Artificial intelligence may fall short when analyzing data across multiple health systems

November 6, 2018, The Mount Sinai Hospital

Artificial intelligence (AI) tools trained to detect pneumonia on chest X-rays suffered significant decreases in performance when tested on data from outside health systems, according to a study conducted at the Icahn School of Medicine at Mount and published in a special issue of PLOS Medicine on machine learning and health care. These findings suggest that artificial intelligence in the medical space must be carefully tested for performance across a wide range of populations; otherwise, the deep learning models may not perform as accurately as expected.

As interest in the use of computer frameworks called convolutional neural networks (CNN) to analyze medical imaging and provide a computer-aided grows, recent studies have suggested that AI image classification may not generalize to new data as well as commonly portrayed.

Researchers at the Icahn School of Medicine at Mount Sinai assessed how AI models identified pneumonia in 158,000 chest X-rays across three medical institutions: the National Institutes of Health; The Mount Sinai Hospital; and Indiana University Hospital. Researchers chose to study the diagnosis of pneumonia on chest X-rays for its common occurrence, clinical significance, and prevalence in the research community.

In three out of five comparisons, CNNs' performance in diagnosing diseases on X-rays from hospitals outside of its own network was significantly lower than on X-rays from the original health system. However, CNNs were able to detect the system where an X-ray was acquired with a high-degree of accuracy, and cheated at their predictive task based on the prevalence of pneumonia at the training institution. Researchers found that the difficulty of using deep learning models in is that they use a massive number of parameters, making it challenging to identify specific variables driving predictions, such as the types of CT scanners used at a hospital and the resolution quality of imaging.

"Our findings should give pause to those considering rapid deployment of platforms without rigorously assessing their performance in real-world clinical settings reflective of where they are being deployed," says senior author Eric Oermann, MD, Instructor in Neurosurgery at the Icahn School of Medicine at Mount Sinai. "Deep learning models trained to perform medical diagnosis can generalize well, but this cannot be taken for granted since patient populations and imaging techniques differ significantly across institutions."

"If CNN systems are to be used for medical diagnosis, they must be tailored to carefully consider clinical questions, tested for a variety of real-world scenarios, and carefully assessed to determine how they impact accurate diagnosis," says first author John Zech, a medical student at the Icahn School of Medicine at Mount Sinai.

This research builds on papers published earlier this year in the journals Radiology and Nature Medicine, which laid the framework for applying computer vision and deep learning techniques, including natural language processing algorithms, for identifying clinical concepts in radiology reports for CT scans.

Explore further: Artificial intelligence platform screens for acute neurological illnesses

Related Stories

Artificial intelligence platform screens for acute neurological illnesses

August 13, 2018
An artificial intelligence platform designed to identify a broad range of acute neurological illnesses, such as stroke, hemorrhage, and hydrocephalus, was shown to identify disease in CT scans in 1.2 seconds, faster than ...

Machine learning techniques generate clinical labels of medical scans

January 31, 2018
Researchers used machine learning techniques, including natural language processing algorithms, to identify clinical concepts in radiologist reports for CT scans, according to a study conducted at the Icahn School of Medicine ...

Pathology test uses AI to predict prostate cancer progression following surgery

October 2, 2018
A pathology test that applies artificial intelligence (AI) to characterize tissue samples can accurately predict clinically significant prostate cancer disease progression following surgery, according to a study conducted ...

Lung ultrasound may be a safe substitute for chest X-ray when diagnosing pneumonia in children

April 12, 2016
Lung ultrasound has been shown to be highly effective and safe for diagnosing pneumonia in children and a potential substitute for chest X-ray, according to a study conducted at the Icahn School of Medicine at Mount Sinai. ...

Artificial intelligence may help diagnose tuberculosis in remote areas

April 25, 2017
Researchers are training artificial intelligence models to identify tuberculosis (TB) on chest X-rays, which may help screening and evaluation efforts in TB-prevalent areas with limited access to radiologists, according to ...

Recommended for you

Precision neuroengineering enables reproduction of complex brain-like functions in vitro

November 14, 2018
One of the most important and surprising traits of the brain is its ability to dynamically reconfigure the connections to process and respond properly to stimuli. Researchers from Tohoku University (Sendai, Japan) and the ...

Gene mutation found to cause macrocephaly and intellectual deficits

November 13, 2018
The absence of one copy of a single gene in the brain causes a rare, as-yet-unnamed neurological disorder, according to new research that builds on decades of work by a University at Buffalo biochemist and his colleagues.

Can scientists change mucus to make it easier to clear, limiting harm to lungs?

November 12, 2018
For healthy people, mucus is our friend. It traps potential pathogens so our airways can dispatch nasty bugs before they cause harm to our lungs. But for people with conditions such as cystic fibrosis (CF) and chronic obstructive ...

Scientists uncover new gatekeeper function of anti-aging molecule

November 12, 2018
The protein klotho has been shown to promote longevity and counteract aging-related impairments. Having more klotho seems to allow for longer and healthier lives, whereas a depletion of this molecule accelerates aging and ...

Mutations, CRISPR, and the biology behind movement disorders

November 12, 2018
Scientists at the RIKEN Center for Brain Science (CBS) in Japan have discovered how mutations related to a group of movement disorders produce their effects. Published in Proceedings of the National Academy of Sciences, the ...

Researchers explain how your muscles form

November 12, 2018
All vertebrates need muscles to function; they are the most abundant tissue in the human body and are integral to movement.

0 comments

Please sign in to add a comment. Registration is free, and takes less than a minute. Read more

Click here to reset your password.
Sign in to get notified via email when new comments are made.