Accurately guide and detect intraoperative bleeding spots during cardiac surgery at a quicker rate than the human eye
2nd Place, MIT Grand Medicine Hackathon, 2024
Problem
It’s a bloody mess during surgery. We can’t even see where the bleeding is coming from. It’s difficult to spot the bleeding spot. Time is of the essence: anesthesia time, cost of surgery in time for insurance and out-of-pocket, recovery time, and complications.
Scale
2 million open heart surgeries are performed each year. The most commonly performed is Coronary artery bypass grafting aka CABG. Do you know what average amount of blood loss is? The range about 1500mL blood loss in cardiac surgery. CABG (Coronary artery bypass grafting) is the most common surgery performed in the United States each year.
Objective
We are looking for a way to more accurately and quickly guide and detect bleeding spots at a quicker rate than human capabilities— not dependent on luck: how much sleep the surgeon got the night before and how much experience they have.
Debunk Assumption: Bleeding does not only happen when there’s an issue. Bleeding is bound to happen during surgery. Bleeding is a part of all surgery; a distraction to get to the actual site. Vessels must be burnt (cauterization) to reach surgical sites, and even after repair, closure is delayed by persistent bleeding.
We are looking to address a recurring problem that occurs in surgery workflows involving bleeding:
Initial Incision
The procedure begins by making an incision, cutting through the epidermis and dermis (outer skin layers). Just beneath the skin lies a network of veins, which are the first blood vessels encountered.
Managing Vein Bleeding
- At this stage, a Bovie electrocautery device is used.
- It has two buttons: one for cutting and one for cauterizing (using electric heat to seal blood vessels).
- The surgeon applies the cauterization for 1-5 seconds to stop bleeding from small veins.
Deeper Layers – Arteries
- As the incision deepens, arteries are reached. Unlike veins, arteries are larger, high-pressure vessels carrying oxygenated blood, so bleeding is more pronounced.
- Controlling arterial bleeding requires additional techniques:
Hemostasis Techniques (Stopping Bleeding)
1. First Approach: Electrocautery (Bovie) – The same device may be used on smaller arteries.
2. Second Approach: Hemostatic Clip ("Pac-Man" Device)
- If bleeding persists, a clip is applied to the artery to clamp it shut.
3. Third Approach: Suturing
- For deeper muscle layers or major arteries, a suture (stitching technique) is used to securely close the vessel and stop bleeding.
4. Fourth Approach (Rare) and high mortality risk
- Substances you give: a mesh surgicell absorbs blood to stop the bleeding
- arista - potatoe starch
- Blood transfusion – poses a huge risk because it causes shock. You can get an adverse reaction from getting blood. When you transfuse, your 7.6% mortality increase for cardiac surgery. Mortality increases
Rapid Prototyping
At the MIT MakerWorkshop, 3D printing an attachment onto current goggles made by Phuc An Dinh. I worked collaboratively and parellel to make a lazer cut piece that with An to make a Lazer Cut piece that is heat bended
Detailed Orthographic Drawings
Challenges
The original design changed with when the electrical components and its dimensions changed. The electrical engineer of the previous design was done by Venkadesh Eswaranandam. The biggest design challenge was creating a compact housing for the electronics, primarily dictated by battery size. I worked closely with electrical engineer Aditya Bangde to precisely fit his components—RFID, camera, and microphone—using calipers to match the components to the CAD. The design focused on being compact, lightweight, and sturdy, with an internal structure that secured components despite movement.
Collaborative Work
Teammates from Stanford Hackathon: Ben Cullen, Elsa Bis, Aditya Bangde
Teammates from previous MIT Hackathon: Eric Swidler, Jerome Andres, Tenzin Chophen, and Venkadesh Eswaranandam
The software, hardware, and CAD design for each hackathon were created separately, following the specific rules of each event, with no reuse of previous work.
Hardware
- RFID-based patient identification – Scans patient wristbands to log patient ID.
- Power efficient audio & video syncing – Captures speech & patient behavior for every interaction throughout the day.
- Xiao ESP32-S3 smart badge – Microphone, camera, and Wi-Fi enables wireless transmission of real-time data.
- Custom 3D-printed housing – Encases ESP32, RFID reader, and patient wristbands for hospital use.
AI-powered software
- Speech recognition (Deepgram API) – Converts nurse-patient dialogue into detailed, diatorized text.
- Vision analysis (Gemini 2.0 Flash) – Identifies helpful details about the patient and room to enhance notes.
- EHR structuring (Mistral AI) – Transforms raw conversation into Epic-compatible notes.
- Doctor’s review portal (Next.js/Vercel v0) – Physicians edit & approve AI-generated documentation.
- Patient summary app (Next.js/Vercel v0) – Provides multilingual, easy-to-read summaries of doctor's notes.
- Patient Q&A (Perplexity Sonar API & Postgres SQL) – Patients can receive reliable medication answers grounded in their doctor's notes and web data.
WingNote fully automates the documentation pipeline, improving workflow efficiency and patient care.