Rapid Grading of Fundus Photographs for Diabetic Retinopathy Using Crowdsourcing

Background

Screening for diabetic retinopathy is both effective and cost-effective, but rates of screening compliance remain suboptimal. As screening improves, new methods to deal with screening data may help reduce the human resource needs. Crowdsourcing has been used in many contexts to harness distributed human intelligence for the completion of small tasks including image categorization.

Objective

Our goal was to develop and validate a novel method for fundus photograph grading.

Methods

An interface for fundus photo classification was developed for the Amazon Mechanical Turk crowdsourcing platform. We posted 19 expert-graded images for grading by Turkers, with 10 repetitions per photo for an initial proof-of-concept (Phase I). Turkers were paid US $0.10 per image. In Phase II, one prototypical image from each of the four grading categories received 500 unique Turker interpretations. Fifty draws of 1-50 Turkers were then used to estimate the variance in accuracy derived from randomly drawn samples of increasing crowd size to determine the minimum number of Turkers needed to produce valid results. In Phase III, the interface was modified to attempt to improve Turker grading.

Results

Across 230 grading instances in the normal versus abnormal arm of Phase I, 187 images (81.3%) were correctly classified by Turkers. Average time to grade each image was 25 seconds, including time to review training images. With the addition of grading categories, time to grade each image increased and percentage of images graded correctly decreased. In Phase II, area under the curve (AUC) of the receiver-operator characteristic (ROC) indicated that sensitivity and specificity were maximized after 7 graders for ratings of normal versus abnormal (AUC=0.98) but was significantly reduced (AUC=0.63) when Turkers were asked to specify the level of severity. With improvements to the interface in Phase III, correctly classified images by the mean Turker grade in four-category grading increased to a maximum of 52.6% (10/19 images) from 26.3% (5/19 images). Throughout all trials, 100% sensitivity for normal versus abnormal was maintained.

Conclusions

With minimal training, the Amazon Mechanical Turk workforce can rapidly and correctly categorize fundus photos of diabetic patients as normal or abnormal, though further refinement of the methodology is needed to improve Turker ratings of the degree of retinopathy. Images were interpreted for a total cost of US $1.10 per eye. Crowdsourcing may offer a novel and inexpensive means to reduce the skilled grader burden and increase screening for diabetic retinopathy.     

Comparison of some biological effects of epidermal growth factor and commercial serum albumin on the induction of alpha-lactalbumin in rat and rabbit mammary explants

Commercial preparations of serum albumin from six species can markedly enhance the prolactin-independent induction of alpha-lactalbumin in mammary explants from pregnant rats, and evoke such induction in the tissue from virgin rats. These effects are similar to those of epidermal growth factor (EGF) reported previously. The stimulatory activity of bovine serum albumin (BSA) resides in a putative impurity in the albumin. Charcoal extraction and gel filtration of the BSA results in complete loss of activity. Of the five milk protein mRNAs studied, only alpha-lactalbumin mRNA is induced by insulin, glucocorticoid and serum albumin in the absence of prolactin. Despite these similarities, the biological effects of serum albumin and EGF on mammary tissue diverge in some respects. They appear to operate by different mechanisms since their effects on the rat system are additive. Furthermore, while both inhibit prolactin-mediated induction of alpha-lactalbumin in rabbit mammary explants, cortisol converts EGF into a stimulatory agent, but merely blocks the inhibitory effect of serum albumin. The results emphasize that commercial serum albumin is not to be regarded simply as an inert protein additive to culture media.     

The risk of fellow eyes in patients with rhegmatogenous retinal detachment

PURPOSE OF REVIEW: A review of the current literature regarding the risk of nontraumatic rhegmatogenous retinal detachment in the fellow eye and prophylaxis of predisposing lesions is presented. RECENT FINDINGS: Although reports vary in study design and inclusion/exclusion criteria, the risk of retinal detachment in fellow eyes is similar to prior findings. Studies on prophylactic therapy also treated different cohorts. This makes drawing definitive conclusions difficult. SUMMARY: Rhegmatogenous retinal detachment poses a significant threat to the fellow eye. Risk factors specifically relevant to the fellow eye are discussed. There is insufficient evidence to recommend prophylactic therapy for prevention of retinal detachment in the fellow eye. To date, no prospective, randomized clinical trial on the prevention of retinal detachment in the fellow eye has been published.     

Dual plating for fractures of the distal third of the humeral shaft

In this study, we present a novel method for performing dual plating of extra-articular fractures of the distal third of the humerus. Since 2006, we have treated 15 such fractures with dual plates from a single posterior midline incision. In the first part of the study, we provide the surgical protocol we have used in addressing these fractures. In the second part, the charts of these patients were reviewed retrospectively to examine their clinical and radiographic outcomes. Using this technique, we have achieved an excellent union rate without significant complications while allowing early and aggressive range of motion.