The Smart Atrium
The heart is the organ in the human body that pumps blood throughout the body. It's the primary organ of the circulatory system. The heart contains four main sections (chambers) made of muscle and powered by electrical impulses. As it's so pivotal many diseases do arise in it and on
2025-06-28 16:29:43 - Adil Khan
The Smart Atrium
Project Area of Specialization Computer ScienceProject SummaryThe heart is the organ in the human body that pumps blood throughout the body. It's the primary organ of the circulatory system. The heart contains four main sections (chambers) made of muscle and powered by electrical impulses. As it's so pivotal many diseases do arise in it and one of the most common heart diseases is Atrial fibrillation or A-fib. It refers to frequent irregular heartbeats, which is called Arrhythmia. A-Fib can lead to Blood Clots, Stroke, and Heart Attacks. The manual process to precisely diagnose and treat atrial fibrillation requires the examination of the human atria and that procedure is tedious, invasive, and an open to error process.To overcome this barrier we will develop a desktop application ‘The Smart Atrium’ for practical use by performing Left Atrium Segmentation using Deep Learning for Medical Image Analysis, which will be available only to licensed Medical Practitioners, Clinics, and Hospitals.
Project ObjectivesThe goal is to replace the manual method of analyzing the left atrium with an automated model based on Deep Learning to provide more significant and reliable results. The Desktop application will be developed for the practical use of our model which will be available only to licensed Medical Practitioners, Clinics, and Hospitals.
Project Implementation MethodAs it is a deep learning based project, firstly after data collection and cleaning we will train our model with the help of Algorithm which will comprise Multi-Stage CNN. After successful training we will test our model for validation and accuracy. And then upon getting desirable results we will deploy our model in a desktop application which will work as described below in flow chart.
We are automatizing a manual procedure for analyzing Left Atrium and the diseases it entails such as Atrial fibrillation and other related severe heart conditions. A manual procedure that has various flaws such as:
- Tedious: Imaging methods such as ECG, EKG, chest X-RAY, Stress test, and Blood tests are conducted per patient which takes an extensive amount of time.
- Error-Prone Process: The initial tests are sometimes non-conclusive even when performed under expert supervision.
- Invasive: Sometimes a test called Cardiac Catheterization is conducted, in this test A thin hollow tube called a catheter is threaded from a blood vessel in the arm, groin, or neck to the heart. It takes 2-3 hours.
By its automation we will be able to overcome many barriers. This transition would save time, money, and most importantly human lives.
Technical Details of Final DeliverableTo use this desktop application, a user will sign in with his/her credentials, after logging in successfully user will upload a scanned MRI scan then our deep learning model will analyze the MRI scan and will provide the desired output which will be a clear vision for doctors about what to do next for his/her patient.
Final Deliverable of the Project Software SystemCore Industry MedicalOther IndustriesCore Technology Artificial Intelligence(AI)Other TechnologiesSustainable Development Goals Good Health and Well-Being for PeopleRequired Resources| Item Name | Type | No. of Units | Per Unit Cost (in Rs) | Total (in Rs) |
|---|---|---|---|---|
| Total in (Rs) | 38000 | |||
| Graphic Card GTX 1050 Ti | Equipment | 1 | 30000 | 30000 |
| Power Supply 300w | Equipment | 1 | 3000 | 3000 |
| Printing of final FYP report in Journal and document | Miscellaneous | 50 | 100 | 5000 |