A cancer diagnosis or a stay in the intensive care unit (ICU) often bring confusion, fear and questions about the best course of treatment. That’s why a group of doctors and scientists at GE Healthcare and Roche Diagnostics are looking for a new way to predict the most effective treatment for an individual by applying data analytics to the problem.
Over the last decade, big data made inroads into personal fitness, energy, politics and other fields. Now it’s moving into healthcare. The idea is that smart algorithms looking for insights in terabytes of medical information will help physicians better serve their patients with earlier diagnoses and customized treatment plans.
The partnership between GE and Roche announced in January will create digital platforms for so-called “precision health” in oncology and critical care. The oncology platform, the first of its kind, will take “in-vivo” data obtained directly from the patient by radiological imaging and monitoring equipment to characterize the tumor at the anatomical and physiological level. It will combine the data with “in-vitro” information from laboratory tests that characterize the tumor at the molecular level by looking at tissue pathology, blood-based biomarkers, genomic alterations (cancer-relevant mutations) and other factors. The system also will integrate data from electronic medical records, medical best practices and the latest research.
Today in oncology, the team of specialists assembled to treat individual cancer patients typically meets on a set schedule to swap details about the patient and test results to make a diagnosis and assess how the disease is progressing. “Everyone gets together and pulls up all these images and slides and spreadsheets,” says Nadeem Ishaque, chief innovation officer at GE Healthcare’s Imaging business. “It’s inefficient when time is of the essence.”
The new platform will provide care teams with a comprehensive data dashboard that pulls together information from diagnostic imaging (such as CT or MRI scans), blood tests, tissue samples, gene sequencing, the latest clinical trials, and other sources. The idea is to give oncologists all the information they need in one place to arrive at an accurate diagnosis and determine the most effective course of treatment for a specific patient. Software apps available on the system, using analytics and machine learning, will compare the patient’s data with historical treatment data and outcomes. “The electronic board will be accessible to everyone [involved in caring for that patient] all the time, no matter where they are,” Ishaque says. “Software apps will integrate all the longitudinal data on the patient, from seven years ago, three years ago and today, so the oncologist can manage the patient throughout their lifetime and through every stage of the disease.”
Roche Diagnostics and GE Healthcare also are working on another platform for critical care that will have applications in the ICU and the emergency room (ER). The new combined dataset can then be integrated into existing clinician workflows and help physicians to identify or even predict infectious diseases. Infectious diseases are a huge risk for patients in the ICU because they are more susceptible. “Say someone comes to the ER with a fever, but it’s not obvious why,” Ishaque says. “All the information — vital signs, physiological parameters, blood tests, X-rays, ultrasounds — will be combined and analyzed immediately to help with a diagnosis for quicker and safer treatment.”
Says Ishaque: “The end game is saving lives and a better quality of life for the patients.”