Dr. Shiwei Wang, CEO of Evomics
Evomix
AI and nuclear medicine
Some of the most innovative applications of artificial intelligence in healthcare today are in the field of nuclear medicine, and thanks to AI, nuclear medicine shows great potential for cancer treatment. About 20 million people are newly diagnosed with cancer and about 10 million die each year. That’s about 1 in 6 of all deaths. The problem, and therefore the opportunity, is vast.
One of the leading companies in this field is Evomics, based in Shanghai and Vienna. The company, which uses the same technology to develop both diagnostics and therapeutics, has ambitious plans for the EV101 compound, which he believes could be his $10 billion blockbuster. increase.
Nuclear medicine compared to radiotherapy
It’s easy to confuse nuclear medicine with radiotherapy, or radiotherapy. Radiation therapists direct radiation from an external source to tissue to remove or shrink cancer cells. In contrast, nuclear medicine injects radioactive molecules into the bloodstream where they act as drugs. Radioactive molecules are known as ‘radioligands’ (from the Latin ‘ligare’ for binding) and can perform a diagnostic or therapeutic function.
Once in the body, the molecule “recognizes” proteins expressed by tumor cells and binds to them. This causes the radioligand to decay, releasing a positron, the antimatter of the electron. When a positron encounters an electron, they annihilate each other, creating a pair of high-energy photons. This is detected by a device called a positron electron tomography (PET) scanner. This is the diagnostic mode of nuclear medicine.
In therapeutic mode, the radioligand irradiates bound tumor cells, causing them to die. Radioactivity is limited, so several treatments are usually needed to tackle cancer.
Theranostics
Evomics uses the same compound for both diagnostic and therapeutic applications, but at different dosages. EV101 is therapeutic application and EV201 is diagnostic version. Combined they are known as “Theranostic” (treatment + diagnosis).
The protein targeted by the Evomics radioligand is known as fibroblast activation protein (FAP). Fibroblasts are long, thin cells that normally help create collagen-containing tissue structures, but when dysfunctional they replicate uncontrollably, which leads to cancer.FAPs targeted by EV101 and EV201 are associated with nearly all forms of cancer.
“Don’t be afraid to fail
Evomics CEO Dr. Shiwei Wang co-founded the company with his twin brother Shifeng and two Vienna-based professors of nuclear medicine, Dr. Li Xiang and Dr. Marcus Hacker. Prior to starting Evomics, he spent 10 years investing in biotech start-ups while working in corporate development in the pharmaceutical industry. His employer was not a financial investor, but he aimed to ensure access and experience in important emerging technologies. This gave Dr. Wang privileged insight into the most promising new technologies. He decided that the combination of nuclear medicine and his AI would bring enormous benefits to patients and spur the rise of major new businesses.
Dr. Wang launched the new company with the support of his former employer, but Evomics remains low profile at the moment, so its funding sources are currently undisclosed.
During his corporate development work, Dr. Wang worked with the venture capital community centered on Silicon Valley’s Sand Hill Road, and one of the things he learned from them was that good technology and good products are successful. was not enough to guarantee On top of that, business founders need extraordinary motivation. Starting and growing a business is not for the faint of heart. Resilience and confidence are essential for anyone looking to do things in new ways. Dr. Wang told me that he at Evomics believes his greatest asset as CEO is “not being afraid to fail.” Interestingly, he thinks part of his confidence comes from being a twin.
Application of AI
Deep learning AI systems are central to Evomics’ philosophy and integral to its work. They optimize the planning and preparation of medical interventions, analyze the images produced, and automate and speed up the generation of reports that can be used by clinicians. By making all stages of the process more efficient, AI can reduce the number of scans required and the radiation dose a patient receives. As you know, speed is essential in cancer treatment. So, by speeding up the analysis process and delivering valuable insights quickly to clinicians, diagnoses can be made faster, more accurately, and save lives.
AI algorithms can also provide valuable second opinions when clinicians disagree about interpreting images or data.
In the future, Evomics is also working on deploying large-scale language models, Transformer AI models that have captivated the world in the form of ChatGPT and GPT-4, analyzing and noting findings for other technologies.
Heart disease
Gamma rays emitted from radioligands may have benefits beyond cancer. Evomics also develops software for SPECT imaging. SPECT stands for single photon emission computed tomography and can detect the presence of blockages in arteries that can lead to heart disease.
hurdles and opportunities
Significant challenges remain to the successful application of AI in nuclear medicine. Models are more effective when trained on large data sets. Also, the amount of data currently available is limited and rarely consistently organized and labeled. Most deep learning models are trained on 2D images, but nuclear medicine scans are 3D. And most importantly, patients, clinicians, and regulators need to trust AI before it is widely deployed. Unsurprisingly, many still view AI systems as inexplicable black boxes.
These hurdles have been eased to some extent by the Covid-19 pandemic, which has greatly facilitated the use of AI techniques in medical imaging, including nuclear medicine.
In 2016, AI researcher Jeff Hinton, known as the father of deep learning AI, famously said: You’re already over the edge, but you’re not looking down yet. There is no ground under it. He was wrong: even today, seven years later, human radiologists are still in demand, and indeed many countries are short of radiologists. Even if AI won’t take your job anytime soon, humans who know how to work with it probably will.
The big picture is that AI makes more and better technology available to all of us.
Follow me please twitter or LinkedIn. check out My website or some of my other works can be found here.
