Nav: Home

Deep learning algorithm does as well as dermatologists in identifying skin cancer

January 25, 2017

It's scary enough making a doctor's appointment to see if a strange mole could be cancerous. Imagine, then, that you were in that situation while also living far away from the nearest doctor, unable to take time off work and unsure you had the money to cover the cost of the visit. In a scenario like this, an option to receive a diagnosis through your smartphone could be lifesaving.

Universal access to health care was on the minds of computer scientists at Stanford when they set out to create an artificially intelligent diagnosis algorithm for skin cancer. They made a database of nearly 130,000 skin disease images and trained their algorithm to visually diagnose potential cancer. From the very first test, it performed with inspiring accuracy.

"We realized it was feasible, not just to do something well, but as well as a human dermatologist," said Sebastian Thrun, an adjunct professor in the Stanford Artificial Intelligence Laboratory. "That's when our thinking changed. That's when we said, 'Look, this is not just a class project for students, this is an opportunity to do something great for humanity.'"

The final product, the subject of a paper in the Jan. 25 issue of Nature, was tested against 21 board-certified dermatologists. In its diagnoses of skin lesions, which represented the most common and deadliest skin cancers, the algorithm matched the performance of dermatologists.

Why skin cancer

Every year there are about 5.4 million new cases of skin cancer in the United States, and while the five-year survival rate for melanoma detected in its earliest states is around 97 percent, that drops to approximately 14 percent if it's detected in its latest stages. Early detection could likely have an enormous impact on skin cancer outcomes.

Diagnosing skin cancer begins with a visual examination. A dermatologist usually looks at the suspicious lesion with the naked eye and with the aid of a dermatoscope, which is a handheld microscope that provides low-level magnification of the skin. If these methods are inconclusive or lead the dermatologist to believe the lesion is cancerous, a biopsy is the next step.

Bringing this algorithm into the examination process follows a trend in computing that combines visual processing with deep learning, a type of artificial intelligence modeled after neural networks in the brain. Deep learning has a decades-long history in computer science but it only recently has been applied to visual processing tasks, with great success. The essence of machine learning, including deep learning, is that a computer is trained to figure out a problem rather than having the answers programmed into it.

"We made a very powerful machine learning algorithm that learns from data," said Andre Esteva, co-lead author of the paper and a graduate student in the Thrun lab. "Instead of writing into computer code exactly what to look for, you let the algorithm figure it out."

The algorithm was fed each image as raw pixels with an associated disease label. Compared to other methods for training algorithms, this one requires very little processing or sorting of the images prior to classification, allowing the algorithm to work off a wider variety of data.

From cats and dogs to melanomas and carcinomas

Rather than building an algorithm from scratch, the researchers began with an algorithm developed by Google that was already trained to identify 1.28 million images from 1,000 object categories. While it was primed to be able to differentiate cats from dogs, the researchers needed it to know a malignant carcinoma from a benign seborrheic keratosis.

"There's no huge dataset of skin cancer that we can just train our algorithms on, so we had to make our own," said Brett Kuprel, co-lead author of the paper and a graduate student in the Thrun lab. "We gathered images from the internet and worked with the medical school to create a nice taxonomy out of data that was very messy - the labels alone were in several languages, including German, Arabic and Latin."

After going through the necessary translations, the researchers collaborated with dermatologists at Stanford Medicine, as well as Helen M. Blau, professor of microbiology and immunology at Stanford and co-author of the paper. Together, this interdisciplinary team worked to classify the hodgepodge of internet images. Many of these, unlike those taken by medical professionals, were varied in terms of angle, zoom and lighting. In the end, they amassed about 130,000 images of skin lesions representing over 2,000 different diseases.

During testing, the researchers used only high-quality, biopsy-confirmed images provided by the University of Edinburgh and the International Skin Imaging Collaboration Project that represented the most common and deadliest skin cancers -- malignant carcinomas and malignant melanomas. The 21 dermatologists were asked whether, based on each image, they would proceed with biopsy or treatment, or reassure the patient. The researchers evaluated success by how well the dermatologists were able to correctly diagnose both cancerous and non-cancerous lesions in over 370 images.

The algorithm's performance was measured through the creation of a sensitivity-specificity curve, where sensitivity represented its ability to correctly identify malignant lesions and specificity represented its ability to correctly identify benign lesions. It was assessed through three key diagnostic tasks: keratinocyte carcinoma classification, melanoma classification, and melanoma classification when viewed using dermoscopy. In all three tasks, the algorithm matched the performance of the dermatologists with the area under the sensitivity-specificity curve amounting to at least 91 percent of the total area of the graph.

An added advantage of the algorithm is that, unlike a person, the algorithm can be made more or less sensitive, allowing the researchers to tune its response depending on what they want it to assess. This ability to alter the sensitivity hints at the depth and complexity of this algorithm. The underlying architecture of seemingly irrelevant photos -- including cats and dogs -- helps it better evaluate the skin lesion images.

Health care by smartphone

Although this algorithm currently exists on a computer, the team would like to make it smartphone compatible in the near future, bringing reliable skin cancer diagnoses to our fingertips.

"My main eureka moment was when I realized just how ubiquitous smartphones will be," said Esteva. "Everyone will have a supercomputer in their pockets with a number of sensors in it, including a camera. What if we could use it to visually screen for skin cancer? Or other ailments?"

The team believes it will be relatively easy to transition the algorithm to mobile devices but there still needs to be further testing in a real-world clinical setting.

"Advances in computer-aided classification of benign versus malignant skin lesions could greatly assist dermatologists in improved diagnosis for challenging lesions and provide better management options for patients," said Susan Swetter, professor of dermatology and director of the Pigmented Lesion and Melanoma Program at the Stanford Cancer Institute, and co-author of the paper. "However, rigorous prospective validation of the algorithm is necessary before it can be implemented in clinical practice, by practitioners and patients alike."

Even in light of the challenges ahead, the researchers are hopeful that deep learning could someday contribute to visual diagnosis in many medical fields.
-end-
Additional Stanford co-authors of this work include Roberto Novoa, clinical assistant professor of dermatology and of pathology, and Justin Ko, clinical associate professor of dermatology. Thrun is also founder and president of Udacity. Blau is also the director of the Baxter Laboratory for Stem Cell Biology and a member of Stanford Bio-X, the Stanford Cardiovascular Institute, the Child Health Research Institute and the Stanford Cancer Institute.

Stanford University

Related Health Care Articles:

Care management program reduced health care costs in Partners Pioneer ACO
Pesearchers at Partners HealthCare published a study showing that Partners Pioneer ACO not only reduces spending growth, but does this by reducing avoidable hospitalizations for patients with elevated but modifiable risks.
Health care leaders predict patients will lose under President Trump's health care plans
According to a newly released NEJM Catalyst Insights Report, health care executives and industry insiders expect patients -- more than any other stakeholder -- to be the big losers of any comprehensive health care plan from the Trump administration.
The Lancet: The weaponisation of health care: Using people's need for health care as a weapon of war over six years of Syrian conflict
Marking six years since the start of the Syrian conflict (15 March), a study in The Lancet provides new estimates for the number of medical personnel killed: 814 from March 2011 to February 2017.
In the January Health Affairs: Brazil's primary health care expansion
The January issue of Health Affairs includes a study that explores a much-discussed issue in global health: the role of governance in improving health, which is widely recognized as necessary but is difficult to tie to actual outcomes.
Advocacy and community health care models complement research and clinical care
Global lung cancer researchers and patient advocates today emphasized that new models of delivering care and communicating about cancer care play an important role in the fight against lung cancer.
About 1 million Texans gained health care coverage due to Affordable Care Act
Texas has experienced a roughly 6 percentage-point increase in health insurance coverage from the Affordable Care Act, according to new research by experts at Rice University and the Episcopal Health Foundation.
In India, training informal health-care providers improved quality of care
Training informal health-care providers in India improved the quality of health care they offered to patients in rural regions, a new study reports.
Affordable Care Act has improved access to health care, but disparities persist
The Affordable Care Act has substantially decreased the number of uninsured Americans and improved access to health care, though insurance affordability and disparities by geography, race/ethnicity, and income persist.
Integrated team-based care shows potential for improving health care quality, use and costs
Among adults enrolled in an integrated health care system, receipt of primary care at integrated team-based care practices compared with traditional practice management practices was associated with higher rates of some measures of quality of care, lower rates for some measures of acute care utilization, and lower actual payments received by the delivery system, according to a study appearing in the Aug.
Study finds quality of care in VA health care system compares well to other settings
The quality of health care provided to US military veterans in Veterans Affairs (VA) facilities compares favorably with the treatment and services delivered outside the VA.

Related Health Care Reading:

Best Science Podcasts 2019

We have hand picked the best science podcasts for 2019. Sit back and enjoy new science podcasts updated daily from your favorite science news services and scientists.
Now Playing: TED Radio Hour

Climate Crisis
There's no greater threat to humanity than climate change. What can we do to stop the worst consequences? This hour, TED speakers explore how we can save our planet and whether we can do it in time. Guests include climate activist Greta Thunberg, chemical engineer Jennifer Wilcox, research scientist Sean Davis, food innovator Bruce Friedrich, and psychologist Per Espen Stoknes.
Now Playing: Science for the People

#527 Honey I CRISPR'd the Kids
This week we're coming to you from Awesome Con in Washington, D.C. There, host Bethany Brookshire led a panel of three amazing guests to talk about the promise and perils of CRISPR, and what happens now that CRISPR babies have (maybe?) been born. Featuring science writer Tina Saey, molecular biologist Anne Simon, and bioethicist Alan Regenberg. A Nobel Prize winner argues banning CRISPR babies won’t work Geneticists push for a 5-year global ban on gene-edited babies A CRISPR spin-off causes unintended typos in DNA News of the first gene-edited babies ignited a firestorm The researcher who created CRISPR twins defends...