Press "Enter" to skip to content

How Two Scientists Are Putting Together Artificial Intelligence and Radiology Creating a Multidisciplinary Community to Fight Brain Cancer

Nowadays, artificial intelligence permeates every aspect of our life, including healthcare. As one of the consequences, the foremost authorities in medicine, biomedical engineering, and artificial intelligence meet annually to share the most recent developments. It is the MICCAI conference, which stands for Medical Image Computing and Computer Assisted Interventions society. The event will take place in Singapore this year.

Dr. Alessandro Crimi (Brain and More Lab) and Spyridon Bakas (University of Pennsylvania) curated  the Brain lesion workshop under MICCAI, this year with the help of young researchers as Dr. Ujjwal Baid, Monika Pytlarz and Sylwia Malec (2 PhD students of Sano, Center for Computational Medicine in Poland). The MICCAI event focuses on brain tumors, multiple sclerosis, traumatic brain injuries, and strokes, and includes  the BRATS (multimodal BRAin Tumor Segmentation) challenge.

The objective was to combine scientific findings related to different brain lesions that, from the perspective of medical imaging, sharing similarities, aiming at gaining fresh perspectives. The event also holds competitions to achieve the best brain tumor segmentation in  magnetic resonance imaging (the BRATS competition). This is relevant as, It is crucial to correctly segment brain tumors since doing so can help doctors save a lot of time, preventing doctor disagreements, and possibly even link  patient outcomes through machine learning.

Since the competition’s inception seven years ago, the event and related community has grown, and now it even includes backing from the Radiological Society of North America (RSNA), and  large hardware firms like Nvidia and Intel.  Due to recent restrictions between the US, China, and Russia and fear of recession, Nvidia and Intel shares have lost over half of their value in 2022. However, they continue to dominate the market for fundamental AI computational units.

AI-driven medical imaging and diagnosis have advanced dramatically in recent years, and it is thought that this competition has contributed to some aspects in this context (as the brain tumor segmentation in MRI). What began as a nerdy competition will eventually benefit patients. We urge you to join this year’s event if you are interested, or read the consensus paper about the past editions (https://arxiv.org/pdf/1811.02629.pdf )

The main organizers are also interested in global health efforts. Dr. Bakas is working with CAMERA, the Consortium for the Advancement of MRI Education and Research in Africa, to establish a continuous network of resources and knowledge that will help Africa utilize MRI to meet its healthcare needs. Indeed, a case competition for global health imaging was started at the University of Pennsylvania, with brain tumor images and other evidence supporting the program.

Dr. Crimi has been lecturing at the African Institute for Mathematical Sciences in Ghana for nine years. He also directed  projects using portable ultrasound machines in some of the country’s rural areas.  The next effort of him and his team are now about  federated learning, even more cutting edge machine learning,  and data from Rwandan EEG scans.

There is a hope that these computational tools, together with therapies like immunotherapies, surgery, and other approaches, may one day help cure brain cancer. Even if they started as  some nerdy, trendy scientific gathering about AI.