Graylight Imaging Data Scientists top ranked in RSNA-ASNR-MICCAI 2021 Brain Tumor Segmentation (BraTS) Challenge

images of automated segmentation performed by an algorithm top ranked in BRATS 2021 challenge

By: Anna Choma


Although we are Graylight Imaging today, before our rebranding, we participated as Future Processing Healthcare in the 10th Brain Tumor Segmentation 2021 (BraTS) challenge jointly organized by the Radiological Society of North America (RSNA), the American Society of Neuroradiology (ASNR), and the Medical Image Computing and Computer Assisted Interventions (MICCAI) society. Our team took sixth place at this prestigious challenge. The results of the competition were presented at the American Congress of Radiology 2021.

Brain Tumor Segmentation Challenge at a glance

The competition attracted 2,600 teams from around the world who downloaded the Brain Tumor Segmentation 2021 (BraTS) challenge database provided by the organizers to prepare and submit their own artificial intelligence (AI) model based on it. The task was to build an algorithm that automatically segments a glioma on magnetic resonance imaging into three regions: oedema, active tumor cells (which enhance when contrast is administered), and necrosis. This challenge was tackled by 1300 teams.

Our BraTS history

We have been actively participating in the Brain Tumor Segmentation Challenge since 2017. “Sixth place is a big achievement for us. We have proven that our skills are world-class. Observing previous editions, you could see that often teams that were ranked high after the validation phase were placed lower in the final results. We managed to maintain our position at the top. I would also like to add that the result is more an outcome of using our experience from other projects that we carry out, rather than direct involvement in the competition” – says Krzysztof Kotowski, leader of the competition team, algorithm and machine learning specialist at Graylight Imaging (former Future Processing Healthcare).

After the validation phase, our team was ranked 5th.

The testing phase included more research into the effectiveness of the developed solution than the validation phase. “We treat this achievement as a great success. Let us remember that the organisers of the initiative are the most important institutions and corporations in the world related to the use of artificial intelligence in medical imaging, and the competition itself is recognised also outside the world of radiology”. – says Marek Pitura, head of project management at GLI.

Find the full leaderboard of BraTS participants here.

BraTS 2021 – ask us anything

If you’re interested in what approaches to AI model for tumor segmentation they’ve developed, what frameworks they’ve used, what ups and downs they’ve experienced, or anything else related to the Brain Tumor Segmentation 2021 (BraTS) challenge and our algorithm – just ask.

We invite you to take part in our AMA! Ups and Downs Brain Tumor Segmentation

AMA (short of “ask me anything”) is a type of interview in which the interviewee is open to questions from the public.

Join Graylight Imaging BraTS 2021 team AMA. You ask, we answer!

We are going to collect questions from you until 11th February:

The answers will be published on our blog and LinkedIn https://www.linkedin.com/company/graylight-imaging on 16th February.

Timeline:

Send questions: 4-11.02.2022

Answers published: 16.02.2022

See the previous post by Szymon Janota: Graylight Imaging – a new brand of Future Processing Healthcare