In a new development, a diagnostic tool developed by Princeton researchers will enable to analyze chest X-rays for patterns in lung diseases. Using this tool, doctors can have useful information about a patient’s condition at point of care, quickly and in an inexpensive manner.
The devastating range of attacks of COVID-19 is behind the development, said the principal investigator of the project. As hospitals are struggling to expand their serving capacity overrun with patients, two basic types of lung damage are observed by doctors: one more life-threatening than the other. As the line of treatment between the two types of lung damage can differ, distinguishing the two could help improve care and allow better allocation of resources.
New Tool beneficial over Time-consuming methods currently used
Meanwhile, current methods to assess differentiation between lung damage are time-consuming, which includes computed tomography (CT) scan. Conversely, the new method machine learning model involves looking at a simple X-ray image and find patterns that are subtle even for the expert human eye. And, with the new tool, doctors would have a new measure to determine the type and severity of COVID-19 pneumonia.
The new diagnostic tool does not involve change in practice, said the principal investigator of the project. Technicians do not have to perform anything differently, hospitals do not have to create any new procedure. For the X-rays that hospitals have and routinely take, the new tool can provide some extra information.
The researchers behind the tool posted the details of their work on medrxiv – a server for scientists to share results of a study in the form of early drafts while the study undergoes a formal editorial process.
The paper for the study is not peer-reviewed at the time of publication of the writing.