ISLE: An Intelligent Streaming Framework for High-Throughput AI Inference in Medical Imaging.

As the adoption of artificial intelligence (AI) systems in radiology grows, the increase in demand for greater bandwidth and computational resources can lead to greater infrastructural costs for healthcare providers and AI vendors. To that end, we developed ISLE, an intelligent streaming framework to address inefficiencies in current imaging infrastructures. Our framework draws inspiration from video-on-demand platforms to intelligently stream medical images to AI vendors at an optimal resolution for inference from a single high-resolution copy using progressive encoding. We hypothesize that ISLE can dramatically reduce the bandwidth and computational requirements for AI inference, while increasing throughput (i.e., the number of scans processed by the AI system per second). We evaluate our framework by streaming chest X-rays for classification and abdomen CT scans for liver and spleen segmentation and comparing them with the original versions of each dataset. For classification, our results show that ISLE reduced data transmission and decoding time by at least 92% and 88%, respectively, while increasing throughput by more than 3.72 × . For both segmentation tasks, ISLE reduced data transmission and decoding time by at least 82% and 88%, respectively, while increasing throughput by more than 2.9 × . In all three tasks, the ISLE streamed data had no impact on the AI system's diagnostic performance (all P > 0.05). Therefore, our results indicate that our framework can address inefficiencies in current imaging infrastructures by improving data and computational efficiency of AI deployments in the clinical environment without impacting clinical decision-making using AI systems.

Efficacy of Low-Dose Streptokinase Infusion in Late-Onset Permanent Tunnel Catheter Dysfunction: A Single-Center Interventional Study.

Introduction Hemodialysis is a vital modality for patients with renal dysfunction, with venous access being a significant factor in its success. While arteriovenous fistulas are preferred, tunneled catheters serve as important alternatives, especially in challenging cases. Late-onset tunneled catheter dysfunction, often due to fibrin sheath formation, impedes hemodialysis efficiency. Streptokinase, a low-cost thrombolytic agent, has shown promise in resolving such complications, yet its efficacy in the Indian context remains unexplored. Methods We conducted a single-center interventional study at Mahatma Gandhi Mission (MGM) Hospital, Aurangabad, India, from May 2023 to October 2023. Ethical approval was obtained, and 10 eligible patients experiencing late-onset permanent tunnel catheter dysfunction were enrolled. Patients were treated with low-dose streptokinase, and outcomes were monitored for 60 days. Results Ten patients, evenly distributed by gender, participated, with a mean age of 48.2 ± 11.96 years. Diabetes was the predominant cause of chronic kidney disease (CKD) at 33% (3/10). All patients achieved the primary endpoint of blood flow rate (BFR) >300 ml/min post-streptokinase treatment, with an overall success rate of 100%. Group A had the highest average gain in catheter days (80.6 ± 7.59), followed by Group B (64 ± 1), while Group C showed variations in catheter days between the first (26.2 ± 6.8) and second insertion (32.5 ± 1.76). Eight patients maintained catheter patency during the 60-day follow-up. Adverse effects, primarily minor, were observed. The dosage rationale involved an eight-hour infusion at 4,000 units per hour. Conclusion Streptokinase emerges as cost-effective and efficacious for maintaining the patency of late-onset tunnel catheter dysfunction in resource-limited settings, particularly in younger patients. Caution is advised for older individuals with prolonged CKD.

Conventional and frugal methods of estimating COVID-19-related excess deaths and undercount factors.

Across the world, the officially reported number of COVID-19 deaths is likely an undercount. Establishing true mortality is key to improving data transparency and strengthening public health systems to tackle future disease outbreaks. In this study, we estimated excess deaths during the COVID-19 pandemic in the Pune region of India. Excess deaths are defined as the number of additional deaths relative to those expected from pre-COVID-19-pandemic trends. We integrated data from: (a) epidemiological modeling using pre-pandemic all-cause mortality data, (b) discrepancies between media-reported death compensation claims and official reported mortality, and (c) the "wisdom of crowds" public surveying. Our results point to an estimated 14,770 excess deaths [95% CI 9820-22,790] in Pune from March 2020 to December 2021, of which 9093 were officially counted as COVID-19 deaths. We further calculated the undercount factor-the ratio of excess deaths to officially reported COVID-19 deaths. Our results point to an estimated undercount factor of 1.6 [95% CI 1.1-2.5]. Besides providing similar conclusions about excess deaths estimates across different methods, our study demonstrates the utility of frugal methods such as the analysis of death compensation claims and the wisdom of crowds in estimating excess mortality.

A Single-Center Study on the Clinical Profiles and Ultrasonographic Assessments of Living Kidney Donors in the Marathwada Region of Maharashtra.

Living donor kidney transplantation plays a vital role in renal replacement therapy, particularly in India, where a substantial increase in kidney transplants has been observed. Thorough assessments of living kidney donors are crucial, focusing on parameters such as kidney size and glomerular filtration rate (GFR). Despite the importance of GFR in donor assessments, there is a noticeable lack of data on normal GFR ranges in the Indian population. This study aims to address the gap in knowledge by establishing a reference range for GFR in healthy kidney donors from the Marathwada region of Maharashtra. The research also explores the clinical profiles and ultrasonographic features of living kidney donors. A retrospective analysis was conducted at the Mahatma Gandhi Mission (MGM) Medical College and Hospital in Aurangabad, involving 134 living kidney donors. Inclusion criteria encompassed healthy donors with a BMI of less than 30 kg/m², while donors with uncontrolled hypertension, diabetes, microalbuminuria, or a measured GFR below 70 mL/min/1.73 m² were excluded. Comprehensive medical histories, demographic parameters, and ultrasonographic assessments were conducted, with GFR measured using 99M technetium diethylenetriamine pentaacetate scans. The study reveals that the majority of donors were females (80.6%), and the highest number fell within the 41-50 age group. Parents constituted the primary donor category (68.7%), reflecting a familial inclination toward organ donation. Ultrasonographic assessments indicated larger kidney sizes compared to other studies, suggesting regional or population-specific differences. The mean GFR for the right and left kidneys, as well as the total GFR, was within the expected range. The negative correlation between age and GFR emphasizes the need to consider age in donor assessments. The findings emphasize the unique features of this population, including a higher average age, female preponderance, and larger kidney sizes. The study contributes to the understanding of living kidney donors' profiles in the region and highlights the importance of individualized assessments in the donor selection process.

Economic and Environmental Costs of Cloud Technologies for Medical Imaging and Radiology Artificial Intelligence.

Radiology is on the verge of a technological revolution driven by artificial intelligence (including large language models), which requires robust computing and storage capabilities, often beyond the capacity of current non-cloud-based informatics systems. The cloud presents a potential solution for radiology, and we should weigh its economic and environmental implications. Recently, cloud technologies have become a cost-effective strategy by providing necessary infrastructure while reducing expenditures associated with hardware ownership, maintenance, and upgrades. Simultaneously, given the optimized energy consumption in modern cloud data centers, this transition is expected to reduce the environmental footprint of radiologic operations. The path to cloud integration comes with its own challenges, and radiology informatics leaders must consider elements such as cloud architectural choices, pricing, data security, uptime service agreements, user training and support, and broader interoperability. With the increasing importance of data-driven tools in radiology, understanding and navigating the cloud landscape will be essential for the future of radiology and its various stakeholders.

Coarse Race and Ethnicity Labels Mask Granular Underdiagnosis Disparities in Deep Learning Models for Chest Radiograph Diagnosis.

See also the editorial by Nikolic in this issue.

Overcoming Challenges and Innovations in Orthopedic Prosthesis Design: An Interdisciplinary Perspective.

Recent advances in the orthopedic prostheses design have significantly improved the quality of life for individuals with orthopedic disabilities. However, there are still critical challenges that need to be addressed to further enhance the functionality of orthopedic prostheses improving biocompatibility to promote better integration with natural tissues, enhancing durability to withstand the demands of daily use, and improving sensory feedback for better control of movement are the most pressing issues. To address these challenges, promising emerging solutions such as smart prosthetics, 3D printing, regenerative medicine, and artificial intelligence have been developed. These innovative technologies hold the potential to significantly enhance the functionality of orthopedic prostheses. Realizing the full potential of these next-generation orthopedic prostheses requires addressing several critical factors. These include interdisciplinary collaboration between experts in orthopedics, materials science, biology, and engineering, increased investment in research and development, standardization of components to ensure quality and reliability, and improved access to prosthetics. A comprehensive review of these challenges and considerations for future orthopedic prosthesis design is s provided in this paper addressing the further advances to the field. By addressing these issues, we can continue to improve the lives of individuals with orthopedic disabilities and further enhance the field of orthopedic prosthetics.

Associations between dairy farm performance indicators and culling rates under policy-driven herd size constraints.

Introduction: This article aimed to study cross-sectional associations between the performance of dairy farms and their corresponding culling proportions under the herd size constraint as imposed in 2018 by the new phosphate regulation in the Netherlands. Methods: To this end, production data from 10,540 Dutch dairy farms were analyzed to capture the inflow and outflow of both primiparous and multiparous cows. Farm performance was measured by 10 indicators structured in four areas of longevity, production, reproduction, and udder health. Farm culling proportions were represented by the overall culling (OC) and the number of culled primiparous cows in relation to (i) the total number of producing cows (PC), (ii) the number of producing primiparous cows (PPC), and (iii) the number of culled producing cows (POC). Spearman's rank correlation and weighted logistic regression were adopted to study associations. Results: In 2018, on average, 28% of producing cows were culled (OC). The number of primiparous cows culled represented 4.5% of the total number of producing cows (PC) and the mean proportion of culled primiparous cows was 18.8% of the total number of producing primiparous cows (PPC), and, of the total number of producing culled cows, 15% were primiparous cows (POC). However, the variance around the mean, and among individual farms, was high (SD 4-15% for all four culling proportions). Results from rank correlation showed very low-rank conformity (<12%) between the areas of production, reproduction, and udder health to the culling proportions. Results from logistic regression showed that higher farm levels of production and higher percentages of cows with poor udder health were associated with more overall culling but with less primiparous culling. For reproduction indicators, the associations were similar for overall and primiparous culling. However, except for the average age of culled animals, the odds ratios for indicators were close to 1 (range: 0.92-1.07 and 0.68-1.07 for OC and PPC, respectively), indicating only weak associations to culling proportions. Discussion: In conclusion, although the introduction of phosphate regulation resulted in an increased outflow of cattle, corresponding culling proportions were not associated with the level of farm performance measured in terms of production, reproduction, or udder health.

Artificial intelligence and machine learning disciplines with the potential to improve the nanotoxicology and nanomedicine fields: a comprehensive review.

The use of nanomaterials in medicine depends largely on nanotoxicological evaluation in order to ensure safe application on living organisms. Artificial intelligence (AI) and machine learning (MI) can be used to analyze and interpret large amounts of data in the field of toxicology, such as data from toxicological databases and high-content image-based screening data. Physiologically based pharmacokinetic (PBPK) models and nano-quantitative structure-activity relationship (QSAR) models can be used to predict the behavior and toxic effects of nanomaterials, respectively. PBPK and Nano-QSAR are prominent ML tool for harmful event analysis that is used to understand the mechanisms by which chemical compounds can cause toxic effects, while toxicogenomics is the study of the genetic basis of toxic responses in living organisms. Despite the potential of these methods, there are still many challenges and uncertainties that need to be addressed in the field. In this review, we provide an overview of artificial intelligence (AI) and machine learning (ML) techniques in nanomedicine and nanotoxicology to better understand the potential toxic effects of these materials at the nanoscale.

Opportunities for Smartphone Sensing in E-Health Research: A Narrative Review.

Recent years have seen significant advances in the sensing capabilities of smartphones, enabling them to collect rich contextual information such as location, device usage, and human activity at a given point in time. Combined with widespread user adoption and the ability to gather user data remotely, smartphone-based sensing has become an appealing choice for health research. Numerous studies over the years have demonstrated the promise of using smartphone-based sensing to monitor a range of health conditions, particularly mental health conditions. However, as research is progressing to develop the predictive capabilities of smartphones, it becomes even more crucial to fully understand the capabilities and limitations of using this technology, given its potential impact on human health. To this end, this paper presents a narrative review of smartphone-sensing literature from the past 5 years, to highlight the opportunities and challenges of this approach in healthcare. It provides an overview of the type of health conditions studied, the types of data collected, tools used, and the challenges encountered in using smartphones for healthcare studies, which aims to serve as a guide for researchers wishing to embark on similar research in the future. Our findings highlight the predominance of mental health studies, discuss the opportunities of using standardized sensing approaches and machine-learning advancements, and present the trends of smartphone sensing in healthcare over the years.

Annular illumination in 2D quantitative phase imaging: a systematic evaluation.

Quantitative phase imaging (QPI) is an invaluable microscopic technology for definitively imaging phase objects such as biological cells and optical fibers. Traditionally, the condenser lens in QPI produces disk illumination of the object. However, it has been realized by numerous investigators that annular illumination can produce higher-resolution images. Although this performance improvement is impressive and well documented, the evidence presented has invariably been qualitative in nature. Recently, a theoretical basis for annular illumination was presented by Bao et al. [Appl. Opt.58, 137 (2019)APOPAI0003-693510.1364/AO.58.000137]. In our current work, systematic experimental QPI measurements are made with a reference phase mask to rigorously document the performance of annular illumination. In both theory and experiment, three spatial-frequency regions are identified: low, mid, and high. The low spatial-frequency region response is very similar for disk and annular illumination, both theoretically and experimentally. Theoretically, the high spatial-frequency region response is predicted to be much better for the annular illumination compared to the disk illumination--and is experimentally confirmed. In addition, the mid-spatial-frequency region response is theoretically predicted to be less for annular illumination than for disk illumination. This theoretical degradation of the mid-spatial-frequency region is only slightly experimentally observed. This bonus, although not well understood, further elevates the performance of annular illumination over disk illumination.

Speech and Language Practitioners' Experiences of Commercially Available Voice-Assisted Technology: Web-Based Survey Study.

BACKGROUND: Speech and language therapy involves the identification, assessment, and treatment of children and adults who have difficulties with communication, eating, drinking, and swallowing. Globally, pressing needs outstrip the availability of qualified practitioners who, of necessity, focus on individuals with advanced needs. The potential of voice-assisted technology (VAT) to assist people with speech impairments is an emerging area of research but empirical work exploring its professional adoption is limited. OBJECTIVE: This study aims to explore the professional experiences of speech and language therapists (SaLTs) using VAT with their clients to identify the potential applications and barriers to VAT adoption and thereby inform future directions of research. METHODS: A 23-question survey was distributed to the SaLTs from the United Kingdom using a web-based platform, eliciting both checkbox and free-text responses, to questions on perceptions and any use experiences of VAT. Data were analyzed descriptively with content analysis of free text, providing context to their specific experiences of using VAT in practice, including barriers and opportunities for future use. RESULTS: A total of 230 UK-based professionals fully completed the survey; most were technologically competent and were aware of commercial VATs (such as Alexa and Google Assistant). However, only 49 (21.3%) SaLTs had used VAT with their clients and described 57 use cases. They reported using VAT with 10 different client groups, such as people with dysarthria and users of augmentative and alternative communication technologies. Of these, almost half (28/57, 49%) used the technology to assist their clients with day-to-day tasks, such as web browsing, setting up reminders, sending messages, and playing music. Many respondents (21/57, 37%) also reported using the technology to improve client speech, to facilitate speech practice at home, and to enhance articulation and volume. Most reported a positive impact of VAT use, stating improved independence (22/57, 39%), accessibility (6/57, 10%), and confidence (5/57, 8%). Some respondents reported increased client communication (5/57, 9%) and sociability (3/57, 5%). Reasons given for not using VAT in practice included lack of opportunity (131/181, 72.4%) and training (63/181, 34.8%). Most respondents (154/181, 85.1%) indicated that they would like to try VAT in the future, stating that it could have a positive impact on their clients' speech, independence, and confidence. CONCLUSIONS: VAT is used by some UK-based SaLTs to enable communication tasks at home with their clients. However, its wider adoption may be limited by a lack of professional opportunity. Looking forward, additional benefits are promised, as the data show a level of engagement, empowerment, and the possibility of achieving therapeutic outcomes in communication impairment. The disparate responses suggest that this area is ripe for the development of evidence-based clinical practice, starting with a clear definition, outcome measurement, and professional standardization.

Enrichment of Methylocystis dominant mixed culture from rice field for PHB production.

Presence of methanotrophs in diverse environmental habitats helps to reduce emissions of greenhouse gas like methane. Isolation and culture of undiscovered wealth of methanotrophic organisms can help in exploitation of these organisms in value added products. The present study focuses on the enrichment of methanotroph dominated mixed microbial community by use of three stage strategy of revival, proliferation, and segregation. During the enrichment process amplicon sequencing of 16 s rRNA V3-V4 region showed relative abundance of mixed culture comprising single methanotrophic species of Methylocystis genus (88.92%) along with only three other species. Methylocystis dominant mixed culture (MMI-11) was observed to produce polyhydroxyalkanoates (PHA). During studies to identify favourable culture conditions, nitrate was found to be preferred nitrogen source for growth and PHA production. Cell growth ability to produce PHA was also evaluated at 14 L fermentor by supplying gas using continuous bubbling and through pressurization in the headspace. The mixed methanotrophic culture was found to accumulate maximum of 22.20% polyhydroxybutyrate (PHB) under nitrate limited condition. The molecular weight of PHB was found to be 2.221 × 105 g mol-1 with polydispersity of 1.82.

Selection of methanotrophic platform for methanol production using methane and biogas.

To develop biotechnological process for methane to methanol conversion, selection of a suitable methanotrophic platform is an important aspect. Systematic approach based on literature and public databases was developed to select representative methanotrophs Methylotuvimicrobium alcaliphilum, Methylomonas methanica, Methylosinus trichosporium and Methylocella silvestris. Selected methanotrophs were further investigated for methanol tolerance and methanol production on pure methane as well as biogas along with key enzyme activities involved in methane utilization. Among selected methanotrophs M. alcaliphilum showed maximum methanol tolerance of 6% v/v along with maximum methanol production of 307.90 mg/L and 247.37 mg/L on pure methane and biogas respectively. Activity of methane monooxygenase and formate dehydrogenase enzymes in M.alcaliphilum was significantly higher up to 98.40 nmol/min/mg cells and 0.87 U/mg protein, respectively. Biotransformation trials in 14 L fermentor resulted in increased methanol production up to 418 and 331.20 mg/L, with yield coefficient 0.83 and 0.71 mg methanol/mg of pure methane and biogas respectively. The systematic selection resulted in haloalkaliphilic strain M. alcaliphilum as one of the potential methanotroph for bio-methanol production.

Survival analysis of dairy cows in the Netherlands under altering agricultural policy.

Culling of underperforming dairy cows by replacement heifers is a fundamental part of Dutch dairy farm management. Changes in national agricultural policies can influence farmers' culling decisions. The objective of this study was to analyse the relevancy of cow-level risk factors for survival of Dutch dairy cows under perturbations due to national policy changes related to the -milk quota abolishment of 2015 and the phosphate regulations since 2017. For this purpose, an accelerated failure time model was fitted on-longitudinal dairy cows' data at national level covering the period 2009-2019. The associated cow-level risk factors for culling such as lactation value (relative production level), parity number, rolling average of inseminations over all parities, very high fat-protein ratio (highFPR) and very low fat-protein ratio (lowFPR) in early lactation, test-day somatic cell count, were fitted in the model. Along with these, a factor representing three target policy periods, namely Milk Quota period (MQ), Post-Milk Quota period (PMQ) and Phosphate regulation period (PH) were fitted. The mean survival age for all producing cows was 441 weeks overall. The predicted median survival time for the policy periods MQ, PMQ and PH were 273 weeks, 271 weeks and 256 weeks, respectively. Risk factors such as lactation value, parity and highFPR, rolling average of inseminations over all parities were positively associated with survival time in all three policy periods. Risk factors such as test-day somatic cell count and lowFPR were negatively associated with survival time in all three policy periods. In conclusion, this study demonstrated the differences in survival of Dutch dairy cows in response to changing agricultural policy. The association of cow-level risk factors for culling was consistent across the three evaluated policy periods.

On the origin of the genetic variation in infectious disease prevalence: Genetic analysis of disease status versus infections for Digital Dermatitis in Dutch dairy cattle.

The purpose of this study was to investigate the origin of the genetic variation in the prevalence of bovine digital dermatitis (DD) by comparing a genetic analysis of infection events to a genetic analysis of disease status. DD is an important endemic infectious disease affecting the claws of cattle. For disease status, we analysed binary data on individual disease status (0,1; indicating being free versus infected), whereas for infections, we analysed binary data on disease transmission events (1,0; indicating becoming infected or not). The analyses of the two traits were compared using cross-validation. The analysis of disease status captures a combination of genetic variation in disease susceptibility and the ability of individuals to recover, whereas the analysis of infections captures genetic variation in susceptibility only. Estimated genetic variances for both traits indicated substantial genetic variation. The GEBV for disease status and infections correlated with only 0.60, indicating that both models indeed capture distinct information. Together, these results suggest the presence of genetic variation not only in disease susceptibility, but also in the ability of individuals to recover from DD. We argue that the presence of genetic variation in recovery implies that breeders should distinguish between infected individuals versus infectious individuals. This is because epidemiological theory shows that selection for recovery is effective only when it targets recovery from being infectious.

Integrative analysis and machine learning on cancer genomics data using the Cancer Systems Biology Database (CancerSysDB).

BACKGROUND: Recent cancer genome studies on many human cancer types have relied on multiple molecular high-throughput technologies. Given the vast amount of data that has been generated, there are surprisingly few databases which facilitate access to these data and make them available for flexible analysis queries in the broad research community. If used in their entirety and provided at a high structural level, these data can be directed into constantly increasing databases which bear an enormous potential to serve as a basis for machine learning technologies with the goal to support research and healthcare with predictions of clinically relevant traits. RESULTS: We have developed the Cancer Systems Biology Database (CancerSysDB), a resource for highly flexible queries and analysis of cancer-related data across multiple data types and multiple studies. The CancerSysDB can be adopted by any center for the organization of their locally acquired data and its integration with publicly available data from multiple studies. A publicly available main instance of the CancerSysDB can be used to obtain highly flexible queries across multiple data types as shown by highly relevant use cases. In addition, we demonstrate how the CancerSysDB can be used for predictive cancer classification based on whole-exome data from 9091 patients in The Cancer Genome Atlas (TCGA) research network. CONCLUSIONS: Our database bears the potential to be used for large-scale integrative queries and predictive analytics of clinically relevant traits.

Challenges in the Setup of Large-scale Next-Generation Sequencing Analysis Workflows.

While Next-Generation Sequencing (NGS) can now be considered an established analysis technology for research applications across the life sciences, the analysis workflows still require substantial bioinformatics expertise. Typical challenges include the appropriate selection of analytical software tools, the speedup of the overall procedure using HPC parallelization and acceleration technology, the development of automation strategies, data storage solutions and finally the development of methods for full exploitation of the analysis results across multiple experimental conditions. Recently, NGS has begun to expand into clinical environments, where it facilitates diagnostics enabling personalized therapeutic approaches, but is also accompanied by new technological, legal and ethical challenges. There are probably as many overall concepts for the analysis of the data as there are academic research institutions. Among these concepts are, for instance, complex IT architectures developed in-house, ready-to-use technologies installed on-site as well as comprehensive Everything as a Service (XaaS) solutions. In this mini-review, we summarize the key points to consider in the setup of the analysis architectures, mostly for scientific rather than diagnostic purposes, and provide an overview of the current state of the art and challenges of the field.

Semi-automated cancer genome analysis using high-performance computing.

Next-generation sequencing (NGS) has turned from a new and experimental technology into a standard procedure for cancer genome studies and clinical investigation. While a multitude of software packages for cancer genome data analysis have been made available, these need to be combined into efficient analytical workflows that cover multiple aspects relevant to a clinical environment and that deliver handy results within a reasonable time frame. Here, we introduce QuickNGS Cancer as a new suite of bioinformatics pipelines that is focused on cancer genomics and significantly reduces the analytical hurdles that still limit a broader applicability of NGS technology, particularly to clinically driven research. QuickNGS Cancer allows a highly efficient analysis of a broad variety of NGS data types, specifically considering cancer-specific issues, such as biases introduced by tumor impurity and aneuploidy or the assessment of genomic variations regarding their biomedical relevance. It delivers highly reproducible analysis results ready for interpretation within only a few days after sequencing, as shown by a reanalysis of 140 tumor/normal pairs from The Cancer Genome Atlas (TCGA) in which QuickNGS Cancer detected a significant number of mutations in key cancer genes missed by a well-established mutation calling pipeline. Finally, QuickNGS Cancer obtained several unexpected mutations in leukemias that could be confirmed by Sanger sequencing.

Aging is associated with highly defined epigenetic changes in the human epidermis.

BACKGROUND: Altered DNA methylation patterns represent an attractive mechanism for understanding the phenotypic changes associated with human aging. Several studies have described global and complex age-related methylation changes, but their structural and functional significance has remained largely unclear. RESULTS: We have used transcriptome sequencing to characterize age-related gene expression changes in the human epidermis. The results revealed a significant set of 75 differentially expressed genes with a strong functional relationship to skin homeostasis. We then used whole-genome bisulfite sequencing to identify age-related methylation changes at single-base resolution. Data analysis revealed no global aberrations, but rather highly localized methylation changes, particularly in promoter and enhancer regions that were associated with altered transcriptional activity. CONCLUSIONS: Our results suggest that the core developmental program of human skin is stably maintained through the aging process and that aging is associated with a limited destabilization of the epigenome at gene regulatory elements.