Diagnostic Reading #28: Five “Must Read” Articles on HIT and Radiology
Reading Time: 3 minutes read
In the news this week: FDA needs to keep pace with mobile apps; machine learning predicts 1 year mortality rates.
This week’s articles in Diagnostic Reading include: FDA: Regulations must respond to technological developments; Several evidence-based studies show equivalent diagnostic accuracy of tablets compared with DICOM-calibrated PACS workstation displays; medical imaging group warns China tariffs will cost manufacturers $138 million and compromise care; sound waves can separate cancer cells from blood with 86% efficiency; and machine learning predicts 1 year mortality rates using EHR data.
FDA: Regulations must respond to technological developments – Clinical Innovation+Technology
Government regulation can’t keep up with innovation in healthcare technology, especially when dealing with mobile medical applications according to top officials at the FDA. An article in JAMA states that efficiency in regulation can encourage development and utilization of mobile medical apps. The article states that current regulatory rules are designed for hardware technologies and need to be adapted for software-based technologies including mobile applications.
Diagnostic Reading on Mobile Devices – Everything Rad
Can radiologists do diagnostic reading on mobile devices with confidence? Several evidence-based studies have shown equivalent diagnostic accuracy of tablets compared with DICOM-calibrated PACS workstation displays. But are there limitations on what can be read on mobile devices?
A leading industry group is pressing the Trump administration to allow medical imaging technology to be exempt from tariffs imposed on imported goods from China, arguing that the policy will cost device manufacturers $138 million this year. The Medical Imaging and Technology Alliance (MITA) asked the Trump administration to establish an exemption process for the tariffs, which could prevent manufacturers from staying competitive and limit patient access to care. The tariffs that tax Chinese imports went into effect in July.
Sound waves can separate cancer cells from blood with 86% efficiency – Clinical Innovation+Technology
Researchers have developed a method to separate circulating tumor cells (CTCs) from blood samples that enable “liquid biopsies” to help with diagnosis, prognosis and suggestions for treatment. The technique can separate CTCs from a 7.5-mL vial of blood with at least 86 percent efficiency in less than an hour.
Machine learning predicts 1-year mortality using EHR data – Health Data Management
University of Minnesota researchers developed a machine learning algorithm using EHR data to improve care delivery for seriously ill patients by accurately predicting the risk of one year mortality. The model leverages EHR data such as vital signs, complete blood count, basic and complete metabolic panel, demographic information, as well as ICD codes. Age, blood urea nitrogen, platelet count, hemoglobin and creatinine were the most important variables in the RF model, according to authors of the study. The model could be a potential clinical decision support tool for improving end-of-life planning.