Mapping the Bibliometrics Landscape of AI in Medicine: Methodological Study

BACKGROUND

Artificial intelligence (AI), conceived in the 1950s, has permeated numerous industries, intensifying in tandem with advancements in computing power. Despite the widespread adoption of AI, its integration into medicine trails other sectors. However, medical AI research has experienced substantial growth, attracting considerable attention from researchers and practitioners.

OBJECTIVE

In the absence of an existing framework, this study aims to outline the current landscape of medical AI research and provide insights into its future developments by examining all AI-related studies within PubMed over the past 2 decades. We also propose potential data acquisition and analysis methods, developed using Python (version 3.11) and to be executed in Spyder IDE (version 5.4.3), for future analogous research.

METHODS

Our dual-pronged approach involved (1) retrieving publication metadata related to AI from PubMed (spanning 2000-2022) via Python, including titles, abstracts, authors, journals, country, and publishing years, followed by keyword frequency analysis and (2) classifying relevant topics using latent Dirichlet allocation, an unsupervised machine learning approach, and defining the research scope of AI in medicine. In the absence of a universal medical AI taxonomy, we used an AI dictionary based on the European Commission Joint Research Centre AI Watch report, which emphasizes 8 domains: reasoning, planning, learning, perception, communication, integration and interaction, service, and AI ethics and philosophy.

RESULTS

From 2000 to 2022, a comprehensive analysis of 307,701 AI-related publications from PubMed highlighted a 36-fold increase. The United States emerged as a clear frontrunner, producing 68,502 of these articles. Despite its substantial contribution in terms of volume, China lagged in terms of citation impact. Diving into specific AI domains, as the Joint Research Centre AI Watch report categorized, the learning domain emerged dominant. Our classification analysis meticulously traced the nuanced research trajectories across each domain, revealing the multifaceted and evolving nature of AI's application in the realm of medicine.

CONCLUSIONS

The research topics have evolved as the volume of AI studies increases annually. Machine learning remains central to medical AI research, with deep learning expected to maintain its fundamental role. Empowered by predictive algorithms, pattern recognition, and imaging analysis capabilities, the future of AI research in medicine is anticipated to concentrate on medical diagnosis, robotic intervention, and disease management. Our topic modeling outcomes provide a clear insight into the focus of AI research in medicine over the past decades and lay the groundwork for predicting future directions. The domains that have attracted considerable research attention, primarily the learning domain, will continue to shape the trajectory of AI in medicine. Given the observed growing interest, the domain of AI ethics and philosophy also stands out as a prospective area of increased focus.

Involving Health Care Professionals in the Development of Electronic Health Records: Scoping Review

BACKGROUND

Electronic health records (EHRs) are a promising approach to document and map (complex) health information gathered in health care worldwide. However, possible unintended consequences during use, which can occur owing to low usability or the lack of adaption to existing workflows (eg, high cognitive load), may pose a challenge. To prevent this, the involvement of users in the development of EHRs is crucial and growing. Overall, involvement is designed to be very multifaceted, for example, in terms of the timing, frequency, or even methods used to capture user preferences.

OBJECTIVE

Setting, users and their needs, and the context and practice of health care must be considered in the design and subsequent implementation of EHRs. Many different approaches to user involvement exist, each requiring a variety of methodological choices. The aim of the study was to provide an overview of the existing forms of user involvement and the circumstances they need and to provide support for the planning of new involvement processes.

METHODS

We conducted a scoping review to provide a database for future projects on which design of inclusion is worthwhile and to show the diversity of reporting. Using a very broad search string, we searched the PubMed, CINAHL, and Scopus databases. In addition, we searched Google Scholar. Hits were screened according to scoping review methodology and then examined, focusing on methods and materials, participants, frequency and design of the development, and competencies of the researchers involved.

RESULTS

In total, 70 articles were included in the final analysis. There was a wide range of methods of involvement. Physicians and nurses were the most frequently included groups and, in most cases, were involved only once in the process. The approach of involvement (eg, co-design) was not specified in most of the studies (44/70, 63%). Further qualitative deficiencies in the reporting were evident in the presentation of the competences of members of the research and development teams. Think-aloud sessions, interviews, and prototypes were frequently used.

CONCLUSIONS

This review provides insights into the diversity of health care professionals' involvement in the development of EHRs. It provides an overview of the different approaches in various fields of health care. However, it also shows the necessity of considering quality standards in the development of EHRs together with future users and the need for reporting this in future studies.

AMICAI: A Method Based on Risk Analysis to Integrate Responsible Research and Innovation into the Work of Research and Innovation Practitioners

The integration of ethics into the day-to-day work of research and innovation (R&I) is an important but difficult challenge. However, with the Aachen method for identification, classification and risk analysis of innovation-based problems (AMICAI) an approach from an engineering perspective is presented that enables the integration of ethical, legal and social implications into the day-to-day work of R&I practitioners. AMICAI appears in particular capable of providing a procedural guidance for R&I practitioners based on a method established in engineering science, breaking down the object of consideration into partial aspects and prioritizing the innovation-based problems in dependence of potential risk. This enables the user to apply AMICAI continuously during all stages of the research and development (R&D) process and to analyze and choose between certain sociotechnical alternatives. In this way, problems that affect ethical, legal, and social aspects can be understood, reflected and considered in the mostly technically focused R&D process. The paper gives a general guidance about AMICAI by describing principles and assumptions, providing the steps of analysis and application aids, giving an example application, explaining the necessary adjustments of AMICAI compared to the methodical basis of failure mode, effects, and criticality analysis and discussing the advantages and limits. AMICAI's simple applications can stimulate interdisciplinary cooperation in the R&D process and be a starting point for the development of an "open RRI risk analysis platform" allowing society to evaluate innovation-based problems.

[Evaluation of a simple screening tool for ambulant fall prevention]

BACKGROUND

An individual's risk of falling is generally difficult to detect and it is likely to be underestimated. Thus, preventive measures are challenging and they demand sufficient integration and implementation into aftercare and outpatient management. The Aachen Falls Prevention Scale (AFPS) is a quick and easy tool for patient-driven fall risk assessment. Older adults' risk of falling is identified in a suitable manner and they then have the opportunity to independently assess and monitor their risk of falling.

OBJECTIVES

The aim of the current study was to evaluate the AFPS as a simple screening tool in geriatric trauma patients via the identification of influencing factors, e.g. objective or subjective fall risk, fear of falling (FOF) and demographic data. In this context, we investigated older adults' willingness to take part in special activities concerning fall prevention.

METHODS

Retrospectively, all patients over 70 years of age who received in-hospital fracture treatment between July 2014 and April 2016 were analyzed at a level I trauma center. After identification of 884 patients, participants completed a short questionnaire (47 questions, yes/no, Likert scale) comprising the AFPS. A history of falls in the past year was considered an indicator of a balance disorder. In addition, ambulant patients were invited to participate between July and August 2016.

RESULTS

In total, 201 patients (mean 80.4 years, range 63-97 years) performed a self-assessment based on the AFPS. After steps 1 and 2 of the AFPS had been completed, 95 (47%) participants rated their subjective risk of falling as high (more than 5 points). Of the participants 84 (42%) were objectively classified as "fallers" with significant effects on their AFPS evaluation and rating of their subjective risk of falling. Furthermore, 67% of the participants identified a general practitioner as their main contact person, and 43% of the respondents viewed the AFPS as a beneficial screening tool in fall risk evaluation (8% negative attitudes). Only 12% of the participants could imagine using the AFPS app version as a feasible option.

CONCLUSION

It would be advantageous to pretest at-risk individuals in their environment using a simple self-assessment approach, with the main purpose of identifying potential balance problems. With this approach, cost savings in the healthcare system are possible, combined with a higher health-related quality of life in the geriatric population.

Predicting technology usage by health information need of older adults: Implications for eHealth technology

BACKGROUND

Information and communication technology increasingly addresses the information needs patients have regarding their personal health. While an understanding of older adults' needs is crucial for developing successful eHealth technology, user research results hardly apply to different systems.

OBJECTIVE

The present study aims at: (1) describing and analysing the context of digital health systems in a general manner, (2) investigating if information need of older adults influences their technology usage to show the relevance of the concept for a general context analysis and (3) testing which demographic variables intervene with their health information need.

METHODS

Survey data from a longitudinal study with older adults (N = 551) were reported descriptively. After showing a significant relationship during chi-square tests, we quantified the ones between general health information need and technology usage, as well as between general health information need and the demographic variables age, education, chronic diseases and gender by means of (multiple) linear regression models.

RESULTS

We predicted older adults' technology usage based on their health information need. The results confirmed this relationship. Higher information need led to a more frequent usage of apps installed on the tablet personal computer (PC), to a frequent use of smartwatches and to the possession of a computer or laptop. Users' education has a higher impact on health information need than amount of chronic diseases, gender and age.

CONCLUSIONS

Information need emerged as a useful object for investigation of context and user requirement analysis across different systems: it predicted technology usage so that design recommendations derived from the descriptive gained in importance.

The Aachen Falls Prevention Scale: Multi-Study Evaluation and Comparison

Background

Fall risk assessment is a time-consuming and resource-intensive activity. Patient-driven self-assessment as a preventive measure might be a solution to reduce the number of patients undergoing a full clinical fall risk assessment.

Objective

The aim of this study was (1) to analyze test accuracy of the Aachen Falls Prevention Scale (AFPS) and (2) to compare these results with established fall risk assessment measures identified by a review of systematic reviews.

Methods

Sensitivity, specificity, and receiver operating curves (ROC) of the AFPS were calculated based on data retrieved from 2 independent studies using the AFPS. Comparison with established fall risk assessment measures was made by conducting a review of systematic reviews and corresponding meta-analysis. Electronic databases PubMed, Web of Science, and EMBASE were searched for systematic reviews and meta-analyses that reviewed fall risk assessment measures between the years 2000 and 2018. The review of systematic reviews was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis statement. The Revised Assessment of Multiple SysTemAtic Reviews (R-AMSTAR) was used to assess the methodological quality of reviews. Sensitivity, specificity, and ROC were extracted from each review and compared with the calculated values of the AFPS.

Results

Sensitivity, specificity, and ROC of the AFPS were evaluated based on 2 studies including a total of 259 older adults. Regarding the primary outcome of the AFPS subjective risk of falling, pooled sensitivity is 57.0% (95% CI 0.467-0.669) and specificity is 76.7% (95% CI 0.694-0.831). If 1 out of the 3 subscales of the AFPS is used to predict a fall risk, pooled sensitivity could be increased up to 90.0% (95% CI 0.824-0.951), whereas mean specificity thereby decreases to 50.0% (95% CI 0.42-0.58). A systematic review for fall risk assessment measures produced 1478 articles during the study period, with 771 coming from PubMed, 530 from Web of Science, and 177 from EMBASE. After eliminating doublets and assessing full text, 8 reviews met the inclusion criteria. All were of sufficient methodological quality (R-AMSTAR score ≥22). A total number of 9 functional or multifactorial fall risk assessment measures were extracted from identified reviews, including Timed Up and Go test, Berg Balance Scale, Performance-Oriented Mobility Assessment, St Thomas’s Risk Assessment Tool in Falling Elderly, and Hendrich II Fall Risk Model. Comparison of these measures with pooled sensitivity and specificity of the AFPS revealed a sufficient quality of the AFPS in terms of a patient-driven self-assessment tool.

Conclusions

It could be shown that the AFPS reaches a test accuracy comparable with that of the established methods in this initial investigation. However, it offers the advantage that the users can perform the self-assessment independently at home without involving trained health care professionals.

Seamless recording of glucometer measurements among older experienced diabetic patients - A study of perception and usability

Self-measurement and documentation of blood-glucose are critical elements of diabetes management, particularly in regimes including insulin. In this study, we analyze the usability of iBG-STAR, the first blood glucose meter connectable to a smartphone. This technology records glucometer measurements, removing the burden of documentation from diabetic patients. This study assesses the potential for implementation of iBG-STAR in routine care. Twelve long-term diabetic patients (4 males; median age of 66.5 years) were enrolled in the study. N = 4/12 reported diabetic polyneuropathy. Reported subjective mental workload for all tasks related to iBG-STAR was on average lower than 12 points, corresponding to the verbal code 'nearly no effort needed'. A "Post Study System Usability Questionnaire", evaluated the glucometer at an average value of 2.06 (SD = 1.02) on a 7-Likert-scale (1 = 'I fully agree' to 7 = 'I completely disagree') for usability. These results represent a positive user-experience. Patients with polyneuropathy may experience physical difficulties in completing the tasks, thereby affecting usability. Technologically savvy patients (n = 6) with a positive outlook on diabetes assessed the product as a suitable tool for themselves and would recommend to other diabetic patients. The main barrier to regular use was treating physicians' inability to retrieve digitally recorded data. This barrier was due to a shortcoming in interoperability of mobile devices and medical information systems.

Satisfying Product Features of a Fall Prevention Smartphone App and Potential Users' Willingness to Pay: Web-Based Survey Among Older Adults

Background

Prohibiting falls and fall-related injuries is a major challenge for health care systems worldwide, as a substantial proportion of falls occur in older adults who are previously known to be either frail or at high risk for falls. Hence, preventive measures are needed to educate and minimize the risk for falls rather than just minimize older adults’ fall risk. Health apps have the potential to address this problem, as they enable users to self-assess their individual fall risk.

Objective

The objective of this study was to identify product features of a fall prevention smartphone app, which increase or decrease users’ satisfaction. In addition, willingness to pay (WTP) was assessed to explore how much revenue such an app could generate.

Methods

A total of 96 participants completed an open self-selected Web-based survey. Participants answered various questions regarding health status, subjective and objective fall risk, and technical readiness. Seventeen predefined product features of a fall prevention smartphone app were evaluated twice: first, according to a functional (product feature is implemented in the app), and subsequently by a dysfunctional (product feature is not implemented in the app) question. On the basis of the combination of answers from these 2 questions, the product feature was assigned to a certain category (must-be, attractive, one-dimensional, indifferent, or questionable product feature). This method is widely used in user-oriented product development and captures users’ expectations of a product and how their satisfaction is influenced by the availability of individual product features.

Results

Five product features were identified to increase users’ acceptance, including (1) a checklist of typical tripping hazards, (2) an emergency guideline in case of a fall, (3) description of exercises and integrated workout plans that decrease the risk of falling, (4) inclusion of a continuous workout program, and (5) cost coverage by health insurer. Participants’ WTP was assessed after all 17 product features were rated and revealed a median monthly payment WTP rate of €5.00 (interquartile range 10.00).

Conclusions

The results show various motivating product features that should be incorporated into a fall prevention smartphone app. Results reveal aspects that fall prevention and intervention designers should keep in mind to encourage individuals to start joining their program and facilitate long-term user engagement, resulting in a greater interest in fall risk prevention.

Prevalence of Health App Use Among Older Adults in Germany: National Survey

Background

Health apps are increasingly becoming an integral part of health care. Especially in older adults, the self-management of chronic diseases by health apps might become an integral part of health care services.

Objective

The aim of this explorative study was to investigate the prevalence of health app use and related demographic factors, as well as health status among older adults in Germany.

Methods

A nationwide postal survey was conducted. Of the 5000 individuals contacted, a total of 576 participants completed this survey. On the basis of their self-indicated assignment to one of the three predefined user groups (health app users, general app users, and nonusers of apps), participants answered various questions regarding app and health app use, including frequency of use and number of installed apps, demographic factors, and health status.

Results

In total, 16.5% (95/576) used health apps, whereas 37.5% (216/576) indicated only using general apps, and 46.0% (265/576) reported using no apps at all. The number of installed health apps was most frequently reported as between 1 and 5 apps per participant, which were usually used on a weekly basis. The most frequently cited type of health apps were exercise-related ones. Individuals using health apps were found to be younger (MeanmHealth 66.6, SD 4.7) and to have a higher level of technical readiness compared with general app users and nonusers of apps (adjusted odds ratio, AOR=4.02 [95% CI 2.23-7.25] for technical readiness, and AOR=0.905 [95% CI 0.85-0.97] for age). The most frequently mentioned sources of information about apps within the group of health and general app users were family and friends. Identified barriers against the use of health apps were a lack of trust, data privacy concerns, and fear of misdiagnosis.

Conclusions

Health apps are already used by older adults in Germany. The main type of apps used are exercise-related ones. Barriers to and incentives for the use of health apps and associations with health status and users’ demographics were revealed.

The "Aachen fall prevention App" - a Smartphone application app for the self-assessment of elderly patients at risk for ground level falls

Background

Fall incidents are a major problem for patients and healthcare. The “Aachen Fall Prevention App” (AFPA) represents the first mobile Health (mHealth) application (app) empowering older patients (persons 50+ years) to self-assess and monitor their individual fall risk. Self-assessment is based on the “Aachen Fall Prevention Scale,” which consists of three steps. First, patients answer ten standardized yes–no questions (positive criterion ≥ 5 “Yes” responses). Second, a ten-second test of free standing without compensatory movement is performed (positive criterion: compensatory movement). Finally, during the third step, patients rate their subjective fall risk on a 10-point Likert scale, based on the results of steps one and two. The purpose of this app is (1) to offer a low-threshold service through which individuals can independently monitor their individual fall risk and (2) to collect data about how a patient-centered mHealth app for fall risk assessment is used in the field.

Results

The results represent the first year of an ongoing field study. From December 2015 to December 2016, 197 persons downloaded the AFPA (iOS^™ and Android^™; free of charge). N = 111 of these persons voluntarily shared their data and thereby participated in the field study. Data from a final number of n = 79 persons were analyzed due to exclusion criteria (age, missing objective fall risk, missing self-assessment). The objective fall risk and the self-assessed subjective risk measured by the AFPA showed a significant positive relationship.

Conclusions

The “Aachen Fall Prevention App” (AFPA) is an mHealth app released for iOS and Android. This field study revealed the AFPA as a promising tool to raise older adults’ awareness of their individual fall risk by means of a low-threshold patient-driven fall risk assessment tool.

Who Is Still Playing Pokémon Go? A Web-Based Survey

Background

Poor physical activity is one of the major health care problems in Western civilizations. Various digital gadgets aiming to increase physical activity, such as activity trackers or fitness apps, have been introduced over recent years. The newest products are serious games that incorporate real-life physical activity into their game concept. Recent studies have shown that such games increase the physical activity of their users over the short term.

Objective

In this study, we investigated the motivational effects of the digital game “Pokémon Go” leading to continued use or abandonment of the game. The aim of the study was to determine aspects that motivate individuals to play augmented reality exergames and how this motivation can be used to strengthen the initial interest in physical activity.

Methods

A total of 199 participants completed an open self-selected Web-based survey. On the basis of their self-indicated assignment to one of three predefined user groups (active, former, and nonuser of Pokémon Go), participants answered various questions regarding game experience, physical activity, motivation, and personality as measured by the Big Five Inventory.

Results

In total, 81 active, 56 former, and 62 nonusers of Pokémon Go were recruited. When asked about the times they perform physical activity, active users stated that they were less physically active in general than former and nonusers. However, based on a subjective rating, active users were more motivated to be physically active due to playing Pokémon Go. Motivational aspects differed for active and former users, whereas fan status was the same within both groups. Active users are more motivated by features directly related to Pokémon, such as catching all possible Pokémon and reaching higher levels, whereas former users stress the importance of general game quality, such as better augmented reality and more challenges in the game. Personality did not affect whether a person started to play Pokémon Go nor their abandonment of the game.

Conclusions

The results show various motivating elements that should be incorporated into augmented reality exergames based on the game Pokémon Go. We identified different user types for whom different features of the game contribute to maintained motivation or abandonment. Our results show aspects that augmented reality exergame designers should keep in mind to encourage individuals to start playing their game and facilitate long-term user engagement, resulting in a greater interest in physical activity.