Neomed employs AI to help physicians to issue test reports in less than 24 hours
About Neomed
Neomed is a platform which enables integration between clinics, hospitals, and physicians to improve the processing of tests with visual representations, such as an electrocardiogram. It was established in 2017 with the aim of incorporating technology to medicine by automating these tests processing, therefore reducing the patient’s waiting period to obtain his or her reports and diagnosis.
Tell us your challenge. We're here to help.
Contact usThanks to Google Cloud solutions and Google for Startups Accelerator mentorship, the startup developed an AI model for its platform in just one month.
Results
- Platform dataset development within a week
- Deployments are now daily and without service interruptions
- 60% improvement in medical efficiency for reports preparation
- 97.5% data processing time reduction
- 90% reduction in infrastructure costs
Reports delivery time from seven working days to 24 hours.
In general, a patient waits at least seven days to receive the report of a graphically represented test, such as an electrocardiogram or an electroencephalogram. Neomed is a startup founded to abridge the period between the patient’s symptoms and the adequate treatment for his or her diagnosis. Established in 2017, the company connects clinics and hospitals of Brazil with physicians of different districts in order to process test results, operating as a sort of marketplace for medical reports.
The idea to create this platform came up when founders Bruno Farias and Gustavo Kuster acknowledged the difficulties experienced by health units to manage thousands of reports, besides the lack of physicians trained to assess tests in locations far away from the major centers. These two factors increase the time to obtain test results and, therefore, delay the diagnosis and treatment of a condition.
After three years of different enhancements on its product, Neomed still had a dream: integrate artificial intelligence with its platform. This aim motivated its founders to enter the fifth class of Google for Startups Accelerator, which took place between March and June 2020. During this acceleration program, the startup learned about the Google Cloud tools which would help to boost a solution.
Takeaways and acceleration
With the tools covered in the program, Neomed was able to rapidly try the countless options to develop its AI product and devise the best path to follow.
“Within a month, we progressed as much as we would have in a year and a half.”
—Bruno Farias, co-founder and CPO, NeomedAs they learned about the options offered by Cloud Functions thanks to the mentorships, they analyzed other solutions, such as Cloud Storage, Google Kubernetes Engine, and TensorFlow, which helped them to create a dataset within a week. During the following weeks, they built a model to train this set to translate waves of an electrocardiogram into data. They also implemented an algorithm without a server operating within the infrastructure.
“It was our ‘Eureka’ moment. As from there on, we managed to create and train an AI algorithm which assessed the degree of severity of an electrocardiogram, therefore simplifying the physician’s work. It was one of the mentorship’s highlights,” stated Diego Volts, CTO of the company.
Security was another aspect thoroughly analyzed at the program. Since the startup had migrated its infrastructure recently, it did not yet have all security resources correctly implemented. Today, the team has solid and formalized procedures for internal control and visibility to ensure they are in compliance with all health regulations and certifications.
“Google Cloud tools suite helped us not only to build the product, but also to improve the environment which we already have.”
—Diego Voltz, CTO, NeomedImprovements for clients and the team
Nowadays, Neomed has thousands of tests being uploaded to the platform by clinics and hospitals on a monthly basis, thus optimizing and centralizing the generation and management of reports in one single location. As the AI model allows for the detection of possible alterations, the solution helps physicians to prioritize more urgent cases, like a sort of triage.
“We managed to improve medical efficiency by two thirds. A physician would spend 60% more time assessing a test outside Neomed,” explains Farias. He also adds that data processing time plummeted by 97.5% and that clinics and hospitals, which usually took up to seven working days to deliver a test report, today do so in less than 24 hours.
Participation in the acceleration program and Google Cloud solutions also had a positive impact on the startup's internal procedures. Kubernetes Engine caused microservices management to become more fluid and streamlined the deployment process. If in the past it was indeed necessary to wait for two weeks to upload an application, now the team manages to do so on a daily basis, without service interruptions. This advantage permits a quicker delivery of improvements and client orders, thus offering a more satisfying experience.
Besides from a productivity enhancement, there was a 90% decrease in infrastructure costs, with more space available for growth. “As of today, app escalation has exceeded the credit card’s costs escalation,” jokes the CTO.
“Our demand now increased and, if we had not migrated to Google Cloud, we certainly would have spent two or three times more.”
—Diego Voltz, CTO, NeomedEvolution of AI model and overseas expansion
The startup maintains its acceleration rate even after the end of mentorships, and it is already working on the evolution of the AI model. It is expected that, apart from indicating whether there are alterations in a test or not, the system also manages to point out such possible anomalies, speeding up the test report generation process.
In addition, Neomed is starting operations in Chile, planning to expand to different parts of the world within the next five years. “We just executed a contract with one of the biggest health operators in Brazil and we want to become the world’s biggest medical technology company in the world,” Farias stated.
Tell us your challenge. We're here to help.
Contact us
About Neomed
Neomed is a platform which enables integration between clinics, hospitals, and physicians to improve the processing of tests with visual representations, such as an electrocardiogram. It was established in 2017 with the aim of incorporating technology to medicine by automating these tests processing, therefore reducing the patient’s waiting period to obtain his or her reports and diagnosis.