51 years ago, the Apollo 13 was launched into space. Shortly thereafter, a powerful explosion shook the spacecraft. As their precious oxygen supplies entered the atmosphere, the three astronauts on board were in serious danger.
While the clock was running, NASA’s Mission Control team in Houston worked endlessly. How can they identify and solve the challenges of a seemingly insurmountable material object floating in space, 200,000 miles away?
Happily, NASA had 15 simulators that trained astronauts to perform the task. These high-performance analysts and their computer systems are ahead of “digital twins.” They were used to replicate the damaged spacecraft and played a key role in the successful rescue of the crew.
Today, digital twins of the human body are the latest trend in life science. Pharmaceutical companies are testing this technology to transform drug development and patient care.
What are digital twins?
A digital twin is a visual representation of a real world person, body structure, or process. For example, the digital twin of an aircraft engine is a precise copy of the machine, powered by artificial intelligence (AI). Stream data collected from sensors in search engines is transmitted to digital twins in real time. This allows aircraft engineers to monitor engine performance and predict when it will fail.
Now, imagine building a real digital twin for a real person to replicate how they will behave and respond to specific situations. You can track their health, diagnose diseases, and plan preventive treatments.
While the full range of human entertainment in the world is impossible (thankfully!), Even partial repetition offers great benefits today.
Why are digital twins so important to health care? We can use them to test new drugs to find out the safety of drugs and their effectiveness. If we had access to this technology by 2020, the COVID-19 drugs would have arrived months ago and saved the lives of millions.
Why clinical trials need some disruption
Today, four key challenges plague clinical trials and make drug development incomplete, slow, unexpected, and unsafe.
1. They are not accurate representations of the real world
By design, temptations can only produce a small representation of many and diverse people in the real world. For billions of people around the world, how each of them would react to a drug based on their unique physical and mental condition would never be repeated in a clinical trial.
2. Few tests require patients to be needed at a time
Some clinical trials have not yielded results because researchers have not been able to find suitable patients. Pharma companies are striving to enter the prescribed number and type of patients to meet the complex definitions of the clinical trial design. In fact, recruitment challenges delay 80 percent of all tests.
3. Not all patients are treated with a new trial drug
In most trials, the drug itself is exchanged with a placebo or a topical drug for up to half of patients. This helps to compare how patients with the disease respond when a test drug is given. This means that at least some patients enrolling for tests, hoping to get new treatments, are not getting it.
4. Not all test drugs work as safely as intended
By design, clinical trials test the safety and efficacy of experimental drugs. In addition to testing and control measures to improve drug safety safety, trials have serious side effects, including death. This prevents many patients from registering for clinical trials, and those who join do so amid safety issues.
Digital twins can face these challenges:
Covering: Digital twins can mimic a variety of patient characteristics, providing a representative view of the effects of the drug on the general population.
Speed: AI can simplify trial design by providing visibility to patient availability through a variety of implants and extraction methods.
Predictability: With digital twins predicting patient response, there will be no need for placebos or dummy drugs. Therefore, every patient in the trial can be confident of receiving a new treatment.
Safety: By reducing the number of patients who need real-world testing, digital twins can reduce the harmful side effects of early drugs.
However, the big question is whether digital twins are ready to carry these burdens in the availability of medical care today.
Fact or myth: The status of digital twins in health care
We are in the early days of using digital twins in the science of life. Today, pilots use simple twins to mimic the cellular and cellular functions of the human body, rather than to imitate the entire patient’s response to medical tests.
Charles Fisher, CEO of Unlearn, states: “We are not yet in a position to mimic the actual chemical composition of man. “There is a lot of biology that we do not yet understand, and there is no data. Therefore, we are not working to predict how patients will respond to new therapies, ”he adds.
Many clinical trials divide patients into two sets: the treatment arm receives the test drug, and the control arm receives the placebo or dummy drug. The control arm helps determine how patients with the disease respond if they do not receive new treatment.
It’s the control arm when pharmaceutical companies first try to convert using digital twins. “We can collect patient data from tens of thousands of people to see how they respond to treatment. This turns the challenge into a common machine learning problem where you have a history of learning from it, ”explains Fisher. “We are trying to predict what will happen to the new patient receiving the existing treatment.”
In recent years, the proposed use of AI for patient care algorithms has been completed. As the practice accelerates during the epidemic, the US Food and Drug Administration (FDA) is reportedly developing regulatory frameworks. “We are currently looking at graduation procedures from the FDA and the European Medicines Agency (EMA) to ensure we are compiled under the current guidelines,” Fisher said.
What diseases can begin to detect the use of digital twins? The potential is high in medical facilities where we have a trial of high-quality, high-quality clinics and real-world data. For example, Unlearn works with emotional problems such as Alzheimer’s disease and multiple sclerosis.
Three roadblocks that can block digital twins
Often, the major challenges of AI design in standard deployment are not related to technology — as evidenced by advances in algorithms and model accuracy. However, data collection, user acceptance, and reliable applications have not yet been adopted.
. Health care challenges with small, unsafe details
“Life is not just about your biology, but also about your genes, social structures, and everything else in between,” said Junaid Bajwa, Microsoft’s Chief Medical Scientist and physician at the UK’s National Health Service (NHS). ). We need access to more information around these boundaries to make the twins more ambitious, ”he adds.
As we move into new patient data sources, we should also refine what we already have. Fisher adds, “Today, electronic health records [EHR] exist primarily to inform patient billing and are not intended to enhance research.” He announces that the next major stages in digital twins will not be in AI research but in solving problems with small, sad details in health care.
There is visible hope. The streaming of digital data from devices such as wearables and mobile phones can allow access to high reliability, real-world data. Companies like 23andMe and Datavant are working to improve access to high quality research healthcare data.
2. It is not easy to change one’s mind and behavior
If innovators were able to create a digital twin that perfectly mimics a patient’s behavior, can we jump on its bandwagon? The historic adoption of new technologies suggests something else.
Bajwa remarks, “In my clinical practice, if I were asked to examine digital twins one day and repeat the same procedure the next day for a real person, I would find it very bizarre.” It would be easy to change technology and processes but not human thinking.
“It has taken a lot of people to use video conferences to discuss online,” said Bajwa. “Technology has been around for more than a decade, and its potential to achieve access and equity was evident. However, these catastrophic changes have taken a lot of time and global hardship. ”
3. AI thorny issues of AI trust, privacy, and bias
We have discussed the effective use of these technologies, but we cannot avoid their potentially harmful consequences. “While we need high-resolution data to incorporate digital twins into meaningful use, we must balance this performance with equality issues of trust, privacy, and bias,” added Shwen Gwee, Vice President and Head of Digital Technology at Bristol Myers Squibb.
Can patients be assured that their digital twins will not be tested for drugs or conditions that do not allow them? Additionally, clinical trial data sets today do not have the best representation from nationality and racial perception. Using such discriminatory information to train digital twins can lead to misconceptions of certain categories of people.
“I don’t think you can create unbiased AI technology,” Fisher said. “As AI is driven by data, it will show data bias. Instead of trying to make perfect algorithms, we should focus on how we use algorithm predictions in clinical trials. “Here again, the solutions are not based on technology but on proper procedures, human care, and involvement.
A bright future is with digital twins driven by AI
While digital twins can initiate change by empowering the control arm of clinical trial experiments, they wield great biological power. Actively addressing the three challenges of data collection, user acceptance, and reliable use can help them move forward.
How can a future be seen far away from digital twins?
“I see the potential to reduce the size of clinical trials safely and reliably,” said 25%, “Fisher said. “This can have a recurring effect on medical research and patients. It will enable all Biotech and Pharma companies to run clinical trials faster and less expensive. ”
Drawing on personal experiences, Bajwa shared the pain that cancer patients go through. “While there are many treatments available today, such as surgery, chemotherapy or immunotherapy, finding the best treatment for a patient can still help.”
Bajwa is looking at how digital twins can change this behavior in the future. “Suppose my multidisciplinary clinical team is given to a cancer patient – a 50-year-old mother of two on a beach in London. Can we use every possible combination of a patient’s digital twin to determine the course of treatment? Doing this quickly and accurately can provide the best quality of life for each person. This could help millions survive cancer. ”
