MEDICA 2019 – DataArt to Present Machine Learning Powered X-ray Stereogrammetry as an Alternative to CT


DataArt the global technology consultancy that designs, develops and supports unique software solutions,  will be speaking at MEDICA 2019 on the topic of machine learning-powered X-ray Stereogrammetry and demonstrating its SkinCareAI prototype app – which shows how AI can be used to support early cancer diagnosis. DataArt will also be demonstrating an AI-powered tool for flagging early warning signs of skin cancer. 

Talk: X-ray stereogrammetry (RSA) powered by machine learning as an alternative to CT

RSA powered by ML is potentially a powerful new tool for better diagnostics and a fascinating area of research that has already caused waves in the industry. DataArt’s Alexander Khmil will outline the huge promise and also address the limitations. While AI and ML will never entirely replace doctors, is this potential new tool going to change things?  And how?

Speaker:        Alexander Khmil,  Senior Solution Consultant at DataArt

Time:             12:30 – 12:45 pm

Date:              20 Nov 2019

Venue:           Hall 13, Stand D45

Session:         AI, BIG DATA and IoT – Unleashing the power of data.

Area:              Medica Connected Healthcare Forum 


SkinCareAI App Demo

With well over 200,000 new cases per year, skin cancer is one of the most common cancers in Europe.  Early detection is essential. To help counter this problem the Healthcare and Life Sciences Practice of DataArt has developed a prototype app which analyses skin types in order to alert users that they need to visit a doctor for further investigation. DataArt is demonstrating the prototype to the trade for the first time at MEDICA.

Venue:           Hall 13, B32


RS and ML

More than a hundred years ago, the German university professor Wilhelm Conrad Roentgen discovered the ray (X-ray), which could pass through most solid objects and cast a shadow onto a piece of paper or a wall. This discovery transformed the field of medical diagnostics and triggered further research. In 1961, the American physician and neurologist William H. Oldendorf developed the basis for computerized tomography (CT) and in 1973 this technique was applied to clinical diagnosis by Godfrey Hounsfield.

Today, we face a new twist in the tale, with research in to Stereophotogrammetry combines with mathematical magic. One such technology is represented by a combination of Roentgen stereophotogrammetry and conventional photogrammetry with some advanced math employed.  The use of ML looks set to transform the field.


Early detection of melanomas

SkinCareAI was developed by DataArt expert Andrey Sorokin for the International Skin Imaging Collaboration (ISIC) Challenge. Based on worldwide data and the latest findings in Machine Learning (ML) technology, the App uses ML algorithms for early detection of melanoma.

“SkinCareAI” brings together the know-how of the world’s best dermatologists in a melanoma diagnosis tool that can be accessed via any smartphone. In principle, it is as if their suspicious-looking skin lesion is being examined by thousands of doctors based on their combined experience and expertise. This is exactly what the prototype of our smartphone app offers,” explains Andrey Sorokin.

With the app, DataArt demonstrates to the healthcare sector and the medical community how machine learning can be used to analyze medical images in an AI diagnostic process. However, the app does not replace a doctor’s diagnosis, but supports him in recognizing the disease at an early stage.


Simple user guidance

The user guidance of the app is very simple. Users either upload photos of their skin lesions or use the app to take them directly. Within seconds, the images are analyzed by the app. This is done by comparing thousands of images, which the app uses to classify the lesion and identify its morphological properties. Based on this, the app then suggests the next steps to the user.

“All functions are fast and user-friendly,” continues Sorokin. “However, we do not recommend relying solely on the skin cancer detection app.

Rather, the app is designed as an encouragement to check yourself out and see a doctor if necessary. At the same time, the application is intended to help the medical community by bringing collective intelligence to individual practitioners.

“We are very confident that the app, if developed with a medical partner, will save lives through early detection,” Sorokin says.


Innovative healthcare apps

In addition to the SkinCareAI App, the experts from DataArt present further Proofs of Concept at MEDICA. DataArt has already made a name for itself with innovative app developments in the healthcare sector. The Telehealth & Remote Monitoring platform, a modern form of telemedicine, also uses artificial intelligence for medical support. KidPRO, a pilot solution for patient engagement in paediatric medicine, is also an award-winning PoC in the field of digital healthcare. Both apps are represented at the MEDICA App Competition.