CerebraAI on the cusp of saving the lives of stroke patientsCerebraAI on the cusp of saving the lives of stroke patients

Meet Doszhan Zhussupov, a 37-year-old serial entrepreneur who is redefining emergency stroke care with an AI-powered platform, CerebraAI. Zhussupov won the Innov8 Start-up Competition recently held at the Global Health Exhibition in Saudi Arabia after battling several contenders from across the globe, proving that CerebraAI holds a promising solution for stroke patients.

CerebraAI is designed for emergency departments in specialised stroke units and community hospitals, exclusively utilising Non-Contrast Computer Tomography (NCCT) scan readings. To date, more than 100,000 patients have been analysed through the CerebraAI program, with over 500 doctors using artificial intelligence, and 46 medical institutions that trust and use the capabilities of CerebraAI.

The journey of new beginnings

While Zhussupov is predominantly based in California, CerebraAI’s origins lie in Kazakhstan, where it was founded in 2018 at Almaty AI Laboratory. In 2019, he set his sights on AI research and development in three focus areas — pulmonological diseases, ophthalmology, and neuroradiology — which led to the rise of three small research groups consisting of machine learning engineers, mathematicians, and software. Soon afterward, an evaluation revealed that neuroradiology showed the most promise, with specialists seeking neuroradiology AI to detect neuroradiology pathologies, notably acute ones.

For Zhussupov, this also became a project close to his heart, as he revealed that his father passed away from a stroke. Trial and error followed but the team at CerebraAI was determined to create a sustainable solution that could change the lives of patients.

“Frankly, the first version of CerebraAI in 2019 was useless but we did not know it at the time. We then realised the problem was in the way we organised the R&D process. We had most of our competence in-house like software engineering, data science, design, and maths, but were missing medical expertise,” he said.

Zhussupov understood the importance of an in-house expert and key physician as part of the development journey and recruited one of Kazakhstan’s top neuroradiologists to the team.

“We started crafting the second version of CerebraAI, and not surprisingly, it worked very well, although we spent almost two years on R&D. The product we released in 2021 was the real CerebraAI that can triage between haemorrhage (ICH) and ischemic stroke based on Non-Contrast CT (NCCT) scan with an accuracy of 99 per cent and within two to three minutes,” he added.

In the case of an ICH stroke, CerebraAI shows which class of that stroke occurred and the calculation of blood volume is made automatically with very accurate precision. And in the case of an ischemic stroke, CerebraAI captures acute and subacute types and tells how many anatomic parts of a brain are damaged at the time Non-Contrast CT is made.

How CerebraAI works

Zhussupov explained that CerebraAI enhances NCCT efficiency, making it as sensitive as MRI and crucial for early stroke detection and treatment within the critical window. This overcomes the limitations of MRI availability and the time-sensitive nature of strokes.

According to the standard of care, NCCT is used mainly to detect haemorrhage, if it is not seen patients must be referred to perform MRI or Contrast CT (CTA/CTP) to detect an ischemic stroke.

“But the problem is that in most cases MRI is not available, and Contrast CT is not performed as well. There is also a time issue related to stroke as every minute stroke patient loses around two million brain cells and neurological deficit becomes permanent after six to eight hours. Early treatment (thrombolysis or thrombectomy) must be provided within this window,” he said.

CerebraAI is solely based on NCCT method, which is available in 94 per cent of emergency departments and it enhances NCCT to be as sensitive as MRI at detecting acute ischemic stroke not just at haemorrhage.

“That is why our technology is so important for humanity as it makes existing hardware more efficient without any capital investments or extensive personnel training. Our technology is easily deployable as any hospital with an internet connection could gain access at once,” he said.

Tackling challenges one step at a time

The path to success was long and hard, Zhussupov recounts, but each obstacle defined the CerebraAI of today. Access to data, notably anonymised brain CT images, was the first hurdle, which cost him a year.

“One year for a start-up is like three to four years for conventional companies. Obstacles like these can threaten the lifeline of any MedTech start-up,” he added.

Access to capital was the second challenge. As one of the first start-ups in deeptech AI for healthcare in Kazakhstan, investors were not too eager to explore their options back in 2018. CerebraAI was resilient and succeeded in raising capital not just from Kazakhstan and Uzbekistan venture capitals (VCs) but also from Singapore VCs.

Regulatory challenges are an obstacle CerebraAI still faces to date, as governments across the globe struggle to adopt and integrate artificial intelligence into their legislation systems. “The ability to pilot your AI product is a hard task but commercialisation becomes a mission impossible thing to do. We still have struggles with regulatory issues in each geography we operate in,” Zhussupov added.

The road ahead

Zhussupov has big goals in mind for CerebraAI in the coming year. His dream is to deploy CerebraAI in Makkah, Madinah, Jeddah, and Riyadh to support individuals attending Hajj, Umrah, and Ramadan.

“Secondly, we aim to penetrate the US market, leveraging our proven solution to enhance stroke care in the largest healthcare economy globally,” he concluded.

Final words...

Zhussupov believes that the health industry is at the cusp of smart changes. “The last decade brought us new ways of drug discovery and vaccines creation, electronic medical records, surgical robots, lots of biotools and medical devices, telemedicine, online prescription and e-pharmacies, PACS, and of course, AI. But some real shift is yet to come. The healthcare revolution, in my personal view, will be brought by a massive adoption of AI and AI-powered robots,” he concluded.

For inquiries or to learn more about Cerebra AI, contact Zhussupov at [email protected]

Back to Technology