📄 Project Overview

<aside> <img src="/icons/cursor-click_gray.svg" alt="/icons/cursor-click_gray.svg" width="40px" /> TiSepX is an X-ray quantitative analysis and tissue separation solution. It provides numerical information of the lungs and lesions with the augmented images of a single X-ray image. It is used for quantification and progress monitoring of tuberculosis and COVID-19 as well as lung volumetry. The application of TiSepX is expected to continue.

</aside>

🎯 Product Goals

<aside> 🎯 1) Development of a windows-based client application

</aside>

<aside> 🎯 2) Incorporation of a novel user experience (UX) design

</aside>

<aside> 🎯 3) Simple and intuitive user interface for running deep learning models

</aside>

<aside> 🎯 4) Optimization of deep Learning models for faster inference

</aside>

<aside> 🎯 5) Continuous monitoring and validation

</aside>

📥 My Contributions

<aside> <img src="/icons/groups_green.svg" alt="/icons/groups_green.svg" width="40px" /> Led a team of 3 junior developers in creating a robust C++ based API to integrate deep learning models with the system software.

</aside>

<aside> 🏎️ Employed TensorRT and ONNX to reduce the inference time of the model from 48 seconds to 18 seconds (3 times speed up), ensuring improved performance.

</aside>

<aside> 🛡️ Developed an innovative encryption strategy to protect the trained deep learning models, securing intellectual property of the company.

</aside>

<aside> 🖥️ Incorporated C++ multi-thread for better utilization of resource

</aside>

<aside> 🛣️ Contributed in the development of product roadmap design

</aside>

<aside> 🏗️ Innovative UX/UI Design for TiSepX Client App

</aside>

<aside> 🏁 In collaboration with the IT team, deployment of the software, testing, verification, and post-deployment refinement to ensure a reliable software.

</aside>

🛠️ Tech Stacks Used

<aside> <img src="https://prod-files-secure.s3.us-west-2.amazonaws.com/cb4bc682-8064-42da-8247-5f27942c5d27/32e2b717-42da-4f29-90d1-4be52bff867e/C.png" alt="https://prod-files-secure.s3.us-west-2.amazonaws.com/cb4bc682-8064-42da-8247-5f27942c5d27/32e2b717-42da-4f29-90d1-4be52bff867e/C.png" width="40px" /> C/C++

</aside>

<aside> <img src="https://prod-files-secure.s3.us-west-2.amazonaws.com/cb4bc682-8064-42da-8247-5f27942c5d27/9acd32ee-f449-47d5-a3d3-7d4f551eb43a/qt.png" alt="https://prod-files-secure.s3.us-west-2.amazonaws.com/cb4bc682-8064-42da-8247-5f27942c5d27/9acd32ee-f449-47d5-a3d3-7d4f551eb43a/qt.png" width="40px" /> QT framework

</aside>

<aside> <img src="https://prod-files-secure.s3.us-west-2.amazonaws.com/cb4bc682-8064-42da-8247-5f27942c5d27/6551a5ba-e238-42c4-93e1-d9a80e75b741/python.png" alt="https://prod-files-secure.s3.us-west-2.amazonaws.com/cb4bc682-8064-42da-8247-5f27942c5d27/6551a5ba-e238-42c4-93e1-d9a80e75b741/python.png" width="40px" /> Python

</aside>

<aside> <img src="https://prod-files-secure.s3.us-west-2.amazonaws.com/cb4bc682-8064-42da-8247-5f27942c5d27/6240f116-b260-4d80-a4c0-7f0578226eb4/docker.png" alt="https://prod-files-secure.s3.us-west-2.amazonaws.com/cb4bc682-8064-42da-8247-5f27942c5d27/6240f116-b260-4d80-a4c0-7f0578226eb4/docker.png" width="40px" /> Docker

</aside>

<aside> <img src="https://prod-files-secure.s3.us-west-2.amazonaws.com/cb4bc682-8064-42da-8247-5f27942c5d27/fe96e258-89e0-4938-b211-9b1fa4c91cae/fastapi.svg" alt="https://prod-files-secure.s3.us-west-2.amazonaws.com/cb4bc682-8064-42da-8247-5f27942c5d27/fe96e258-89e0-4938-b211-9b1fa4c91cae/fastapi.svg" width="40px" /> Fast API

</aside>

<aside> <img src="https://prod-files-secure.s3.us-west-2.amazonaws.com/cb4bc682-8064-42da-8247-5f27942c5d27/435bd915-7351-43bf-b1c0-7bb37e27d23f/PyTorch_logo_icon.svg.png" alt="https://prod-files-secure.s3.us-west-2.amazonaws.com/cb4bc682-8064-42da-8247-5f27942c5d27/435bd915-7351-43bf-b1c0-7bb37e27d23f/PyTorch_logo_icon.svg.png" width="40px" /> LibTorch & Torchscript

</aside>

<aside> <img src="https://prod-files-secure.s3.us-west-2.amazonaws.com/cb4bc682-8064-42da-8247-5f27942c5d27/34f66807-94c2-408b-a902-54208874ee0f/version-control-tortoise-svn.png" alt="https://prod-files-secure.s3.us-west-2.amazonaws.com/cb4bc682-8064-42da-8247-5f27942c5d27/34f66807-94c2-408b-a902-54208874ee0f/version-control-tortoise-svn.png" width="40px" /> Tortoise SVN

</aside>

<aside> <img src="https://prod-files-secure.s3.us-west-2.amazonaws.com/cb4bc682-8064-42da-8247-5f27942c5d27/5e0da7cd-de65-4980-ac72-f21daf00e6a4/gitlab.png" alt="https://prod-files-secure.s3.us-west-2.amazonaws.com/cb4bc682-8064-42da-8247-5f27942c5d27/5e0da7cd-de65-4980-ac72-f21daf00e6a4/gitlab.png" width="40px" /> Gitlab

</aside>

<aside> <img src="https://prod-files-secure.s3.us-west-2.amazonaws.com/cb4bc682-8064-42da-8247-5f27942c5d27/8b22e90c-6534-4563-9c2f-c10d89026041/AWS_Simple_Icons_AWS_Cloud.svg.png" alt="https://prod-files-secure.s3.us-west-2.amazonaws.com/cb4bc682-8064-42da-8247-5f27942c5d27/8b22e90c-6534-4563-9c2f-c10d89026041/AWS_Simple_Icons_AWS_Cloud.svg.png" width="40px" /> AWS SagkeMaker, IAM, EC2 and Lambda Function

</aside>

<aside> <img src="https://prod-files-secure.s3.us-west-2.amazonaws.com/cb4bc682-8064-42da-8247-5f27942c5d27/2a3854ad-4823-48ae-a08e-18af6513219d/redmine.png" alt="https://prod-files-secure.s3.us-west-2.amazonaws.com/cb4bc682-8064-42da-8247-5f27942c5d27/2a3854ad-4823-48ae-a08e-18af6513219d/redmine.png" width="40px" /> RedMine

</aside>

<aside> 🎯 Data science

</aside>

<aside> <img src="https://prod-files-secure.s3.us-west-2.amazonaws.com/cb4bc682-8064-42da-8247-5f27942c5d27/89076378-e166-4668-baa0-a9b336f920b0/machine_learning.png" alt="https://prod-files-secure.s3.us-west-2.amazonaws.com/cb4bc682-8064-42da-8247-5f27942c5d27/89076378-e166-4668-baa0-a9b336f920b0/machine_learning.png" width="40px" /> Machine Learning

</aside>

<aside> <img src="https://prod-files-secure.s3.us-west-2.amazonaws.com/cb4bc682-8064-42da-8247-5f27942c5d27/852d3ad9-75b9-4a94-9858-add1a882171c/deep-learning-1524275-1290822.png" alt="https://prod-files-secure.s3.us-west-2.amazonaws.com/cb4bc682-8064-42da-8247-5f27942c5d27/852d3ad9-75b9-4a94-9858-add1a882171c/deep-learning-1524275-1290822.png" width="40px" /> Deep Learning

</aside>

🖌️API Design Overview

ITC (Information Technology Center) is the department responsible for software post-deployment and bug-fixes.
The data is transferred to the API from ITC, processed on TATA API and the desired result such as segmentation, classification or tissue separation is returned to the the ITC.
Inspired by AWS bucket concept, I designed C++ based bucket style data structure for efficient exchange of data between TATA API and Medip software.

📝 Introduction

TiSepX is an X-ray quantitative analysis and tissue separation solution. It provides numerical information of the lungs and lesions with the augmented images of a single X-ray image. It is used for quantification and progress monitoring of tuberculosis and COVID-19 as well as lung volumetry. The application of TiSepX is expected to continue.

🌏 International Relation

Beside my research and development responsibilities, I was also involved in expanding the global presence of Medical IP by building partnership with various international organisations.
On behalf Medical IP, I attended international meetings and conferences to discuss the usability and implementation feasibility of Medical IP products.
I played a key role in building the first partnership between Medical IP company and NVIDIA, which led to the deployment of Medical IP’s software onto NVIDIA’s marketplace.
I led the development of the first partnership between Medical IP and Amazon, which led to distribution of Medical IP’s software on Amazon web service.

🖇️ Methodology

Dataset
- Data from Seoul National University Hospital
  - A pair of X-ray and CT scans
  - Single Input to AI model
  - Multi Output from the AI model
  - Histogram normalization
Deep Learning Algorithm
- Pixel2Pixel
Applications
- Lung volumetric - measurement of the lung capacity (volume) from single X-ray image
- Aorta aneurysm - It can be used to measure the existence of aneurysm from X-ray image
- Lung vessel - It can be used to assess the quality of pulmonary vessel from X-ray images.
- Lung vessel digital twin - It can be used to create digital twin of lung vessels to assist in guided surgery using x-ray only image.
- Osteoprosis analysis - Using L1 bone it can accurately analyze osteoprosis in the body
- Prognosis analysis - The separated lung can be used to predict mortality rate. This segmented lung can be further used to improve the performance of other x-ray related deep learning models.
- Educational purpose - The tool can be used for educational purposes such aiding medical students understanding lung, heart and body anatomy from x-ray image.
⚒️ TiSepX Development

Three types of view
1. Monochromic View (Grayscale)
2. Cinematic View (Colorful)
3. Contour View (Drawings)
⛓️ API Features
- Multithreading
- API status monitoring
- Error and exception handling
- Best practices of C++ data structure such as vector, tuple and pointers.
- Extensibility and modularity
- Easy debugging and enhancement