Breaking News
February 18, 2019 - How Inactivity and Junk Food Can Harm Your Brain
February 18, 2019 - Diabetes tops common conditions for frequent geriatric emergency patients
February 18, 2019 - Longer-lived sperm produces offspring with healthier lifespans
February 18, 2019 - New dental adhesive prevents tooth decay around orthodontic brackets
February 18, 2019 - New eHealth tool shows potential to improve quality of asthma care
February 18, 2019 - New Australian initiative helps emergency clinicians to improve patient care
February 17, 2019 - Apellis Pharmaceuticals’ APL-2 Receives Fast Track Designation from the FDA for the Treatment of Patients with Paroxysmal Nocturnal Hemoglobinuria
February 17, 2019 - Researchers identify faulty ‘brake’ that interferes with heart muscle’s ability to contract and relax
February 17, 2019 - Support from trusted adults can reduce risk of dying in suicidal teens, finds study
February 17, 2019 - Heart attack awareness improved since 2008
February 17, 2019 - Exercise gives a better brain boost to older men than women
February 17, 2019 - New research disproves previous assumptions of how looks influence personality
February 17, 2019 - Cannabis use as a teenager linked to depression later in life
February 17, 2019 - Sinks by Toilets in ICU Patient Rooms Harbor Harmful Bacteria
February 17, 2019 - Cancer cells’ plasticity makes them harder to stop
February 17, 2019 - Young cannabis users have increased risk of depression and suicidal behavior
February 17, 2019 - Tasmanian Devils Likely to Survive Cancer Scourge
February 17, 2019 - Neoadjuvant PD-1 blockade seems effective in glioblastoma
February 17, 2019 - Personal, social factors play role in enabling sustainable return to work after ill health
February 17, 2019 - Mouse studies show ‘inhibition’ theory of autism wrong
February 17, 2019 - Study shows how neuroactive steroids inhibit activity of pro-inflammatory proteins
February 17, 2019 - Use of liver grafts from older donors decreased despite better outcomes in recipients
February 17, 2019 - MUSC researchers discover new mechanism for a class of anti-cancer drugs
February 17, 2019 - HPV misconceptions are causing women to miss smear tests
February 17, 2019 - Sanofi and Regeneron Offer Praluent (alirocumab) at a New Reduced U.S. List Price
February 17, 2019 - Researchers say auditory testing can identify children for autism screening
February 17, 2019 - New method analyzes how single biological cells react to stressful situations
February 17, 2019 - WVU gynecologic oncologist investigates novel treatment for cervical and vaginal cancers
February 17, 2019 - ADHD diagnoses poorly documented
February 17, 2019 - Majority of gender minority youth do not identify with traditional sexual identity labels
February 17, 2019 - AbbVie, Teneobio enter into strategic transaction to develop potential treatment for multiple myeloma
February 17, 2019 - Lower Birth Weight May Up Risk for Psychiatric Disorders
February 17, 2019 - Scientists identify reversible molecular defect underlying rheumatoid arthritis
February 17, 2019 - Moffitt researchers shed light on how CAR T cells function mechanistically
February 16, 2019 - Female Anatomy May Play Big Role in Sperm’s Success
February 16, 2019 - BMI may mediate inverse link between fiber intake, knee OA
February 16, 2019 - Movement impairments in autism can be reversed through behavioral training
February 16, 2019 - Studies address racial disparities in postpartum period and cardiovascular health
February 16, 2019 - Scientists implicate hidden genes in the severity of autism symptoms
February 16, 2019 - Decreased deep sleep linked to early signs of Alzheimer’s disease
February 16, 2019 - Neuroscientists show how the brain responds to texture
February 16, 2019 - Gilead Announces Topline Data From Phase 3 STELLAR-4 Study of Selonsertib in Compensated Cirrhosis (F4) Due to Nonalcoholic Steatohepatitis (NASH)
February 16, 2019 - What Can I Do About Sweating? (for Teens)
February 16, 2019 - Companies navigate dementia conversations with older workers
February 16, 2019 - Newly developed stem cell technologies show promise for treating PD patients
February 16, 2019 - Collaborative material research could advance self-assembling nanomaterials
February 16, 2019 - Researchers take major step in creating technology that mimics the human brain
February 16, 2019 - Erasing memories associated with cocaine use reduces drug seeking behavior
February 16, 2019 - Artificial intelligence can accurately predict prognosis of ovarian cancer patients
February 16, 2019 - Racial disparities in cancer deaths on the decline for America
February 16, 2019 - FDA authorizes new interoperable insulin pump for children, adults with diabetes
February 16, 2019 - Coexisting Medical Conditions, Smoking Explain PTSD-CVD Link
February 16, 2019 - Skin Cancer Screening: MedlinePlus Lab Test Information
February 16, 2019 - ‘Happiness’ exercises can boost mood in those recovering from substance use disorder
February 16, 2019 - Cell manipulation could soon halt or reverse aging
February 16, 2019 - Pumped Breast Milk Falls Short of Breastfed Version
February 16, 2019 - Men’s porn habits could fuel partners’ eating disorders, study suggests
February 16, 2019 - Rapid progression of age-related diseases may result from formation of vicious cycles
February 16, 2019 - Immune checkpoint molecule protects against future development of cancer
February 16, 2019 - New method produces hydrogels that have properties similar to cells’ environment
February 16, 2019 - $4.1 million funding for heart research on Valentine’s Day
February 16, 2019 - General anesthesia in early infancy unlikely to have lasting effects on developing brains
February 16, 2019 - New breakthroughs for muscular dystrophy research
February 16, 2019 - First Opinion: Embryo editing for higher IQ is a fantasy. Embryo profiling for it is almost here
February 16, 2019 - Vapers develop cancer-related gene deregulation as cigarette smokers
February 16, 2019 - Bringing Antimicrobial Susceptibility Testing (AST) to the Community
February 16, 2019 - Decolonization protocol after hospital discharge can prevent dangerous infections
February 16, 2019 - Children with ASD more likely to face maltreatment, study finds
February 16, 2019 - Study finds genetic vulnerability to use of menthol cigarettes
February 16, 2019 - Promising drug developed to rejuvenate muscle cells
February 16, 2019 - H-RT should be the standard of care for men with low risk prostate cancer, study shows
February 16, 2019 - New technique using patients’ own modified cells could help treat Crohn’s disease
February 16, 2019 - Therapeutic endoscopy has an expanding role in the treatment of IBD
February 16, 2019 - Blood clot discovery could lead to development of better treatments for blood diseases
February 16, 2019 - Intervention can increase exclusive breastfeeding rates
February 16, 2019 - New project explores how gaming technologies can help cancer patients communicate better
February 16, 2019 - Catalyst Biosciences Presents Updated Data from Its Phase 2/3 Trial of Subcutaneous Marzeptacog Alfa (Activated) in Individuals with Hemophilia A or B with Inhibitors at the 12th Annual EAHAD Congress
February 16, 2019 - Rerouting nerves during amputation reduces phantom limb pain before it starts
February 16, 2019 - A Hormone Produced When We Exercise Might Help Fight Alzheimer’s
February 16, 2019 - Millions of British people breathe toxic air travelling to GPs
AI matched, outperformed radiologists in screening X-rays for certain diseases | News Center

AI matched, outperformed radiologists in screening X-rays for certain diseases | News Center

image_pdfDownload PDFimage_print

“Usually, we see AI algorithms that can detect a brain hemorrhage or a wrist fracture — a very narrow scope for single-use cases,” said Matthew Lungren, MD, MPH, assistant professor of radiology. “But here we’re talking about 14 different pathologies analyzed simultaneously, and it’s all through one algorithm.”

The goal, Lungren said, is to eventually leverage these algorithms to reliably and quickly scan a wide range of image-based medical exams for signs of disease without the backup of professional radiologists. And while that may sound disconcerting, the technology could eventually serve as high-quality digital “consultations” to resource-deprived regions of the world that wouldn’t otherwise have access to a radiologist’s expertise. Likewise, there’s an important role for AI in fully developed health care systems too, Lungren added. Algorithms like CheXNeXt could one day expedite care, empowering primary care doctors to make informed decisions about X-ray diagnostics faster, without having to wait for a radiologist.

“We’re seeking opportunities to get our algorithm trained and validated in a variety of settings to explore both its strengths and blind spots,” said graduate student Pranav Rajpurkar. “The algorithm has evaluated over 100,000 X-rays so far, but now we want to know how well it would do if we showed it a million X-rays — and not just from one hospital, but from hospitals around the world.”

A paper detailing the findings of the study was published online Nov. 20 in PLOS Medicine. Lungren and Andrew Ng, PhD, adjunct professor of computer science at Stanford, share senior authorship. Rajpurkar and fellow graduate student Jeremy Irvin are the lead authors.

Practice makes perfect

Lungren and Ng’s diagnostic algorithm has been in development for more than a year. It builds on their work on a previous iteration of the technology that could outperform radiologists when diagnosing pneumonia from a chest X-ray. Now, they’ve boosted the abilities of the algorithm to flag 14 ailments, including masses, enlarged hearts and collapsed lungs. For 11 of the 14 pathologies, the algorithm made diagnoses with the accuracy of radiologists or better.

Back in the summer of 2017, the National Institutes of Health released a set of hundreds of thousands of X-rays. Since then, there’s been a mad dash for computer scientists and radiologists working in artificial intelligence to deliver the best possible algorithm for chest X-ray diagnostics.

We need to be thinking about how far we can push these AI models to improve the lives of patients anywhere in the world.

The scientists used about 112,000 X-rays to train the algorithm. A panel of three radiologists then reviewed a different set of 420 X-rays, one by one, for the 14 pathologies. Their conclusions served as a “ground truth”— a diagnosis that experts agree is the most accurate assessment — for each scan. This set would eventually be used to test how well the algorithm had learned the telltale signs of disease in an X-ray. It also allowed the team of researchers to see how well the algorithm performed compared to the radiologists.

“We treated the algorithm like it was a student; the NIH data set was the material we used to teach the student, and the 420 images were like the final exam,” Lungren said. To further evaluate the performance of the algorithm compared with human experts, the scientists asked an additional nine radiologists from multiple institutions to also take the same “final exam.”

“That’s another factor that elevates this research,” Lungren said. “We weren’t just comparing this against other algorithms out there; we were comparing this model against practicing radiologists.”

What’s more, to read all 420 X-rays, the radiologists took about three hours on average, while the algorithm scanned and diagnosed all pathologies in about 90 seconds.

Next stop: the clinic

Now, Lungren said, his team is working on a subsequent version of CheXNeXt that will bring the researchers even closer to in-clinic testing. The algorithm isn’t ready for that just yet, but Lungren hopes that it will eventually help expedite the X-ray-reading process for doctors diagnosing urgent care or emergency patients who come in with a cough.

“I could see this working in a few ways. The algorithm could triage the X-rays, sorting them into prioritized categories for doctors to review, like normal, abnormal or emergent,” Lungren said. Or the algorithm could sit bedside with primary care doctors for on-demand consultation, he said. In this case, Lungren said, the algorithm could step in to help confirm or cast doubt on a diagnosis. For example, if a patient’s physical exam and lab results were consistent with pneumonia, and the algorithm diagnosed pneumonia on the patient’s X-ray, then that’s a pretty high-confidence diagnosis and the physician could provide care right away for the condition. Importantly, in this scenario, there would be no need to wait for a radiologist. But if the algorithm came up with a different diagnosis, the primary care doctor could take a closer look at the X-ray or consult with a radiologist to make the final call.

“We should be building AI algorithms to be as good or better than the gold standard of human, expert physicians. Now, I’m not expecting AI to replace radiologists any time soon, but we are not truly pushing the limits of this technology if we’re just aiming to enhance existing radiologist workflows,” Lungren said. “Instead, we need to be thinking about how far we can push these AI models to improve the lives of patients anywhere in the world.”

Other Stanford authors of the study are biostatistician Robyn Ball, PhD; undergraduate student Kaylie Zhu; former research assistant Brandon Yang; data scientist Hershel Mehta; research assistants Tony Duan and Daisy Ding; former research assistant Aarti Bagul; professor of radiology and of medicine Curtis Langlotz, PhD; assistant professor of radiology Bhavik Patel, MD; associate professor of radiology Kristen Yeom, MD; research associate Katie Shpanskaya; associate professor of radiology Francis Blackenberg, MD; clinical assistant professor of radiology Jayne Seekins, MD; clinical associate professor of radiology Safwan Halabi, MD; and clinical assistant professor of radiology Evan Zucker, MD.

Researchers from Duke University and from the University of Colorado also contributed to the study.

Lungren is a member of Stanford Bio-X, the Stanford Child Health Research Institute and the Stanford Cancer Institute.

Stanford’s departments of Radiology and of Computer Science along with the Stanford Center for Artificial Intelligence in Medicine & Imaging supported the work.

Tagged with:

About author

Related Articles