Email
Password
Remember meForgot password?
    Log in with Twitter

article imageQ&A: How AI can be a lifeline for overworked radiologists Special

By Tim Sandle     Apr 22, 2019 in Health
Artificial intelligence can analyze hundreds of images in a matter of minutes and triage abnormalities so that radiologists can prioritize potentially critical cases and save lives. An expert from the startup Aidoc discusses the potential.
The number of medical images taken at hospitals has exploded, according to one study, from 1999 until 2010, the average CT exam increased from 82 images to 679, yet the number of radiologists has remained static. As a result, critical cases can often wait an hour or more until they are seen by a radiologist, and more than 97% of all medical images go without being analyzed.
To help address this and to aid the busy radiologist, artificial intelligence tools are filling the gap. In this rapidly emerging field, Israeli startup Aidoc is pioneering ways to improve medical image analysis.
To understand more, Digital Journal spoke with Aidoc CEO Elad Walach.
Digital Journal: What are the main demands facing radiologists?
Elad Walach: The biggest demand for radiologists is the massive increase in the amount of imaging performed on patients. Advanced imaging techniques like CT and MRI were once used only for the most serious of cases but they are now routine diagnostic tools. They also take more images per patient: In 1999, an average patient CT contained 82 images; by 2010, the number had increased to 679 per patient! Radiologists are faced with an ever-increasing workload as medical imaging becomes cheaper, more accessible and more advanced.
DJ: Do these demands result in errors?
Walach: Radiologists are tremendously talented physicians but obviously, if they have less time to spend on each patient and each image, there is an impact. That impact has been offset by the improvements in imaging tech itself. Patient outcomes are still improving because radiologists see both more images and higher-quality images. But the overload means that some images are never even looked at, and sometimes a doctor can spend only a couple of seconds per image.
DJ: How has radiology been affected by digital transformation?
Walach: Radiology was early to undergo digital transformation. Instead of old-fashioned X-Ray transparencies, most hospitals in advanced countries use picture archiving and communication systems (PACS) that digitize the whole process from image capture, storage and retrieval. PACS work for X-Rays, CTs and MRIs and use standards file formats. This actually makes it an excellent discipline for integrating Machine Learning because the data and infrastructure are already digital.
DJ: How is machine learning shaping radiology?
Walach: Machine Learning is helping radiologists in a few different ways. First, ML can help identify, flag and triage abnormalities in scans, adding an AI ‘second reader’ that helps radiologists confirm their work. In some cases, a computer-vision deep learning solution can detect abnormalities that may not be obvious to the human eye.
ML is also being used in radiology reports. Natural Language Processing is analyzing reports written by radiologists and understanding what they mean, enabling advanced data processing.
More about radiologists, Medical, Artificial intelligence, Radiology, aidoc
More news from
Latest News
Top News