We propose a methodology for digitally fusing diagnostic decisions made by multiple medical experts in order to improve accuracy of diagnosis. Toward this goal, we report an experimental study involving nine experts, where each one was given more than 8,000 digital microscopic images of individual human red blood cells and asked to identify malaria infected cells. The results of this experiment reveal that even highly trained medical experts are not always self-consistent in their diagnostic decisions and that there exists a fair level of disagreement among experts, even for binary decisions (i.e., infected vs. uninfected). To tackle this general medical diagnosis problem, we propose a probabilistic algorithm to fuse the decisions made by trained medical experts to robustly achieve higher levels of accuracy when compared to individual experts making such decisions. By modelling the decisions of experts as a three component mixture model and solving for the underlying parameters using the Expectation Maximisation algorithm, we demonstrate the efficacy of our approach which significantly improves the overall diagnostic accuracy of malaria infected cells. Additionally, we present a mathematical framework for performing 'slide-level' diagnosis by using individual 'cell-level' diagnosis data, shedding more light on the statistical rules that should govern the routine practice in examination of e.g., thin blood smear samples. This framework could be generalized for various other tele-pathology needs, and can be used by trained experts within an efficient tele-medicine platform.

Background

Obesity and depression are complex conditions with stronger comorbid relationships among women than men. Inflammation and cardiometabolic dysfunction are likely mechanistic candidates for increased depression risk, and their prevalence differs by sex. Whether these relationships extend to depressive symptoms is poorly understood. Therefore, we analyzed sex in associations between inflammation and metabolic syndrome (MetS) criteria on depressive symptomatology. Specifically, we examined whether sex positively moderates the relationship between depressive symptoms and inflammation among women, and whether MetS has parallel effects among men.

Methods

Depressive symptoms, MetS, and inflammation were assessed in 129 otherwise healthy adults. Depressive symptoms were assessed using Beck Depression Inventory (BDI-Ia). Monocyte inflammation regulation (BARIC) was quantified using flow cytometry measurement of TNF-α suppression by β-agonist. Moderation effects of sex on associations between BARIC, MetS criteria, and BDI were estimated using two-way ANOVA and linear regression, adjusting for BMI, and by sex subgroup analyses.

Results

Obese individuals reported more depressive symptoms. Sex did not formally moderate this relationship, though BDI scores tended to differ by BMI among women, but not men, in subgroup analysis. Poorer inflammation control and higher MetS criteria were correlated with somatic depressive symptoms. Sex moderated associations between MetS criteria and somatic symptoms; among men, MetS criteria predicted somatic symptoms, not among women. Subgroup analysis further indicated that poorer inflammation control tended to be associated with higher somatic symptoms in women.

Conclusions

These results indicate that obesity-related inflammation and MetS factors have sex-specific effects on depressive symptoms in a non-clinical population. Although pathophysiological mechanisms underlying sex differences remain to be elucidated, our findings suggest that distinct vulnerabilities to depressive symptoms exist between women and men, and highlight the need to consider sex as a key biological variable in obesity-depression relationships. Future clinical studies on comorbid obesity and depression should account for sex, which may optimize therapeutic strategies.

In this work we investigate whether the innate visual recognition and learning capabilities of untrained humans can be used in conducting reliable microscopic analysis of biomedical samples toward diagnosis. For this purpose, we designed entertaining digital games that are interfaced with artificial learning and processing back-ends to demonstrate that in the case of binary medical diagnostics decisions (e.g., infected vs. uninfected), with the use of crowd-sourced games it is possible to approach the accuracy of medical experts in making such diagnoses. Specifically, using non-expert gamers we report diagnosis of malaria infected red blood cells with an accuracy that is within 1.25% of the diagnostics decisions made by a trained medical professional.