Using artificial intelligence to bring doctors into patients’ homes
The ability to monitor patients at home is critical for future healthcare systems, which will be increasingly burdened by ageing populations. Traditional remote patient monitoring is complicated and onerous however, often involving painful integrations with various medical devices. This can make the process difficult for patients, and at times impossible. In the EU-funded Doctomatic project, researchers developed a new software-as-a-service (SaaS) platform, which integrates artificial intelligence (AI) and enables remote patient monitoring without the need for intricate medical devices for diagnostics. Patients can simply download an app and share data from any domestic health monitor with medical practitioners. Doctomatic aims to employ AI to enhance preventive healthcare, bringing healthcare costs down while accelerating patient diagnostics. Remote patient monitoring tools can give insights to clinicians for the early detection of disease, which can avoid complications and the development of chronic conditions. “The vision is to be expert providers of the data-acquisition capabilities, to enhance healthcare services for patients,” says Carmen Pauline Rios Benton, co-founder and CEO of the company Doctomatic, and also the project coordinator.
Using artificial intelligence to analyse medical images
The Doctomatic project funded the design of a patent-pending proprietary visual algorithm able to transform images captured by domestic medical devices into usable data. The AI can analyse and understand information from screen snapshots of widely available and low-cost devices, such as scales, heart rate monitors, glucometers, pulse oximeters and thermometers. Through the project, the Doctomatic team improved the technology, specifically by enhancing the system’s accuracy by 26 %. “This objective was achieved through a multipronged approach,” explains Rios Benton. “This consisted of evaluating and choosing the most appropriate AI model for Doctomatic, developing an automatic data set generation system, creating a smart system around the core AI model and finally developing a benchmarking software for evaluating the different models.” The team developed the system using their own data sets, along with newly generated data gathered during the development of the project. Doctomatic seamlessly integrates into healthcare provider apps and electronic health records. “Its unique selling proposition lies in being device-agnostic, making access to care easier, lowering costs and ensuring seamless interoperability,” says Rios Benton.
Capturing real-world data during palliative care
Thanks to improvements in accuracy through the project, the Doctomatic team were able to improve the algorithm in order to read new devices. The technology successfully analysed screens from artificial ventilators in the palliative care unit of Sant Joan de Déu Hospital in Barcelona, which the team are now working to replicate and scale globally. “I believe this is one of our proudest moments and probably one of the best use cases in regard to the impact it delivers for the families involved,” adds Rios Benton.
Bringing novel palliative healthcare technology to the market
The team have so far developed the application for demonstration purposes and have taken the product to market. Doctomatic currently operates in Colombia, Brazil, Mexico and Spain. “This is a continuous learning and improvement project,” notes Rios Benton. “We are very grateful for the support of the EU via the Womentech funding, which has helped us improve our technology. “We are in conversations with healthcare technology providers in order to at least be present in two more European countries by the end of this fiscal year.”
Keywords
Doctomatic, remote, patient, monitoring, health, chronic, palliative care, artificial intelligence, analyse, images