Startups Demonstrate Power of AI and Machine Learning to Reshape Healthcare
By Suzy Engwall
In healthcare, some problems are so enormous and complex that a solution has long seemed impossible. Artificial intelligence and machine learning promise new insights into these multifactorial challenges, as we learned at our recent Startup Demo Day: AI and Machine Learning. Not only that, but artificial intelligence in healthcare is enabling humans to do their jobs better and improve patient outcomes in some really exciting ways.
As Bradd Busick, senior vice president and CIO of MultiCare Health System, explained, it’s all about “the ability to turn insight into action, particularly when that improves patient safety.”
Our recent Startup Demo Day, co-hosted with MultiCare, put the spotlight on 10 companies using AI and machine learning to implement incredible innovations in healthcare.
In just seven minutes each, these startups showed us technology that promises to:
- Predict when an MRI or CT machine will malfunction before it does. As a result, a hospital can anticipate downtime and fix problems, avoiding unplanned outages and creating a better patient experience. Glassbeam takes data from all kinds of hospital systems, regardless of vendor, and places the analytics “on a single pane of glass,” said Cary Lucian, Glassbeam’s senior vice president of sales and business development, Americas. In short, Lucian explained, Glassbeam takes clinical insights from connected systems and applies expert rules and ML- and AI-type learning to increase machine uptime and use, as well as patient throughput. For healthcare systems, the result is increased revenue and reduced operational costs.
- Document the wound healing process and accurately predict recovery timelines. Pressure ulcers are a pervasive problem that requires expensive, time-intensive treatment. Current procedures involve scanning, measuring and uploading wound data by hand. This requires 20 to 130 minutes per patient, on average. The Wound Capture application provides rapid wound imaging and analysis, using a machine learning algorithm to classify wounds and predict how long each will take to heal.
- Encourage patients to adhere to their prescribed treatment regimen. Patient non-adherence costs $100 billion per year, according to conservative estimates. “Everybody pays the price,” said Megan Thorp, Hcare co-founder. Hcare’s friendly AI system, called Dr. ANIE, helps healthcare systems improve adherence by providing information about what’s driving individual patients’ behaviors: avoidance, control, risk-adversity or something else. As patients adopt the right practices, they heal faster and healthcare systems see lower readmission rates and a higher quality of care.
- Identify patients’ real risk of heart attacks. Heart attack prevention has been largely unchanged for decades, said Cleery CEO James Min, who was a cardiologist for 15 years. Doctors look for risk factors, such as cholesterol levels. But 70% of people who suffer heart attacks have normal cholesterol, Min noted. The strongest predictor of heart attack risk is atherosclerosis (plaque in the arteries), which hasn’t typically measured by heart doctors. Cleerly uses a non-invasive, rapid procedure called coronary CT angiography to detect harmful plaque. Cleerly’s AI algorithms produce a detailed analysis for clinicians as well as easy-to-understand insights for patients.
- Streamline healthcare operations with smart forecasting and scheduling. “What if you could forecast future patient demand? What if you could forecast patient no-shows, staff no-shows and manage operations efficiently by harnessing the information that’s already in your EHR – and take that information and seamlessly integrate it into your operational flow?” asked Darryl Kirsh, CEO of Aidan Health. That’s what the Aidan AI-powered platform does to improve planning, scheduling and patient engagement, he said, thus creating a better experience for patients and staff and boosting the bottom line.
- Empower people to monitor their own health or their loved one’s at home. “There are no solutions today to support home-based care effectively,” said Widy Medina, co-founder and CEO of Telebionix. His company’s Remosense is a simple-to-use, clinical-grade device that captures seven different vital signs and transmits the data to providers via a secure app. This can increase healthcare access for underserved communities and improve patient engagement and quality of care.
- Prevent patients from falling in healthcare settings. Falls are the leading cause of injury-related death among adults ages 65 and older, according to the Centers for Disease Control and Prevention. In hospital settings alone, around 250,000 injuries occur from falls each year. Helpp.ai prevents falls from occurring by providing early warnings that an at-risk patient is trying to get out of bed. The AI system analyzes patients’ movements to predict these exit attempts, then notifies nursing staff so they can intervene.
- Provide continuous patient monitoring to prevent falls and pressure injuries. The Cognito Health system affordably turns every bed in a healthcare setting into a smart bed. A sensor placed on hospital beds detects patients’ movement. A bedside control box sends alerts to nurses’ mobile devices and the nurse station dashboard if a patient tries to get out of bed or needs to be repositioned to prevent pressure injuries. And a dashboard monitors all patients in the unit simultaneously.
- Make last-mile delivery more efficient using robots and AI. Interplai is a smart, collaborative logistics platform that uses AI and machine learning technology to optimize efficiency for last-mile delivery. Last-mile delivery means moving goods from a transport hub to the last 20 feet of their journey — and accounts for 53% of overall shipping costs, noted COO Eddy Chiang. Interplai offers three solutions:
- AI-powered vehicle route optimization (VRO).
- Capacity-as-a-service to find cheaper third-party logistics companies.
- Robots-as-a-service, in which autonomous vehicles carry and deploy sidewalk delivery robots.
- Improve patient outcomes with nutrition, monitoring and coaching. By 2025, half the population will be diabetic or prediabetic, said Eziah Syed, co-founder and CEO of Mend. Add to that other chronic conditions caused by or related to nutrition, and “we’re in code red. It’s a crisis.” Mend seeks to solve the problem with digital nutritherapy: a platform that uses AI, machine learning and the cloud to deploy clinical nutrition to patients. For example, a patient who has just had a knee replacement would receive a formulation designed to lower muscle atrophy, enhance wound healing and lower pain, among other benefits. Smart devices would remotely monitor patients, enabling digital behavioral support to improve outcomes.
Meet the Supporters of Startup Demo Day:
Watch a recording of our Startup Demo Day here!