Hasil untuk "Computer applications to medicine. Medical informatics"

Fei Wang, A. Preininger

Summary Introduction: Artificial intelligence (AI) technologies continue to attract interest from a broad range of disciplines in recent years, including health. The increase in computer hardware and software applications in medicine, as well as digitization of health-related data together fuel progress in the development and use of AI in medicine. This progress provides new opportunities and challenges, as well as directions for the future of AI in health. Objective: The goals of this survey are to review the current state of AI in health, along with opportunities, challenges, and practical implications. This review highlights recent developments over the past five years and directions for the future. Methods: Publications over the past five years reporting the use of AI in health in clinical and biomedical informatics journals, as well as computer science conferences, were selected according to Google Scholar citations. Publications were then categorized into five different classes, according to the type of data analyzed. Results: The major data types identified were multi-omics, clinical, behavioral, environmental and pharmaceutical research and development (R&D) data. The current state of AI related to each data type is described, followed by associated challenges and practical implications that have emerged over the last several years. Opportunities and future directions based on these advances are discussed. Conclusion: Technologies have enabled the development of AI-assisted approaches to healthcare. However, there remain challenges. Work is currently underway to address multi-modal data integration, balancing quantitative algorithm performance and qualitative model interpretability, protection of model security, federated learning, and model bias.

Iulian Întorsureanu, S. Oprea, A. Bâra et al.

In this paper, we investigated the role of generative AI in education in academic publications extracted from Web of Science (3506 records; 2019–2024). The proposed methodology included three main streams: (1) Monthly analysis trends; top-ranking research areas, keywords and universities; frequency of keywords over time; a keyword co-occurrence map; collaboration networks; and a Sankey diagram illustrating the relationship between AI-related terms, publication years and research areas; (2) Sentiment analysis using a custom list of words, VADER and TextBlob; (3) Topic modeling using Latent Dirichlet Allocation (LDA). Terms such as “artificial intelligence” and “generative artificial intelligence” were predominant, but they diverged and evolved over time. By 2024, AI applications had branched into specialized fields, including education and educational research, computer science, engineering, psychology, medical informatics, healthcare sciences, general medicine and surgery. The sentiment analysis reveals a growing optimism in academic publications regarding generative AI in education, with a steady increase in positive sentiment from 2023 to 2024, while maintaining a predominantly neutral tone. Five main topics were derived from AI applications in education, based on an analysis of the most relevant terms extracted by LDA: (1) Gen-AI’s impact in education and research; (2) ChatGPT as a tool for university students and teachers; (3) Large language models (LLMs) and prompting in computing education; (4) Applications of ChatGPT in patient education; (5) ChatGPT’s performance in medical examinations. The research identified several emerging topics: discipline-specific application of LLMs, multimodal gen-AI, personalized learning, AI as a peer or tutor and cross-cultural and multilingual tools aimed at developing culturally relevant educational content and supporting the teaching of lesser-known languages. Further, gamification with generative AI involves designing interactive storytelling and adaptive educational games to enhance engagement and hybrid human–AI classrooms explore co-teaching dynamics, teacher–student relationships and the impact on classroom authority.

Kecheng Huang, Chujun Wu, Rongpeng Pi et al.

This viewpoint article explores the transformative role of artificial intelligence (AI) in predicting perioperative hypoxemia through the integration of deep learning with multimodal clinical data, including lung imaging, pulmonary function tests, and arterial blood gas (ABG) analysis. Perioperative hypoxemia, defined as arterial oxygen partial pressure <60 mmHg or oxygen saturation <90%, poses significant risks of delayed recovery and organ dysfunction. Traditional diagnostic methods such as radiological imaging and ABG analysis often lack integrated predictive accuracy. AI frameworks, particularly convolutional neural networks and hybrid models like TD-CNNLSTM-LungNet, demonstrate exceptional performance in detecting pulmonary inflammation and stratifying hypoxemia risk, achieving up to 96.57% accuracy in pneumonia subtype differentiation and an area under the curve of 0.96 for postoperative hypoxemia prediction. Multimodal AI systems, such as DeepLung-Predict, unify computed tomography scans, pulmonary function tests, and ABG parameters to enhance predictive precision, surpassing conventional methods by 22%. However, challenges persist, including dataset heterogeneity, model interpretability, and clinical workflow integration. Future directions emphasize multicenter validation, explainable AI frameworks, and pragmatic trials to ensure equitable and reliable deployment. This AI-driven approach not only optimizes resource allocation but also mitigates financial burdens on health care systems by enabling early interventions and reducing intensive care unit admission risks.

Rui-Fang Lu, Chao-Yin She, Dan-Ni He et al.

Abstract Hepatocellular carcinoma (HCC) ultrasound screening encounters challenges related to accuracy and the workload of radiologists. This retrospective, multicenter study assessed four artificial intelligence (AI) enhanced strategies using 21,934 liver ultrasound images from 11,960 patients to improve HCC ultrasound screening accuracy and reduce radiologist workload. UniMatch was used for lesion detection and LivNet for classification, trained on 17,913 images. Among the strategies tested, Strategy 4, which combined AI for initial detection and radiologist evaluation of negative cases in both detection and classification phases, outperformed others. It not only matched the high sensitivity of original algorithm (0.956 vs. 0.991) but also improved specificity (0.787 vs. 0.698), reduced radiologist workload by 54.5%, and decreased both recall and false positive rates. This approach demonstrates a successful model of human-AI collaboration, not only enhancing clinical outcomes but also mitigating unnecessary patient anxiety and system burden by minimizing recalls and false positives.

R. A. Ghayda, R. Cannarella, A. Calogero et al.

Artificial intelligence (AI) in medicine has gained a lot of momentum in the last decades and has been applied to various fields of medicine. Advances in computer science, medical informatics, robotics, and the need for personalized medicine have facilitated the role of AI in modern healthcare. Similarly, as in other fields, AI applications, such as machine learning, artificial neural networks, and deep learning, have shown great potential in andrology and reproductive medicine. AI-based tools are poised to become valuable assets with abilities to support and aid in diagnosing and treating male infertility, and in improving the accuracy of patient care. These automated, AI-based predictions may offer consistency and efficiency in terms of time and cost in infertility research and clinical management. In andrology and reproductive medicine, AI has been used for objective sperm, oocyte, and embryo selection, prediction of surgical outcomes, cost-effective assessment, development of robotic surgery, and clinical decision-making systems. In the future, better integration and implementation of AI into medicine will undoubtedly lead to pioneering evidence-based breakthroughs and the reshaping of andrology and reproductive medicine.

Jose Manuel PINILLOS RUBIO, Minerva VIGUERA MORENO

This project develops a cloud-based solution for securely managing clinical data and patient-reported outcomes (PROMs) for multiple sclerosis (MS) patients. Utilizing REDCap for data collection, we incorporated clinical outcomes and PROMs from 300 MS patients over 18 months, supporting a machine learning (ML) based clinical decision support system. Our cloud architecture, featuring segregated data handling and enhanced security protocols using AWS, ensures robust data integrity and confidentiality. Key improvements include streamlined data ETL processes and an interactive online-based dashboard that facilitates the visualization of clinical data and PROMs, crucial for effective clinical decision-making. Initial results indicate a successful implementation in enhancing data management, with implications for personalized and predictive medicine. This framework not only elevates clinical data handling efficiency but also integrates PROMs into clinical practice effectively.

Hairong ZOU

In the design of DR systems, high-voltage generators and X-ray tubes are two key components whose performance directly affects the system's exposure capabilities. This article first explains why it is necessary to conduct an in-depth evaluation of the performance of these two parts during the design of DR products. Then, it provides a detailed description of the evaluation methods, including the assessment of single exposure capability and continuous exposure capability. Finally, the article offers some suggestions and measures to enhance continuous exposure capability. These methods and suggestions not only help in selecting appropriate key components but also provide an important reference for the design, integration, and testing of DR systems.

Sainan Duan, Dongxu Chen, Jinping Wang et al.

BackgroundCognition disorders not only lead to adverse health consequences but also contribute to a range of socioeconomic challenges and diminished capacity for performing routine daily activities. In the digital era, understanding the impact of digital exclusion on cognitive function is crucial, especially in developing countries. ObjectiveThis study aimed to evaluate the association between digital exclusion and cognitive function among elderly populations in developing countries. MethodsUsing data from CHARLS (China Health and Retirement Longitudinal Study) from 2011 to 2020 and MHAS (Mexican Health & Aging Study) from 2012 to 2021, we defined digital exclusion as self-reported absence from the internet. Cognitive function was assessed through 5 tests: orientation, immediate verbal recall, delayed verbal recall, serial 7s, and figure recall. Cognitive function was assessed in 2 categories: worse cognition (a categorical variable that classifies cognition as either better or worse compared to the entire cohort population) and cognitive scores (a continuous variable representing raw cognitive scores across multiple follow-up waves). Logistic regression analyses and generalized estimating equation (GEE) analyses were used to examine the relationship between cognitive function and digital exclusion, adjusting for potential confounders, including demographics, lifestyle factors, history of chronic diseases, basic activities of daily living (BADL) disability, instrumental activities of daily living (IADL) disability, and basic cognitive abilities. ResultsAfter excluding participants with probable cognitive impairment at baseline and those who did not have a complete cognitive assessment in any given year (ie, all tests in the cognitive assessment must be completed in any follow-up wave), a total of 24,065 participants in CHARLS (n=11,505, 47.81%) and MHAS (n=12,560, 52.19%) were included. Of these, 96.78% (n=11,135) participants in CHARLS and 70.02% (n=8795) in MHAS experienced digital exclusion. Adjusted logistic regression analyses revealed that individuals with digital exclusion were more likely to exhibit worse cognitive performance in both CHARLS (odds ratio [OR] 2.04, 95% CI 1.42-2.99; P<.001) and MHAS (OR 1.40, 95% CI 1.26-1.55; P<.001). Gender and age did not significantly modify the relationship between digital exclusion and worse cognition (intervention P>.05). The fully adjusted mean differences in global cognitive scores between the 2 groups were 0.98 (95% CI 0.70-1.28; P<.001) in CHARLS and 0.50 (95% CI 0.40-0.59; P<.001) in MHAS. ConclusionsA substantial proportion of older adults, particularly in China, remain excluded from internet access. Our study examined longitudinal changes in cognitive scores and performed cross-sectional comparisons using Z-score standardization. The findings suggest that digital exclusion is linked to an increased risk of cognitive decline among older adults in developing countries. Promoting internet access may help mitigate this risk and support better cognitive health in these populations.

Johannes von Büren, Inga Hansen, Julian Kött et al.

Objective The use of direct-to-consumer (DTC) teledermatology platforms has increased, particularly for androgenetic alopecia (AGA). However, little is known about the efficacy and safety of these platforms. This study aimed to investigate the patient-reported treatment outcomes and safety of DTC teledermatology for the finasteride treatment of male AGA. Methods This retrospective, cross-sectional study used data from a German DTC platform for finasteride treatment between December 2021 and January 2023. Patient-reported outcomes were collected through voluntary follow-up questionnaires provided to the patients six weeks after the first prescription to assess treatment outcomes and safety. Results Data collection included 2269 patients. Of all patients who answered the follow-up questionnaire ( n  = 191), 79% (150 out of 191) self-reported positive changes in hair appearance, and 59% (113 out of 191) reported an improvement in self-esteem under treatment. Patients with self-reported positive changes in hair appearance were more likely to report improved self-esteem ( P  < 0.0001). Treatment-related adverse events occurred in 12% (22 out of 191) of the patients. Full treatment adherence was reported in 87% (167 out of 191) of patients. Conclusion From the patient's perspective, DTC teledermatology has the potential to improve hair appearance and self-esteem. Our results suggest that it may be an effective and safe treatment option for men with AGA, justifying low-threshold access. However, treatment-related adverse events should be closely monitored during follow-up. Further studies are required to evaluate the long-term effects of the DTC teledermatology treatment. By collecting real-world data, teledermatology platforms could be useful beyond their primary focus and could play an important role in the context of future research.

Maram A. Aljabry, Khulud Alkadi, Mohammed Awawdeh

Objective: The aim of the study was to assess dental patients’ knowledge and perceptions of the risks associated with orthodontic treatment and oral hygiene instructions given at the start of their treatment regimen. Methods: The sample group included dental patients aged 12–16 years. Half of the participants were taught using Augmented Reality (AR) technology, while the other half were taught using traditional methods, such as paper-based leaflets. We then assessed retention of the Oral Hygiene Instructions (OHI) and their awareness of the risks associated with orthodontic treatment. This objective was achieved through a knowledge-based survey. Results: The feedback from the participants was highly positive and favorable toward AR. Conclusion: The videos and leaflet materials were provided to the patients, these two materials effects on patients to understand the information given, but, the group watching the videos had greater scores than those participants reading the leaflet.

Annalisa Maraone, Alessandro Trebbastoni, Antonella Di Vita et al.

BackgroundObsessive-compulsive disorder (OCD) is a psychiatric syndrome characterized by unwanted and repetitive thoughts and repeated ritualistic compulsions for decreasing distress. Symptoms can cause severe distress and functional impairment. OCD affects 2% to 3% of the population and is ranked within the 10 leading neuropsychiatric causes of disability. Cortico-striatal-thalamo-cortical circuitry dysfunction has been implicated in OCD, including altered brain activation and connectivity. Complex glutamatergic signaling dysregulation within cortico-striatal circuitry has been proposed in OCD. Data obtained by several studies indicate reduced glutamatergic concentrations in the anterior cingulate cortex, combined with overactive glutamatergic signaling in the striatum and orbitofrontal cortex. A growing number of randomized controlled trials have assessed the utility of different glutamate-modulating drugs as augmentation medications or monotherapies for OCD, including refractory OCD. However, there are relevant variations among studies in terms of patients’ treatment resistance, comorbidity, age, and gender. At present, 4 randomized controlled trials are available on the efficacy of memantine as an augmentation medication for refractory OCD. ObjectiveOur study’s main purpose is to conduct a double-blind, randomized, parallel-group, placebo-controlled, monocenter trial to assess the efficacy and safety of memantine as an augmentative agent to a selective serotonin reuptake inhibitor in the treatment of moderate to severe OCD. The study’s second aim is to evaluate the effect of memantine on cognitive functions in patients with OCD. The third aim is to investigate if responses to memantine are modulated by variables such as gender, symptom subtypes, and the duration of untreated illness. MethodsInvestigators intend to conduct a double-blind, randomized, parallel-group, placebo-controlled, monocenter trial to assess the efficacy and safety of memantine as an augmentative agent to a selective serotonin reuptake inhibitor in the treatment of patients affected by severe refractory OCD. Participants will be rated via the Yale-Brown Obsessive Compulsive Scale at baseline and at 2, 4, 6, 8, 10, and 12 months. During the screening period and T4 and T6 follow-up visits, all participants will undergo an extensive neuropsychological evaluation. The 52-week study duration will consist of 4 distinct periods, including memantine titration and follow-up periods. ResultsRecruitment has not yet started. The study will be conducted from June 2023 to December 2024. Results are expected to be available in January 2025. Throughout the slow-titration period, we will observe the minimum effective dose of memantine, and the follow-up procedure will detail its residual efficacy after drug withdrawal. ConclusionsThe innovation of this research proposal is not limited to the evaluation of the efficacy and safety of memantine as an augmentation medication for OCD. We will also test if memantine acts as a pure antiobsessive medication or if memantine’s ability to improve concentration and attention mimics an antiobsessive effect. Trial RegistrationClinicalTrials.gov NCT05015595; https://clinicaltrials.gov/ct2/show/NCT05015595 International Registered Report Identifier (IRRID)PRR1-10.2196/39223

Fotios S. Konstantakopoulos, Eleni I. Georga, Dimitrios I. Fotiadis

<italic>Goal</italic>: The modern way of living has significantly influenced the daily diet. The ever-increasing number of people with obesity, diabetes and cardiovascular diseases stresses the need to find tools that could help in the daily intake of the necessary nutrients. <italic>Methods:</italic> In this paper, we present an automated image-based dietary assessment system of Mediterranean food, based on: 1) an image dataset of Mediterranean foods, 2) on a pre-trained Convolutional Neural Network (CNN) for food image classification, and 3) on stereo vision techniques for the volume and nutrition estimation of the food. We use a pre-trained CNN in the Food-101 dataset to train a deep learning classification model employing our dataset Mediterranean Greek Food (MedGRFood). Based on the EfficientNet family of CNNs, we use the EfficientNetB2 both for the pre-trained model and its weights evaluation, as well as for classifying food images in the MedGRFood dataset. Next, we estimate the volume of the food, through 3D food reconstruction of two images taken by a smartphone camera. The proposed volume estimation subsystem uses stereo vision techniques and algorithms, and needs the input of two food images to reconstruct the point cloud of the food and to compute its quantity. <italic>Results:</italic> The classification accuracy where true class matches with the most probable class predicted by the model (Top-1 accuracy) is 83.8&#x0025;, while the accuracy where true class matches with any one of the 5 most probable classes predicted by the model (Top-5 accuracy) is 97.6&#x0025;, for the food classification subsystem. The food volume estimation subsystem achieves an overall mean absolute percentage error 10.5&#x0025; for 148 different food dishes. <italic>Conclusions:</italic> The proposed automated image-based dietary assessment system provides the capability of continuous recording of health data in real time.

Yiming Bao, Yongbiao Xue

Luca Barni, María Ruiz-Muñoz, Manuel Gonzalez-Sanchez et al.

Abstract Introduction There is no systematic review that analyzes the psychometric properties of questionnaires in Italian. Previous studies have analyzed the psychometric characteristics of instruments for the measurement of pathologies of upper limbs and their joints in different languages. The aim of the present study was to analyze the psychometric properties of the questionnaires published in Italian for the evaluation of the entire upper limb or some of its specific regions and related dysfunctions. Evidence acquisition For the development of this systematic review, the following databases were used: PubMed, Scopus, Cochrane, Dialnet, Cinahl, Embase and PEDro. The selection criteria used in this study were: studies of transcultural adaptation to Italian of questionnaires oriented to the evaluation of upper limbs or any of their structures (specifically shoulder, elbow and wrist/hand), and contribution of psychometric variables of the questionnaire in its Italian version. Evidence synthesis After reading the titles and applying the inclusion and exclusion criteria to the complete documents, 16 documents were selected: 3 for the upper limb, 8 for the shoulder, 1 for the elbow and 4 for the wrist and hand. The cross-sectional psychometric variables show levels between good and excellent in all the questionnaires. Longitudinal psychometric variables had not been calculated in the vast majority of the analyzed questionnaires. Conclusions Italian versions of the questionnaires show good basic structural and psychometric characteristics for the evaluation of patients with musculoskeletal disorders of the upper limb and its joints (shoulder, elbow and wrist/hand).

Harriet Unsworth, Bernice Dillon, Lucie Collinson et al.

Objective In 2018, the UK National Institute for Health and Care Excellence (NICE), in partnership with Public Health England, NHS England, NHS Improvement and others, developed an evidence standards framework (ESF) for digital health and care technologies (DHTs). The ESF was designed to provide a standardised approach to guide developers and commissioners on the levels of evidence needed for the clinical and economic evaluation of DHTs by health and care systems. Methods The framework was developed using an agile and iterative methodology that included a literature review of existing initiatives and comparison of these against the requirements set by NHS England; iterative consultation with stakeholders through an expert working group and workshops; and questionnaire-based stakeholder input on a publicly available draft document. Results The evidence standards framework has been well-received and to date the ESF has been viewed online over 55,000 times and downloaded over 19,000 times. Conclusions In April 2021 we published an update to the ESF. Here, we summarise the process through which the ESF was developed, reflect on its global impact to date, and describe NICE’s ongoing work to maintain and improve the framework in the context for a fast moving, innovative field.

Chu, Cherry, Cram, Peter, Pang, Andrea et al.

BackgroundThe COVID-19 pandemic has led to a notable increase in telemedicine adoption. However, the impact of the pandemic on telemedicine use at a population level in rural and remote settings remains unclear. ObjectiveThis study aimed to evaluate changes in the rate of telemedicine use among rural populations and identify patient characteristics associated with telemedicine use prior to and during the pandemic. MethodsWe conducted a repeated cross-sectional study on all monthly and quarterly rural telemedicine visits from January 2012 to June 2020, using administrative data from Ontario, Canada. We compared the changes in telemedicine use among residents of rural and urban regions of Ontario prior to and during the pandemic. ResultsBefore the pandemic, telemedicine use was steadily low in 2012-2019 for both rural and urban populations but slightly higher overall for rural patients (11 visits per 1000 patients vs 7 visits per 1000 patients in December 2019, P<.001). The rate of telemedicine visits among rural patients significantly increased to 147 visits per 1000 patients in June 2020. A similar but steeper increase (P=.15) was observed among urban patients (220 visits per 1000 urban patients). Telemedicine use increased across all age groups, with the highest rates reported among older adults aged ≥65 years (77 visits per 100 patients in 2020). The proportions of patients with at least 1 telemedicine visit were similar across the adult age groups (n=82,246/290,401, 28.3% for patients aged 18-49 years, n=79,339/290,401, 27.3% for patients aged 50-64 years, and n=80,833/290,401, 27.8% for patients aged 65-79 years), but lower among younger patients <18 years (n=23,699/290,401, 8.2%) and older patients ≥80 years (n=24,284/290,401, 8.4%) in 2020 (P<.001). There were more female users than male users of telemedicine (n=158,643/290,401, 54.6% vs n=131,758/290,401, 45.4%, respectively, in 2020; P<.001). There was a significantly higher proportion of telemedicine users residing in relatively less rural than in more rural regions (n=261,814/290,401, 90.2% vs n=28,587/290,401, 9.8%, respectively, in 2020; P<.001). ConclusionsTelemedicine adoption increased in rural and remote areas during the COVID-19 pandemic, but its use increased in urban and less rural populations. Future studies should investigate the potential barriers to telemedicine use among rural patients and the impact of rural telemedicine on patient health care utilization and outcomes.

Hanne Lefrère, Giuseppe Floris, Marjanka K. Schmidt et al.

Postpartum breast cancer (PPBC) - which according to new data, can extend to 5–10 years after the birth - are estimated to represent 35–55% of all cases of breast cancer in women younger than 45 years. Increasing clinical evidence indicates that PPBC represents a high-risk form of breast cancer in young women with an approximately 2-fold increased risk for metastasis and death. Yet, the exact mechanisms that underlay this poor prognosis are incompletely understood and, hence, it is unknown why postpartum breast cancer has an enhanced risk for metastasis or how it should be effectively targeted for improved survival. This article is an accompanying resource of the original article entitled “Breast cancer diagnosed in the post-weaning period is indicative for a poor outcome” and present epidemiological data that compare standard prognostic parameters, first site of metastatic disease and survival and metastatic rates in young women with primary invasive breast cancer diagnosed within two years postpartum (PP-BC), in young women diagnosed during pregnancy (Pr-BC) and nulliparous women (NP-BC). Via an international collaboration of 13 centres participating in the International Network on Cancer, Infertility and Pregnancy (INCIP), retrospective data of 1180 patients with primary invasive breast cancer, aged 25–40 years and diagnosed between January 1995 and December 2017 were collected. In particular, tumour-, patient, and therapy-related characteristics were collected. Furthermore, patient files were reviewed thoroughly to assess, for each parity, if and for how long breastfeeding was given. For PP-BC patients, breastfeeding history was used to differentiate breast cancers identified during lactation (PP-BCDL) from those diagnosed post-weaning (PP-BCPW). Primary exposures were prior childbirth or no childbirth, time between most recent childbirth and breast cancer diagnosis, time between cessation of lactation and breast cancer diagnosis and time between breast cancer diagnosis and metastasis or death. Distribution of standard prognostic parameters and first site of distant metastasis among study groups was determined applying fisher's exact, chi-squared, One-Way ANOVA or Kruskal-Wallis tests or logistic regression models, where applicable. The risks for metastasis and death were assessed using Cox proportional hazards models. A subgroup analysis was performed in PP-BCPW patients that never lactated (PP-BCPW/NL), lactated ≤3 months (PP-BCPW/Lshort) or lactated >3 months (PP-BCPW/Llong).

Juri Yanase, Evangelos Triantaphyllou

Z. Hammal, Di Huang, Kévin Bailly et al.

The goal of Face and Gesture Analysis for Health Informatics's workshop is to share and discuss the achievements as well as the challenges in using computer vision and machine learning for automatic human behavior analysis and modeling for clinical research and healthcare applications. The workshop aims to promote current research and support growth of multidisciplinary collaborations to advance this groundbreaking research. The meeting gathers scientists working in related areas of computer vision and machine learning, multi-modal signal processing and fusion, human centered computing, behavioral sensing, assistive technologies, and medical tutoring systems for healthcare applications and medicine.

Kristiina Häyrinen