Studies in health technology and informatics最新文献

The paper describes a cohort of patients with post-acute COVID-19 syndrome, evaluated for the first time between week 3 and week 12 from the onset of symptoms following the acute COVID-19 infection. The patient's baseline clinical features were used as predictors. The analysis showed that older patients with comorbidities are at higher risk of developing more long-lasting post COVID-19 symptoms. Further integration with a personal monitoring device and combination with the Fast Healthcare Interoperability Resources extends the standardization, interoperability and possibility of integration and harmonization with other hospital systems. By employing advanced machine learning techniques, insights can be derived and further examined to improve the outcome and early treatment options for patients.

The PROSurvival project aims to improve the prediction of recurrence-free survival in prostate cancer by applying federated machine learning to whole slide images combined with selected clinical data. Both the image and clinical data will be aggregated into an anonymized dataset compliant with the General Data Protection Regulation and published under the principles of findable, accessible, interoperable, and reusable data. The DICOM standard will be used for the image data. For the accompanying clinical data, a human-readable, compact and flexible standard is yet to be defined. From the set of existing standards, mostly extendable with varying degrees of modifications, we chose oBDS as a starting point and modified it to include missing data points and to remove mandatory items not applicable to our dataset. Clinical and survival data from clinic-specific spreadsheets were converted into this modified standard, ensuring on-site data privacy during processing. For publication of the dataset, both image and clinical data are anonymized using established methods. The key challenges arose during the clinical data anonymization and in identifying research repositories meeting all of our requirements. Each clinic had to coordinate the publication with their responsible data protection officers, requiring different approval processes due to the individual states' differing interpretations of the legal regulations. The newly established German Health Data Utilization Act is expected to simplify future data sharing in a responsible and powerful way.

The present study explored alternative methods for photographing skin lesions in the absence of specialized instruments like dermatoscopes, aiming to enhance remote diagnostic capabilities, particularly in light of the increasing incidence of melanoma cases annually. Using two lenses attached to a smartphone camera, one macroscopic and the other microscopic, study images of nevus formations from one individual were captured, and, in the absence of a collaboration with a dermatologist, subsequently labeled as melanoma or non-melanoma using a Convolutional Neural Network (CNN) which was trained, with dermoscopic images of melanoma and non-melanoma formations, to see on which image set better performances would be attained. The CNN demonstrated better performance on microscopic images, with 75% of the dataset being labeled correctly, compared to the macroscopic one, with 63% of the dataset being labeled correctly. These findings highlight the potential of smartphone-based imaging with specialized micro lenses to improve diagnostic accuracy for melanoma and other dermatological conditions in remote healthcare settings.

Pediatric growth hormone deficiency (PGHD) is a chronic condition where the pituitary gland fails to produce sufficient growth hormone, leading to delayed growth and developmental challenges. Patient journey maps can provide insight into pain points and potential opportunities for new or improved interventions to enhance care. However, a patient journey map does not yet exist for PGHD. Secondary data analysis was performed on interviews and focus groups from five cohorts in Sweden, the United Kingdom, Luxembourg, France, and The Netherlands. Participants included 62 patients and caregivers who used a prototype digital health solution, which was used to guide discussions. Grounded theory was used to analyze the data, resulting in a patient journey map comprising six stages: awareness, diagnosis, treatment planning, treatment initiation, treatment maintenance and transition. This provides the first detailed PGHD patient journey map, revealing emotional sensitivities and challenges at each stage, and suggesting areas for targeted interventions to improve adherence and long-term outcomes.

Structured data are the capital of empirical health research. The value of these data relates to their quality and to their fit for use. A German guideline for the management of data quality in registries and cohort studies lists 51 quality indicators organized into the categories organization, integrity, and trueness. An update of the guideline will take into account the current view on dimensions of data, the appropriate structure for the definition of an indicator, and the collection of quality indicators itself. In the next version, the collection will explicitly address measures of metadata quality. The first step of a literature review revealed a high number of potential sources of evidence. These will be categorized into the topics dimensions, structure, and indicators respectively. Special attention will be paid to new challenges of data quality control arising from big data and artificial intelligence.

Multiple sclerosis (MS) is a complex neurodegenerative disease with a variable prognosis that complicates effective management and treatment. This study leverages machine learning (ML) to enhance the understanding of disease progression and uncover gender-based differences in MS by analyzing clinical data integrated with patient-reported outcomes (PROMs). We conducted a prospective cohort study involving 250 MS patients at a secondary care hospital in Spain over an 18-month period. Using REDCap for data management, we collected comprehensive demographic, clinical, and PROMs data. Our analysis utilized Decision Trees, Random Forest, and Support Vector Machine algorithms to classify patients based on disease evolution and infer Expanded Disability Status Scale (EDSS) levels. Additionally, we employed propensity score matching to analyze gender differences, focusing on clinical outcomes and quality of life measures. The results could indicate that integrating diverse data sets through ML would significantly improve the diagnostic accuracy and serve as a support for clinician's decision making. Our models achieved high accuracy in classifying MS types and predicting disability levels, demonstrating the potential of ML in personalized treatment planning. Furthermore, our findings suggest notable gender differences in disease progression and response to treatment. These insights advocate for a gender-specific approach in MS management and highlight the importance of personalized medicine. This study underscores the transformative potential of ML in enhancing the understanding and management of MS through integrated data analysis.

Sesterterpenoids, a subset of the terpene family, exhibit notable biological activities. These natural compounds are present in a variety of organisms such as plants, fungi, bacteria, insects and marine life. The therapeutic potential and structural diversity of sesterterpenoids have attracted considerable interest in pharmacological and chemical research. This study illustrates the development of a database to structure and manage data on these compounds. The design process involves the collection of user requirements, creation of a conceptual model with and Entity-Relationship Diagram (ERD), development of a logical model, and implementation in Microsoft SQL Server 2022. Data collection began with an extensive literature review and organization in an Excel spreadsheet. The resulting database improves data acquisition, organization, and accessibility. Future work will include building a website to facilitate data entry, editing, reading and extraction, and automation of data updates via external web services.

This paper proposes to create an Robotic Process Automation style application that can digitalize and extract data from handwritten medical forms. The RPA robot uses OpenAI ChatGPT4o model to extract handwritten medical data and transform it into typed data. The handwritten data is transcribed correctly at a rate of 100%. The data interpretation is accomplished by the UiPath machine learning API. By creating new nonstandard form templates and associated taxonomies the system can be scaled as desired. After the data extraction process the saved data can be sent to a database, spreadsheet. The access to this medical data is restricted to the physicians and medical nurses employed at the medical facility.

The development of medical science allows the treatment of more and more health problems that in the past were not a factor of consumption of health resources, because at that time medical science did not have protocols for their treatment. Health problems that are now treatable, hereafter referred as "new health problems", often affect large population groups and require increased consumption of health resources. It therefore becomes necessary to increase the number of staff providing health services (doctors, nurses, etc.) and other resources. This raises the question: is it feasible to manage the "new health problems" by the existing medical staff? If not, are there other solutions? Could technology help the existing Medical Staff to sufficiently manage the "new health problems"? We will examine a pilot system "Recording and visualizing of outpatient monitoring data with smart mobile phones", which seeks to ensure the competence of existing medical staff in the effective treatment of the ever-increasing volume of transplant patients.

Healthcare systems worldwide face escalating costs and demographic changes, necessitating effective evaluation tools to understand their underlying challenges. Switzerland's high-quality yet costly healthcare system underscores the need for robust assessment methods. Existing international rankings often lack transparency and comparability, highlighting the value of structured frameworks like the Health System Performance Assessment (HSPA) by the World Health Organization (WHO). This framework evaluates healthcare systems across multiple dimensions including governance, resource generation, financing, and service delivery. This paper aims to integrate Swiss healthcare indicators from the Swiss Health Observatory (Obsan) into the HSPA framework, addressing the central research question: How can these indicators be mapped to the HSPA framework, and what insights does this integration provide? Our methodology includes selecting and categorizing Obsan indicators, manually mapping them to HSPA sub-functions, and validating these mappings using word embeddings and cosine similarity. An R Shiny application was developed for interactive visualization. Results demonstrate accurate indicator assignment, enabling intuitive visualization and enhancing data structuring.