Several trends are causing seismic changes in the health care sector, one of them mainly being technology. The impact of IoT and artificial intelligence (AI) on health care involves medical electronics such as robot-assisted surgeries, diagnostics, medical research, connected devices, and energy harvesting.
As one of the earliest robot surgical systems, the da Vinci was launched in 2000. Since then, companies such as Auris/Hansen, Corindus, Stereotaxis, Zimmer-Biomet/Medtech, TransEnterix, Titan Medical, Stryker, Smith & Nephew, Medrobotics, and Mazor have adopted their robotics surgery platforms. These platforms have broadened the application of surgery robotics from abdominal or pelvic procedures to cardiology, colorectal, gastrointestinal, gynecology, orthopedics, pediatrics, spine, transplant, urology and oncology fields.
The surgical field has had a net positive movement towards using robotics, as most surgeons acknowledge that robotics help doctors operate, train medical students, encourage collaboration, reduce complications, pain, and blood loss, and accelerate patient recovery. However, the high unit cost of the platforms remains a barrier for most of the hospitals, and some studies have even cast doubt on the return on investment of the platforms. In addition, some surgeons feel restricted by the unchangeable surgical protocols in the software. Therefore, doctors are currently viewing robotics as an aid, not a replacement of manual surgery.
Leading hospitals are using data analytics, machine learning and artificial intelligence to identify and treat patients who are most at-risk more and more. Recently, the American College of Radiology Data Science Institute (ACR DSI) has released several high-value use cases for AI in medical imaging, such as identifying cardiovascular abnormalities, detecting fractures and musculoskeletal injuries, aiding the diagnosis of neurological disease such as Lou Gehrig's, flagging thoracic complications in patients, and screening for common cancer.
AI is also going beyond medical imaging. A Danish AI company is developing an algorithm that analyzes the vocal tone and background noise of emergency-callers to determine the likelihood of cardiac arrests. So far, the program is having a higher success rate than humans. Also, the British government is generating an algorithm to identify patients who are at risk of developing cancer by examining the medical records, lifestyle habits, and genetic information pooled from health charities and the National Health Service.
Using AI to diagnose patients is still in the burgeoning stage and has few use cases. However, it shows great potential in mining big data and implementing preventive programs.
IBM’s Watson was the first AI platform to enter the field of medical research. Since then, other tech companies such as Intel, Google, and Microsoft have joined the effort. Intel is working with Philips to use AI inference to improve medical imaging analysis, identify organs and patterns, and screen skin cancer. Google is working with universities such as UC San Francisco, University of Chicago, Stanford, and Harvard on diabetic retinopathy, cancer metastasis, and cardiovascular diseases. Meanwhile, Microsoft and its academic partners are using machine learning to study 3D radiological images, epidemics and neurological disorders.
While there has been some debate over whether AI platforms such as Watson will help cure diseases like cancer, an increasing number of research institutions are using AI platforms for patient recruitment for clinical trials, genomic studies and cancer diagnosis. There is evidence that AI will accelerate research: Recently, a machine-learning algorithm developed by an MIT team can analyze 3D scans 1,000 times faster than the current rate and may become independent of learning from tissue samples in the future.
Telehealth had a humble beginning, where it was just a group of VA nurses calling veterans on a landline at home and asking them a list of health-related questions. Then telemedicine expanded to video-conferencing, remote patient monitoring, and patient health information sharing via mobile devices. The prevalence of telemedicine has increased almost six times from 2014 (13 percent) to 2017 (75 percent). Connected health has evolved in-step with government policies like Obamacare. For example, Partners is working with Hitachi to develop a solution to monitor post-operative heart failure patients at home to reduce their chance of readmission within 30-days. Also, there is a broader use of smart pills to monitor patient compliance. With the increasing prevalence of IoT devices such as Alexa, Siri, Apple Watch, and Google Home, many tech companies have an easy avenue to get into telecare and provide functionalities including remote monitoring, telemedicine, or behavior modification, that will improve the quality of care and reduce expensive emergency care episodes.
One other important aspect is the testing requirement of the new hardware and software used in medical electronics. Testing ensures the performance and regulatory compliance of the device software. VectorCAST is a TÜV SÜD-certified product testing platform by Vector. VectorCAST/QA extends the coverage of code analysis to 100 percent and focuses on code areas that have the highest risk of failure. VectorCAST/C++ allows engineers to set up low- and high-level unit tests that cover singular or multiple functions, respectively. Together, these will significantly reduce the risk of code failure. Regardless of the regulatory class of a medical device, VectorCAST tools provide a dependable and repeatable testing process on its software development.
The medical electronics field has tremendous potential and opportunity, as evidenced by the many large and small companies that are pursuing AI technology to further advance their testing and treatment. While the effectiveness of medical technology is uneven across different fields and applications, overall, users and practitioners see benefits and it will also offer developers great opportunities.