Algorithmic discrimination: examining its types and regulatory measures with emphasis on US legal practices.

Introduction: Algorithmic decision-making systems are widely used in various sectors, including criminal justice, employment, and education. While these systems are celebrated for their potential to enhance efficiency and objectivity, they also pose risks of perpetuating and amplifying societal biases and discrimination. This paper aims to provide an indepth analysis of the types of algorithmic discrimination, exploring both the challenges and potential solutions. Methods: The methodology includes a systematic literature review, analysis of legal documents, and comparative case studies across different geographic regions and sectors. This multifaceted approach allows for a thorough exploration of the complexity of algorithmic bias and its regulation. Results: We identify five primary types of algorithmic bias: bias by algorithmic agents, discrimination based on feature selection, proxy discrimination, disparate impact, and targeted advertising. The analysis of the U.S. legal and regulatory framework reveals a landscape of principled regulations, preventive controls, consequential liability, self-regulation, and heteronomy regulation. A comparative perspective is also provided by examining the status of algorithmic fairness in the EU, Canada, Australia, and Asia. Conclusion: Real-world impacts are demonstrated through case studies focusing on criminal risk assessments and hiring algorithms, illustrating the tangible effects of algorithmic discrimination. The paper concludes with recommendations for interdisciplinary research, proactive policy development, public awareness, and ongoing monitoring to promote fairness and accountability in algorithmic decision-making. As the use of AI and automated systems expands globally, this work highlights the importance of developing comprehensive, adaptive approaches to combat algorithmic discrimination and ensure the socially responsible deployment of these powerful technologies.

Unlabeled data selection for active learning in image classification.

Active Learning has emerged as a viable solution for addressing the challenge of labeling extensive amounts of data in data-intensive applications such as computer vision and neural machine translation. The main objective of Active Learning is to automatically identify a subset of unlabeled data samples for annotation. This identification process is based on an acquisition function that assesses the value of each sample for model training. In the context of computer vision, image classification is a crucial task that typically requires a substantial training dataset. This research paper introduces innovative selection methods within the Active Learning framework, aiming to identify informative images from unlabeled datasets while minimizing the number of required training data. The proposed methods, namely Similari-ty-based Selection, Prediction Probability-based Selection, and Competence-based Active Learning, have been extensively evaluated through experiments conducted on popular datasets like Cifar10 and Cifar100. The experimental results demonstrate that the proposed methods outperform random selection and conventional selection techniques. The superior performance of the novel selection methods underscores their effectiveness in enhancing the Active Learning process for image classification tasks.

Digital intervention in improving the outcomes of mental health among LGBTQ+ youth: a systematic review.

LGBTQ+ youth experience mental health disparities and higher rates of mental disorders due to barriers to accessing care, including insufficient services and the anticipated stigma of revealing their identities. This systematic review incorporated 15 empirical studies on digital interventions' impact on LGBTQ+ youth mental health, examining their potential to address these inequities. This study innovatively categorized existing digital interventions into four streams: Structured Formal (telehealth, online programs), Structured Informal (serious games), Unstructured Formal (mobile applications), and Unstructured Informal (social media). We found that S&F and U&F effectively reduced symptoms. U&F showed potential but required enhancement, while U&I fostered resilience but posed risks. Further integration of emerging technologies like virtual reality may strengthen these interventions. This review identifies the characteristics of effective digital health interventions and evaluates the overall potential of digital technologies in improving LGBTQ+ youth mental health, uniquely contributing insights on digital solutions advancing LGBTQ+ youth mental healthcare.

Meta-analysis of Influencing Factors ofF-Fluorodeoxyglucose Uptake of Brown Adipose Tissue in PET/CT Imaging / 中国医学科学院学报

Objective To explore the influencing factors ofF-Fluorodeoxyglucose(F-FDG) uptake of brown adipose tissue(BAT) in PET/CT imaging by performing meta-analysis.Methods Relevant articles published between 1980 and 2016 were searched in PubMed,EMBASE,Web of Science,CBM,WanFang Data,CNKI,and VIP.The factors of gender,season and age were extracted according to the inclusive criteria.Meta-analysis was performed to analyze the factors.Results The merger OR value and 95% CI ofF-FDG uptake of BAT with gender,age,season were 2.67 (2.30-3.11),2.17 (1.38-3.39),and 15.35 (4.72-49.95) respectively.TheF-FDG uptake rate of BAT was 2.16 times in females as that in males,8.67 times in the minors as that in the adults,and 1.94 times in winter as that in summer.Conclusion Gender,season,and age are risk factors forF-FDG uptake of BAT.

Patterns and characteristics of brown adipose tissue uptake of 18F-FDG positron emission tomograph/computed tomography imaging / 中国医学科学院学报

<p><b>OBJECTIVE</b>To characterize the brown adipose tissue (BAT) uptake of 18F-FDG on positron emission tomography (PET)/computed tomography (CT) imaging.</p><p><b>METHODS</b>We retrospectively analyzed the clinical data of 1 080 patients who received all whole-body PET/CT studies in Peking Union Medica College Hospital from July 2008 to February 2009.</p><p><b>RESULTS</b>Forty-one patients (3.8%) were identified to be with BAT uptake, especially during cold seasons. BAT uptake was mostly observed at the neck regions symmetrically (n = 39), and was also seen at paravertebral junctions (n = 30), perinephric regions (n = 21), and mediastinum (n = 10). Patients with BAT uptake had significantly lower male/female ratio (P = 0.0030), younger age (P = 0.0001), and less body mass index (BMI) (P = 0.0415). Three patients with high BAT uptake underwent repeat PET/CT scans 3-7 days later. By keeping warm and full relaxation, BAT uptake disappeared in 2 cases and dramatically decreased in the other case.</p><p><b>CONCLUSIONS</b>BAT uptake commonly occurs during cold seasons in Beijing, especially in young females with low BMI. It usually has specific patterns, and can be avoided or remarkably reduced by asking the patients keep warm and full relaxation a few days before the scanning.</p>

Characteristics of physiological uptake of uterus and ovaries on 18F-fluorodeoxyglucose positron emission tomography / 中国医学科学院学报

<p><b>OBJECTIVE</b>To observe the characteristics of the physiological uptake of uterus and ovaries on 18F-fluorodeoxyglucose positron emission tomography (FDG PET).</p><p><b>METHODS</b>A total of 288 PET examinations performed in 247 women (164 with malignancies, 44 with benign diseases, and 39 without remarkable abnormality) were included for analysis, and clinical follow-ups were applied for at least 10 months to exclude pelvic diseases. The menstrual statuses, menstrual cycles, and related pelvic examinations with other modalities were inquired before each PET examination. PET scanning was performed from pelvis to neck with a Siemens ECAT EXACT HR + system. The uptake levels of uterus and ovaries were set as intense, moderate, and mild by comparing to liver uptake.</p><p><b>RESULTS</b>In 116 patients (131 examinations ) with regular menstruation, the endometrial uptake, usually in inverted cone shape surrounded by relatively low-uptake uterine wall, was observed with two peaks in the early menstrual flow phase and in the mid-cycle respectively; the ovarian uptake was more prominent in the mid-cycle, with the foci of uptake in ovoidal shape and located at the left and/or right side superior-posterior to the bladder. From the early menstrual flow phase to the late secretory phase of the menstrual cycles, the probabilities of mild uptake in both endometrium and ovaries were 7%, 86%, 80%, 58%, 20%, 40%, 64%, and 59%, respectively, indicating that the late menstrual flow phase and the early proliferative phase had the least probability of intense or moderate uptake. No intense uptake was observed in the 17 patients (19 examinations) presenting remarkably irregular menstrual cycle, 112 patients (136 studies) in menopause for 3 months to 39 years, and 2 patients without menstruation yet. Only one patient within 1 year of menopause and a 14-year-old girl expected to start menstruation showed mild to moderate uptake in the endometrium.</p><p><b>CONCLUSION</b>The physiological endometrial and ovarian uptakes have specific shapes and positions on 18F-FDG PET images, which correlates well with the menstrual phases.</p>