管理学最新文献

Objective: This study aimed to develop a novel multi-stage self-supervised learning model tailored for the accurate classification of optical coherence tomography (OCT) images in ophthalmology reducing reliance on costly labeled datasets while maintaining high diagnostic accuracy.

Materials and methods: A private dataset of 2719 OCT images from 493 patients was employed, along with 3 public datasets comprising 84 484 images from 4686 patients, 3231 images from 45 patients, and 572 images. Extensive internal, external, and clinical validation were performed to assess model performance. Grad-CAM was employed for qualitative analysis to interpret the model's decisions by highlighting relevant areas. Subsampling analyses evaluated the model's robustness with varying labeled data availability.

Results: The proposed model outperformed conventional supervised or self-supervised learning-based models, achieving state-of-the-art results across 3 public datasets. In a clinical validation, the model exhibited up to 17.50% higher accuracy and 17.53% higher macro F-1 score than a supervised learning-based model under limited training data.

Discussion: The model's robustness in OCT image classification underscores the potential of the multi-stage self-supervised learning to address challenges associated with limited labeled data. The availability of source codes and pre-trained models promotes the use of this model in a variety of clinical settings, facilitating broader adoption.

Conclusion: This model offers a promising solution for advancing OCT image classification, achieving high accuracy while reducing the cost of extensive expert annotation and potentially streamlining clinical workflows, thereby supporting more efficient patient management.

Empirical studies of two-sided platform markets, like the video game console industry, typically rely on software and platform sales data, thereby overlooking today’s managerial focus on engagement. This present research leverages a unique dataset tracking the daily engagement of over 14,000 users of Microsoft’s Xbox One and Xbox Series video game platforms to remedy this gap. We investigate how software development and release characteristics affect consumers’ engagement with software titles and the platforms on which they release. Our analysis finds that releasing software on subscription services is the strongest determinant of engagement, overshadowing established determinants like software quality or exclusivity. While superstar software and exclusive titles generate engagement, their relative importance is smaller compared to sales-based findings, reported in prior literature. Instead, franchises, non-superstars, and multihomed software perform much better on engagement than on sales, especially when included in a subscription service. These findings have important industry implications. 

The 3.11 disaster in Fukushima Prefecture, Japan, is a useful case study through which to interrogate research ethics. This region has been the site of a high degree of research interest, which sometimes presented a source of stress to local communities. This study examines researcher perspectives on the ethics of post-disaster health research. Qualitative interviews were conducted with these informants, who noted that recovering communities experienced significant over-research, particularly in the form of survey fatigue, which was seen to influence viewpoints concerning both recovery and agency. Efforts to integrate better into community needs reoriented reflexive research towards ‘post-normal’ forms of working. Simultaneously, researchers had to navigate funding and reward structures that prioritised the swift production of results. Focusing on community engagement and feedback, and managing this ethical complexity, were seen as essential forms of ethical practice to mitigate the negative impacts that the influx of research activity can have on a recovering community.

Financial literacy programs aim to prevent consumer overspending by teaching and encouraging fiscally sound habits (purchase restraint, responsible credit use, savings). Unfortunately, trying not to spend is at odds with the emotions consumers experience in a tempting marketplace. The theory of trying considers attitudes and intentions, but not emotions, when trying to consume. To address this gap, we examine indebted consumers opting into formal financial literacy training explicitly designed for debt repayment and avoidance of future debt. Through indebted consumers’ diary reflections and interviews with clients and debt management counselors, we show that financial literacy’s emphasis on budgeting needs versus wants is not sufficient when consumers try not to spend. To reconcile budgets with actual purchasing behavior when faced with temptations in the marketplace, consumers often adopt a linguistic exercise of imaginatively bending and blending utilitarian and hedonic discourses to justify purchases by recategorizing wants as needs. Further, consumers trying not to spend experience negative emotions; how they regulate those emotions impacts their success in getting out of debt. While financial literacy courses only give consumers budget-setting tools, indebted consumers cannot be successful without tools for trying not to spend in the marketplace.