Aunt Minnie (6/29) reports researchers have used artificial intelligence (AI) technology to “analyze how radiologists read mammograms,” according to a study published in the Journal of Medical Imaging. The researchers from Oak Ridge National Laboratory “used an artificial intelligence tool to assess the level of bias” – such as context bias and the radiologists’ previous diagnostic experiences – in mammography interpretations by tracking the “eye movements of three board-certified radiologists and seven radiology residents.” The study “found that the radiologists’ analyses of mammograms were significantly influenced by context bias, and while radiology trainees were perhaps most vulnerable to the phenomenon, even more-experienced radiologists fell victim to some degree.”
The Radiology Business Journal (6/29, Walter) reported that research suggests “breast imaging specialists provide significant value to patients when conducting second-opinion review of imaging studies, even before the patient is diagnosed with breast cancer.” The researchers wrote, “Overall, 51 percent of patients who sought a second opinion by a breast imaging specialist at our cancer center received a significant interpretation change.” The findings were published in the Journal of the American College of Radiology.
Ruth Margalit writes in the New Yorker (6/28) that many Palestinian women from Gaza and the West Bank seeking treatment for cancer in Israel are no longer able to receive care. Since the beginning of the year, “with no announcement of a change in policy, more than half of applications for medical permits from Gaza have been turned down or left unanswered, according to Physicians for Human Rights – Israel,” although in 2012 “Israel approved ninety-two per cent of medical permits for Gazans.” Margalit says that for women who are refused permission for treatment in Israel, the “options in Gaza are dismal: its public hospitals have no MRI machines and only a handful of functioning mammogram machines, so...[they have] no way of receiving a diagnosis, let alone treatment.”
Aunt Minnie (6/28) reports, “Social factors such as race and economic position influence women’s vulnerability to breast cancer, but they are rarely considered in breast cancer research, according to a commentary published in...Cancer Causes and Control.” Aunt Minnie adds, “Not enough is being done to understand how race, income level, and other social factors contribute to cancer susceptibility, wrote a team led by Lorraine Dean, ScD.” The authors “highlighted a 2014 review of more than 20 years of U.S. National Cancer Institute (NCI) clinical trials that found only 20% of randomized controlled studies reported results that took race and ethnicity into account.”
HealthImaging (6/28, O'Connor) also covers the story.
The Radiology Business Journal (6/27, Walter) reports on a study published in Current Problems in Diagnostic Radiology finding that “management recommendations for high-risk lesions detected by a core needle biopsy of the breast are inconsistent.” The study was based on a survey of “breast imagers from 41 academic institutions throughout the United States” of “their own management recommendations when core needle biopsy of the breast detects a high-risk lesion.” Overall, “ninety-five percent of respondents recommended surgical excision for atypical ductal hyperplasia (ADH) and papilloma with associated atypia.” Surgical excision was also recommended by majorities for flat epithelial atypia (76%), radial scar/complex sclerosing lesion (73%), lobular carcinoma in-situ (71%), and atypical lobular hyperplasia (61%) while “39 percent recommended it for intraductal papilloma without atypia.” The researchers suggest that developing a consensus response to all of these diagnoses might improve patient care.