Breaking News
October 21, 2018 - More accurate and less stressful way to measure a baby’s heartbeat
October 21, 2018 - Researchers show better cardiorespiratory fitness leads to longer life
October 21, 2018 - Healthy candies for diabetic patients
October 21, 2018 - Environment impact of microplastics remains unclear
October 21, 2018 - Antibiotics for appendicitis? Surgery often not needed
October 21, 2018 - AHA and AMA recognize more than 800 medical practices, health systems for blood pressure control
October 21, 2018 - Scientists obtain clearest ever image of Ebola virus protein
October 21, 2018 - Study reveals connection between two proteins known to be hyperactive in cancer
October 21, 2018 - Gabapentin Beats Pregabalin for Chronic Sciatica
October 21, 2018 - Cosmetic surgeons offering incomplete information for breast augmentation customers
October 21, 2018 - Chronic sleep disruption in early adult life accelerates AD-related tau pathology
October 21, 2018 - Take 10 for Mindfulness – Drugs.com MedNews
October 21, 2018 - Length of breathing disruption in OSA may be better predictor of mortality risk
October 21, 2018 - ApoE4 gene linked with chronic inflammation increases risk for Alzheimer’s disease
October 21, 2018 - Mother-daughter conflict associated with suicide risk in abused adolescent girls
October 21, 2018 - Scientists molding bacteria into unnatural shapes
October 21, 2018 - Discharged mental health patients ‘at greater risk of dying’
October 21, 2018 - As billions in tax dollars flow to private Medicaid plans, Who’s minding the store?
October 21, 2018 - Neuroscientists identify brain region that appears to be related to food preference decisions
October 21, 2018 - Deaths related to air pollution in the U.S. decreased by 47% between 1990 and 2010
October 21, 2018 - Study shows correlation between spatial memory and the sense of smell
October 21, 2018 - Increased cardiorespiratory fitness associated with reduced long-term mortality
October 21, 2018 - IU researchers receive $1.55 million from NIH to improve chronic-disease management
October 21, 2018 - Income and wealth affect the mental health of Australians, study shows
October 21, 2018 - Patients with hypertension and psoriasis more often require cardiovascular interventions
October 20, 2018 - Leading hip-hop videos depict use of tobacco and marijuana products, study finds
October 20, 2018 - Dose Range of IV Ketamine for Adjunct Tx of Depression Tested
October 20, 2018 - Infants can distinguish between leaders and bullies, study finds
October 20, 2018 - Mad Cow disease found on Aberdeenshire farm
October 20, 2018 - Study identifies factors associated with prescription opioid misuse among students
October 20, 2018 - Scientists uncover key regulator of mTORC1 in cancer growth
October 20, 2018 - Pounds Regained After Weight-Loss Op Can Tell Your Doc a Lot
October 20, 2018 - Sending parents letters to fight childhood obesity doesn’t work
October 20, 2018 - Supervised aerobic exercise can support major depression treatment
October 20, 2018 - Mindfulness-based program effective for reducing stress in infertile women
October 20, 2018 - Molecule capable of halting and reverting neurodegeneration caused by Parkinson’s disease identified
October 20, 2018 - Midazolam-mediated alterations of PER2 expression may have functional consequences during myocardial ischemia
October 20, 2018 - Sweat bees are ideal for studying the genes underlying social behavior
October 20, 2018 - Weight loss success associated with brain areas involved in self-control
October 20, 2018 - KHN’s ‘What the Health?’ Republicans’ preexisting political problem
October 20, 2018 - Research provides a more complete picture of suffering caused by terrorist attacks
October 20, 2018 - Eradicating Helicobacter pylori infections may be a key treatment for Parkinson’s disease
October 20, 2018 - Breast Cancer as a Dynamic Disease
October 20, 2018 - University of Pittsburgh wins NSF grant for big data research to prevent complications from anesthesia
October 20, 2018 - Skin-to-skin contact may promote attachment between parents and preterm infants
October 20, 2018 - Recommendations Developed to Verify NGT Placement in Children
October 20, 2018 - Weight loss can be boosted fivefold thanks to novel mental imagery technique
October 20, 2018 - Children with autism are more likely to be overweight, obese
October 20, 2018 - Nurses making conscientious objections to ethically-relevant policies lack support
October 20, 2018 - Prion strain diversity may be greater than previously thought
October 20, 2018 - Antidepressant treatment may lead to improvements in sleep quality of patients with depression
October 20, 2018 - Study reports increased risk of death in children with inflammatory bowel disease
October 20, 2018 - Number of Autism Genes Now Tops 100
October 20, 2018 - Total diet replacement programmes are effective for treating obesity
October 20, 2018 - CLARIOstar used for fluorescence measurements on CSIRO’s purpose-built research vessel
October 20, 2018 - People with more copies of AMY1 gene digest starchy carbohydrates faster
October 20, 2018 - Case Comprehensive Cancer Center wins NIH grant to study health disparities
October 20, 2018 - Newly discovered compound shows potential for treating Parkinson’s disease
October 20, 2018 - High rate of non-adherence to hormonal therapy found among premenopausal early breast cancer patients
October 20, 2018 - Immunotherapy medicine found to be effective in treating uveitis
October 20, 2018 - The Pistoia Alliance Calls for Greater Collaboration to Realise Benefits of Innovation and Announces Winners of the 2018 President’s Startup Challenge
October 20, 2018 - Female internists consistently earn less than men
October 20, 2018 - Stanford team looks at dangers of teens’ vaping habits
October 20, 2018 - New approach to understanding cancers will accelerate development of better treatments
October 20, 2018 - LJI and UC San Diego awarded $ 4.5 million as part of NCI’s Cancer Moonshot initiative
October 20, 2018 - School-based HPV vaccination did not increase risky sexual behaviors among adolescent girls
October 20, 2018 - Eye discovery to pave way for more successful corneal transplants
October 20, 2018 - New analysis examines the importance of location in the opioid crisis
October 20, 2018 - Green filters increase reading speed for children with dyslexia
October 19, 2018 - Bariatric Sx Cuts Macrovascular Complications in Obesity, T2DM
October 19, 2018 - Better assessments for early age-related macular degeneration
October 19, 2018 - Visible and valued: Stanford Medicine’s first-ever LGBTQ+ Forum | News Center
October 19, 2018 - Understanding of metal-free enzymes used by bacteria could lead to new effective antibiotics
October 19, 2018 - Beckman Coulter Life Sciences announces new research-focused website
October 19, 2018 - Study finds link between refined soluble fibers, gut microbiota and liver cancer
October 19, 2018 - Social media reduces risk of depression among seniors with pain
October 19, 2018 - Newly developed synthetic DNA molecule may one day be used as ‘vaccine’ for prostate cancer
October 19, 2018 - Preoperative weight loss may not provide health benefits after surgery
October 19, 2018 - U.S. Birth Rates Continue to Drop as Age of New Moms Rises
October 19, 2018 - New technology can keep an eye on babies’ movements in the womb
Researchers determine ‘sweet spot’ could enhance diagnosis of melanoma skin lesions

Researchers determine ‘sweet spot’ could enhance diagnosis of melanoma skin lesions

image_pdfDownload PDFimage_print

Too much, too little, just right. It might seem like a line from “Goldilocks and the Three Bears,” but actually describes an important finding from researchers in Florida Atlantic University’s College of Engineering and Computer Science. They have developed a technique using machine learning – a sub-field of artificial intelligence (AI) – that will enhance computer-aided diagnosis (CADx) of melanoma. Thanks to the algorithm they created – which can be used in mobile apps that are being developed to diagnose suspicious moles – they were able to determine the “sweet spot” in classifying images of skin lesions.

This new finding, published in the Journal of Digital Imaging, will ultimately help clinicians more reliably identify and diagnose melanoma skin lesions, distinguishing them from other types of skin lesions. The research was conducted in the NSF Industry/University Cooperative Research Center for Advanced Knowledge Enablement (CAKE) at FAU and was funded by the Center’s industry members.

Melanoma is a particularly deadly form of skin cancer when left undiagnosed. In the United States alone, there were an estimated 76,380 new cases of melanoma and an estimated 6,750 deaths due to melanoma in 2016. Malignant melanoma and benign skin lesions often appear very similar to the untrained eye. Over the years, dermatologists have developed different heuristic classification methods to diagnose melanoma, but to limited success (65 to 80 percent accuracy). As a result, computer scientists and doctors have teamed up to develop CADx tools capable of aiding physicians to classify skin lesions, which could potentially save numerous lives each year.

“Contemporary CADx systems are powered by deep learning architectures, which basically consist of a method used to train computers to perform an intelligent task. We feed them massive amounts of data so that they can learn to extract meaning from the data and, eventually, demonstrate performance comparable to human experts – dermatologists in this case,” said Oge Marques, Ph.D., lead author of the study and a professor in FAU’s Department of Computer and Electrical Engineering and Computer Science who specializes in machine cognition, medical imaging and health care technologies. “We are not trying to replace physicians or other medical professionals with artificial intelligence. We are trying to help them solve a problem faster and with greater accuracy, in other words enabling augmented intelligence.”

Images of skin lesions often contain more than just skin lesions – background noise, hair, scars, and other artifacts in the image can potentially confuse the CADx system. To prevent the classifier from incorrectly associating these irrelevant artifacts with melanoma, the images are segmented into two parts, separating the lesion from the surrounding skin, hoping that the segmented lesion can be more easily analyzed and classified.

“Previous studies have produced conflicting results: some research suggests that segmentation improves classification while other research suggests that segmentation is detrimental, due to a loss of contextual information around the lesion area,” said Marques. “How much we segment an image can either help or impede skin lesion classification.”

With that in mind, Marques and his collaborators Borko Furht, Ph.D., a professor in FAU’s Department of Electrical and Computer Engineering and Computer Science and director of the NSF-sponsored CAKE at FAU; Jack Burdick, a second-year master’s student at FAU; and Janet Weinthal, an undergraduate student at FAU, tested their hypothesis: “How much segmentation is too much?”

To test their hypothesis, the researchers first compared the effects of no segmentation, full segmentation, and partial segmentation on classification and demonstrated that partial segmentation led to the best results. They then proceeded to determine how much segmentation would be “just right.” To do that, they used three degrees of partial segmentation, investigating how a variable-sized non-lesion border around the segmented skin lesion affects classification results. They performed comparisons in a systematic and reproducible manner to demonstrate empirically that a certain amount of segmentation border around the lesion could improve classification performance.

Their findings suggest that extending the border beyond the lesion to include a limited amount of background pixels improves their classifier’s ability to distinguish melanoma from a benign skin lesion.

“Our experimental results suggest that there appears to be a ‘sweet spot’ in the degree to which the surrounding skin included is neither too great nor too small and provides a ‘just right’ amount of context,” said Marques.

Their method showed an improvement across all relevant measures of performance for a skin lesion classifier.

Source:

http://www.fau.edu/newsdesk/articles/melanoma-diagnosis.php

Tagged with:

About author

Related Articles