Diagnostic Robotics: Reimagining healthcare with medical predictive analytics on Google Cloud
About Diagnostic Robotics
Diagnostic Robotics is a medical triage and clinical predictions platform using AI to make healthcare systems more affordable, efficient, effective, and widely available, amidst strained budgets and a global shortage of healthcare workers.
Tell us your challenge. We're here to help.
Contact usDiagnostic Robotics adapts and scales its clinical triage platform to help healthcare stakeholders predict, analyze, and monitor the global spread of COVID-19 in real time using Google Cloud.
Google Cloud results
- Allows patients anywhere in the world to be diagnosed in a day by supporting workload and geographic scalability
- Quicker access to insights for mobile users worldwide thanks to the delay-free experience provided by Google Cloud CDN
- Frees developers from operational tasks by automating data provisioning so that applications can scale quickly without intervention
- Helps limit COVID-19 cases by monitoring 50 million+ people’s symptoms, scaling up in days from initial 8 million
Scales to 50M connected data sources in two days
Even before coronavirus brought the world to a halt in 2020, there were questions on how the global healthcare industry should account for a growing and aging population, the rising prevalence of chronic diseases, and higher labor costs amid workforce shortages all over the world. A Deloitte 2020 report proposes that healthcare systems work toward a future where disease prevention and early intervention, not just treatment, are its core focus. The golden question is how to support the sector in making that crucial shift. Enter Diagnostic Robotics, an Israeli company exploring the most advanced technologies in the field of artificial intelligence (AI) to help the global healthcare delivery system heal itself.
“We imagine a world where AI is harnessed to make healthcare better, more affordable, and more widely available,” says Yonatan Amir of his vision as CEO at Diagnostic Robotics, which he founded in August 2017. Back then, he had developed an algorithm for medical predictive analytics at the world-renowned Technion, the Israel Institute of Technology, where he studied computer engineering. Yonatan’s aim was to create a smart platform that hospitals, clinics, and urgent care centers could use to automate the prediction of patients’ diagnosis based on medical data already available to them. This layer of automation, he hoped, would help ease the load of overworked physicians and prevent the spread of diseases while reducing the cost of medical triage systems. As it turns out, he was right.
Two and a half years later, Diagnostic Robotics had grown into a team of nearly 100 experts bringing Yonatan and co-founder Dr. Kira Radinsky’s vision to life. This included 35 clinicians and physicians specialized in medical diagnostics, based in Tel Aviv and New York City. In the early days of the COVID-19 pandemic, the company partnered with the Israeli Ministry of Health and extended its successful clinical triage platform to monitor coronavirus-related symptoms specifically. They needed a way to monitor the health of the Israeli population of eight million, on a daily basis, to help keep COVID-19 from spreading if people started showing symptoms. “This was two weeks before patient zero in Israel,” Yonatan recalls. “We knew that we were dealing with something new and that we had to move quickly to protect people from it.”
So began the company’s race against time to design a smart healthcare delivery system for COVID-19. This entailed understanding the behavior pattern of a virus about which very little was known at this stage and simultaneously scaling its platform from supporting several hospitals and clinics, to monitoring the country’s entire population in real time, all within a matter of days. To do this, Diagnostic Robotics relied on the scalability and managed services of Google Cloud.
“Our technology is very robust, so we needed the scalability of Google Cloud to quickly deploy our system nationwide. Our infrastructure enabled us to provide all the national health organizations with a clinical snapshot of the entire population within just two to three days.”
—Yonatan Amir, CEO and co-founder, Diagnostic RoboticsScaling in a matter of days to monitor the spread of COVID-19
Diagnostic Robotics uses AI to create monitoring systems that identify the symptoms that could lead to a specific disease or deterioration of a patient’s health and to offer guidance to physicians based on these insights. To do this, Diagnostic Robotics imports medical data from health organization partners via a broad set of plugins, APIs, and integration components that can link their databases to Diagnostic Robotic’s platform. “When it comes to understanding and preventing disease, efficiency is everything. So we built our solution with flexibility in mind for our partners to use it as a plug-and-play and access insights very quickly, no matter what kind of technological setup they have,” says Yonatan.
This flexibility became even more essential for Diagnostic Robotics to monitor the spread of COVID-19, when it needed to feed its predictive algorithm with all the data available about the symptoms that could point to a positive coronavirus diagnosis. The most important source of data to study the behavior pattern of the virus, explains Dr. Kira Radinsky, Chairwoman and CTO at Diagnostic Robotics, turned out to be the Israeli people themselves.
“People in Israel are receiving a daily text message with a link to Diagnostic Robotic’s clinical triage system asking them about their symptoms. After filling in the questionnaire, each person receives a personalized risk profile and next-step guidance, and Diagnostic Robotics uses their anonymized information to create a heatmap of where COVID-19 is every day. This helps the Ministry of Health make decisions regarding the appropriate quarantine measures to be taken in each city and allocate resources such as medical supplies and COVID-19 tests,” Dr. Radinsky explains.
Of a population of eight million, 25% of people fill in the questionnaire once a week and 83% of respondents complete the questionnaire each time. This provides Diagnostic Robotics with enough data to identify the virus’ hotspots earlier than official statistics, which helps inform the national response to COVID-19 in a timely manner while Diagnostic Robotics runs predictive analysis on how and where the virus is likely to spread. Enabling this is Diagnostic Robotic’s capacity to scale its platform to ingest and process this amount of data in real time, an ability that Yonatan credits to its infrastructure setup on Google Cloud.
“Our technology is very robust, so we needed the scalability of Google Cloud to quickly deploy our system nationwide. Our infrastructure enabled us to provide all the national health organizations with a clinical snapshot of the entire population within just two to three days,” says Yonatan, who explains that Diagnostic Robotics has relied on Google Cloud since its founding.
“The outbreak of COVID-19 emphasizes that we made the right decision in terms of our infrastructure. We suddenly had hundreds of thousands of patients across the country being diagnosed by our system because Google Cloud helped us to scale and take action at the same speed or faster than the virus itself,” he adds.
Today, the end-to-end solution created by Diagnostic Robotics connects patients, healthcare providers, healthcare institutions, and public health agencies online on a medical triage platform that can be accessed via a web browser or app on almost any device. The Israeli population uses the platform to self-assess and monitor symptoms without the need for an in-person consultation, increasing endurance to public health guidelines of social distancing. Based on the platform’s predictions, coupled with up-to-date epidemiological maps, government decision-makers were able to make informed decisions on where to impose and lift lockdowns, allowing for a successful, hyper-localized containment approach. By moving the medical triage online, Diagnostic Robotics also reduces the load on healthcare call centers (by approximately 30%) and protects the safety of both patients and physicians by avoiding the overcrowding of clinics and hospitals amidst the outbreak of a highly contagious virus.
Because speed matters when it comes to monitoring a pandemic, Diagnostic Robotic's anonymous system can be installed in a single day by new partners who wish to use it. Its non-anonymous form can take up to three days to be ready for use.
“Google Kubernetes Engine was a major component of our success because it enabled us to scale very quickly in terms of workload and geography, using its managed services. We went from monitoring the symptoms of eight million people in Israel to 50 million in India in a matter of days.”
—Yonatan Amir, CEO and co-founder, Diagnostic RoboticsUsing predictive analysis to support healthcare systems globally
Soon, the word was out about how Diagnostic Robotics was able to support Israeli health agencies in monitoring the pandemic. The company needed to continue scaling as more stakeholders grew interested in how the company’s predictive algorithm could help them too, in their race against time to prevent the spread of the disease.
This situation, Diagnostic Robotics notes, highlights the value of AI in tackling one serious global problem. “We just don’t have enough doctors,” says Dr. Radinsky, who shares that the average waiting time for patients in emergency departments in Israel is 3.6 hours, and in China, eight hours. “Clinicians simply can’t care for thousands of patients at once and determine who needs urgent care. AI can help them by identifying how the virus spreads in real time and moving clinical triage to online platforms, protecting the limited healthcare personnel we have, and easing their load with a layer of automation,” she explains.
Additionally, the smart system has helped the world understand this new virus, Dr. Radinsky explains. “Our system identified specific COVID-19 symptoms very early on, before academic publications, symptoms such as losing the sense of smell and taste, and extreme fatigue. This kind of information changed the guidelines for Israel to run COVID-19 tests,” she explains.
Diagnostic Robotics develops its own data processing and predictive algorithms in-house and uses a host of Google Cloud solutions to scale their applications. “The Google Cloud systems we use today were mostly already in place before the pandemic, all we had to do was scale them up to manage the new demand for our platform during the coronavirus outbreak,” Yonatan explains. “Google Kubernetes Engine was a major component of our success because it enabled us to scale very quickly in terms of workload and geography, using its managed services. We went from monitoring the symptoms of eight million people in Israel to 50 million in India in a matter of days,” he adds.
Diagnostic Robotics integrates Cloud SQL, a fully managed relational database that helps the company to scale and automate its database provisioning without adding to the operational hurdles of its in-house team of developers. Diagnostic Robotics also uses Compute Engine to ensure that all applications have the capacity they need as the company scales globally, and Google Cloud operations tools (formerly known as Stackdriver) to monitor, troubleshoot, and improve the performance of applications on its Google Cloud environment. Finally, Cloud CDN brings a consistently fast experience to all users of the Diagnostic Robotics platform by acting as a complement to the private network of Google Cloud and improving site performance for mobile users around the world.
“Our global expansion is happening very fast. We’re in conversation with stakeholders in the US, Spain, Argentina, Brazil, and many others. Google Cloud helps us to be where support is needed, and we will continue relying on its infrastructure to make healthcare better, cheaper, and more widely available worldwide.”
—Yonatan Amir, CEO and co-founder, Diagnostic RoboticsTracing new paths for the healthcare sector wherever support is needed
In the midst of a global pandemic, the value that Diagnostic Robotics brings to the world is undeniable and resonates with many. But the company is doing more than monitoring the spread of COVID-19. It is trailing a new path for the healthcare system to advance in the future. Its key legacies include a preventative approach that clinicians can embrace by using AI to predict clinical events based on their own medical database, the ability for patients to access personalized medical recommendations faster, online, without exposing themselves to the waiting rooms of hospitals, and a support system for physicians to make better-informed and faster decisions using AI technology during consultations.
Proof of the company's success is the efficient way it is helping the world to understand coronavirus. But perhaps more telling of how it’s driving change in the healthcare sector is the growing demand for the platform’s support worldwide. “Our global expansion is happening very fast,” Yonatan shares. ”We’re in conversation with stakeholders in the US, Spain, Argentina, Brazil, and many others. Google Cloud helps us to be where support is needed, and we will continue relying on its infrastructure to make healthcare better, cheaper, and more widely available worldwide.”
Speaking of the company’s latest major partnership, he adds, “We’re excited to collaborate with Mayo Clinic and implement our triage platform. This collaboration reflects the synergy between our technological vision and Mayo Clinic’s cutting-edge medical expertise.”
Tell us your challenge. We're here to help.
Contact us
About Diagnostic Robotics
Diagnostic Robotics is a medical triage and clinical predictions platform using AI to make healthcare systems more affordable, efficient, effective, and widely available, amidst strained budgets and a global shortage of healthcare workers.