Sistercheck

1. Launch Our First Clinical Pilot

   • Deploy MVP in 1–2 partner clinics
     
   • Test the AI model, dashboard, and chatbot in real-world scenarios

2. Enhance Our AI Model with Real-World Data

   • Use pilot insights to refine the prediction algorithm
     
   • Incorporate more diverse, anonymized patient data
     
   • Improve accuracy, robustness, and clinical value

Built With

• ai
• flutter
• machine-learning
• mistral
• mongodb
• node.js
• python