# Od koła naukowego do spółki akcyjnej | Barbara Klaudel | TEDxPolitechnika Gdańska

Source: https://www.youtube.com/watch?v=mG8tbeLURjw
Recap page: https://rapidrecap.app/video/mG8tbeLURjw
Generated: 2026-01-26T17:41:37.869+00:00

---
## Quick Overview

The speaker recounts her journey from a high school student interested in AI to co-founding Medallion Technology, a company developing an AI platform named TITAN to assess the malignancy of kidney tumors, ultimately raising 400,000 PLN for the project.

**Key Points:**
- The speaker, Barbara Klaudel, was inspired by the topic of Artificial Intelligence, which became widely discussed around 2016, and she always wanted to do something she could build herself.
- The team created the TITAN (Technology In Tumor Analysis) system, an AI tool designed to assess the malignancy of kidney tumors, which is crucial for deciding between observation and surgery.
- The project received funding of 400,000 PLN, which allowed the team to develop the system, including creating a library of DICOM files for anonymization.
- The speaker contrasted her early aspiration of becoming a veterinarian with her current focus on AI, emphasizing that the AI passion was a natural step after learning about the technology.
- The team, composed of students and scientists from the Gdańsk University of Technology, collaborated with urologists from a local hospital to solve the problem of unnecessary surgeries due to uncertain tumor diagnoses.
- The developed system achieved 95% accuracy in recognizing tumors as malignant or benign, potentially reducing unnecessary surgical procedures, which previously accounted for 20% of cases.
- The speaker encourages the audience to pursue their own passions, referencing the movie *Fight Club* quote, "The question, Reynolds, what did you want to be?" to prompt self-reflection.

![Screenshot at 00:54: The speaker displays a document detailing the 'Obecnie stosowany proces diagnostyki i leczenia guzów nerek' \(Currently used process for diagnosis and treatment of kidney tumors\) juxtaposed with the proposed improved process, highlighting the need for their AI solution.](https://ss.rapidrecap.app/screens/mG8tbeLURjw/00-00-54.jpg)

**Context:** This TEDx talk, delivered by Barbara Klaudel at TEDxPolitechnika Gdańska, chronicles her path into developing artificial intelligence solutions for healthcare, specifically focusing on the creation of the TITAN system. The context revolves around addressing uncertainty in diagnosing kidney tumors, where current protocols often lead to unnecessary surgeries, and how her student project evolved into a funded startup, Medallion Technology.

## Detailed Analysis

Barbara Klaudel begins by sharing an anecdote from *Fight Club* to illustrate the importance of asking what one truly wants to do, noting that she always wanted to create something herself. Her interest was sparked around 2016 when AI gained mainstream attention. She initially considered veterinary medicine but found her passion in technology she could build, leading her to pursue AI. She discusses how the Gdańsk University of Technology was not a natural hub for AI development then, but that she and her team started creating their own path. They formed the student scientific circle 'Project Management' and later co-founded Medallion Technology. The core of their work involved creating the TITAN (Technology In Tumor Analysis) system, an AI tool that assesses the malignancy of kidney tumors based on CT scans, aiming to reduce unnecessary surgeries. She contrasts the existing clinical workflow, which involves uncertainty (e.g., Hounsfield Units), with their improved algorithm which achieved 95% accuracy. The project secured 400,000 PLN in funding, which helped them build their own library of DICOM files for data anonymization. The ultimate goal is to provide a revolution in Polish healthcare by offering a tool that helps doctors make better decisions regarding intervention versus observation for kidney masses.

### Inspiration and Early Days

- Recounts the *Fight Club* scene prompting self-reflection
- Always wanted to build things herself
- AI discussion started gaining traction around 2016
- Gdańsk was not an AI hub then
- Formed the 'Project Management' student circle

### Project TITAN Development

- Co-founded Medallion Technology with peers interested in similar topics
- Collaborated with urologists from a local hospital
- Goal was to create a system that solves a real problem
- Developed an AI model to assess kidney tumor malignancy

### Technical and Financial Milestones

- Secured 400,000 PLN in funding
- Built their own library of anonymized DICOM files
- Model achieved 95% accuracy in differentiating benign/malignant tumors
- Aimed to reduce unnecessary surgeries (which accounted for 20% of cases)

### Conclusion and Future Outlook

- Emphasizes that technology, quantum physics, and game development are interesting, but the current focus is medicine
- Urges the audience to find their own passion and path
- Stresses that their success was due to self-investment and collaboration with experts

![Screenshot at 00:05: Title slide for TEDx Politechnika Gdańska event.](https://ss.rapidrecap.app/screens/mG8tbeLURjw/00-00-05.jpg)
![Screenshot at 00:12: Speaker on stage at the TEDx event, with the TEDx Politechnika Gdańska backdrop.](https://ss.rapidrecap.app/screens/mG8tbeLURjw/00-00-12.jpg)
![Screenshot at 00:54: A document showing the current and proposed diagnostic process for kidney tumors, highlighting the problem TITAN addresses.](https://ss.rapidrecap.app/screens/mG8tbeLURjw/00-00-54.jpg)
![Screenshot at 01:01: A meme slide used to illustrate the difference between public perception of AI \(fear-mongering\) and the speaker's practical work \(a cat misclassified as a dog\).](https://ss.rapidrecap.app/screens/mG8tbeLURjw/00-01-01.jpg)
![Screenshot at 02:00: A composite image showing the team working: data analysts in surgical gear on the left, and team members in a meeting on the right, with a whiteboard titled 'What's in my day?'](https://ss.rapidrecap.app/screens/mG8tbeLURjw/00-02-00.jpg)