11 research outputs found
Spectral artefacts post sputter-etching and how to cope with them – A case study of XPS on nitride-based coatings using monoatomic and cluster ion beams
No full textThe issue of artefacts due to sputter-etching has been investigated for a group of AlN-based thin film materials with varying thermodynamical stability. Stability of the materials was controlled by alloying AlN with the group 14 elements Si, Ge or Sn in two different concentrations. The coatings were sputter-etched with monoatomic Ar+ with energies between 0.2 and 4.0 keV to study the sensitivity of the materials for sputter damage. The use of Ar-n(+) clusters to remove an oxidised surface layer was also evaluated for a selected sample. The spectra were compared to pristine spectra obtained after in-vacuo sample transfer from the synthesis chamber to the analysis instrument. It was found that the all samples were affected by high energy (4 keV) Ar+ ions to varying degrees. The determining factors for the amount of observed damage were found to be the materials' enthalpy of formation, where a threshold value seems to exist at approximately -1.25 eV/atom (similar to-120 kJ/mol atoms). For each sample, the observed amount of damage was found to have a linear dependence to the energy deposited by the ion beam per volume removed material. Despite the occurrence of sputter-damage in all samples, etching settings that result in almost artefact-free spectral data were found; using either very low energy (i.e. 200 eV) monoatomic ions, or an appropriate combination of ion cluster size and energy. The present study underlines that analysis post sputter-etching must be carried out with an awareness of possible sputter-induced artefacts.</p
Distribution of first and last author sex.
No full textArtificial intelligence (AI) and machine learning are central components of today’s medical environment. The fairness of AI, i.e. the ability of AI to be free from bias, has repeatedly come into question. This study investigates the diversity of members of academia whose scholarship poses questions about the fairness of AI. The articles that combine the topics of fairness, artificial intelligence, and medicine were selected from Pubmed, Google Scholar, and Embase using keywords. Eligibility and data extraction from the articles were done manually and cross-checked by another author for accuracy. Articles were selected for further analysis, cleaned, and organized in Microsoft Excel; spatial diagrams were generated using Public Tableau. Additional graphs were generated using Matplotlib and Seaborn. Linear and logistic regressions were conducted using Python to measure the relationship between funding status, number of citations, and the gender demographics of the authorship team. We identified 375 eligible publications, including research and review articles concerning AI and fairness in healthcare. Analysis of the bibliographic data revealed that there is an overrepresentation of authors that are white, male, and are from high-income countries, especially in the roles of first and last author. Additionally, analysis showed that papers whose authors are based in higher-income countries were more likely to be cited more often and published in higher impact journals. These findings highlight the lack of diversity among the authors in the AI fairness community whose work gains the largest readership, potentially compromising the very impartiality that the AI fairness community is working towards.</div
Global dispersion of first and last author countries.
No full textBase map found at https://www.openstreetmap.org/#map=2/43.8/3.2 and data from OpenStreetMap and OpenStreetMap Foundation. Contains information from OpenStreetMap and OpenStreetMap Foundation, which is made available under the Open Database License, found here: https://www.openstreetmap.org/copyright.</p
A scientometric analysis of fairness in health AI literature.
No full textArtificial intelligence (AI) and machine learning are central components of today's medical environment. The fairness of AI, i.e. the ability of AI to be free from bias, has repeatedly come into question. This study investigates the diversity of members of academia whose scholarship poses questions about the fairness of AI. The articles that combine the topics of fairness, artificial intelligence, and medicine were selected from Pubmed, Google Scholar, and Embase using keywords. Eligibility and data extraction from the articles were done manually and cross-checked by another author for accuracy. Articles were selected for further analysis, cleaned, and organized in Microsoft Excel; spatial diagrams were generated using Public Tableau. Additional graphs were generated using Matplotlib and Seaborn. Linear and logistic regressions were conducted using Python to measure the relationship between funding status, number of citations, and the gender demographics of the authorship team. We identified 375 eligible publications, including research and review articles concerning AI and fairness in healthcare. Analysis of the bibliographic data revealed that there is an overrepresentation of authors that are white, male, and are from high-income countries, especially in the roles of first and last author. Additionally, analysis showed that papers whose authors are based in higher-income countries were more likely to be cited more often and published in higher impact journals. These findings highlight the lack of diversity among the authors in the AI fairness community whose work gains the largest readership, potentially compromising the very impartiality that the AI fairness community is working towards
Additional information and data.
No full textArtificial intelligence (AI) and machine learning are central components of today’s medical environment. The fairness of AI, i.e. the ability of AI to be free from bias, has repeatedly come into question. This study investigates the diversity of members of academia whose scholarship poses questions about the fairness of AI. The articles that combine the topics of fairness, artificial intelligence, and medicine were selected from Pubmed, Google Scholar, and Embase using keywords. Eligibility and data extraction from the articles were done manually and cross-checked by another author for accuracy. Articles were selected for further analysis, cleaned, and organized in Microsoft Excel; spatial diagrams were generated using Public Tableau. Additional graphs were generated using Matplotlib and Seaborn. Linear and logistic regressions were conducted using Python to measure the relationship between funding status, number of citations, and the gender demographics of the authorship team. We identified 375 eligible publications, including research and review articles concerning AI and fairness in healthcare. Analysis of the bibliographic data revealed that there is an overrepresentation of authors that are white, male, and are from high-income countries, especially in the roles of first and last author. Additionally, analysis showed that papers whose authors are based in higher-income countries were more likely to be cited more often and published in higher impact journals. These findings highlight the lack of diversity among the authors in the AI fairness community whose work gains the largest readership, potentially compromising the very impartiality that the AI fairness community is working towards.</div
Distribution of first, last, and all authors’ race.
No full textDistribution of first, last, and all authors’ race.</p
Fig 5 -
No full textDistribution of Male Authorship (A) and White Authorship (B) per article with number of citations and percentage.</p
Distribution of citations among last authors income level.
No full textDistribution of citations among last authors income level.</p
Distribution of citations among last authors based on gender and ethnicity.
No full textDistribution of citations among last authors based on gender and ethnicity.</p
Distribution of each nationality.
No full textBase map layer data found at https://go-cart.io/cartogram, courtesy of go-cart.io, and is available under Creative Commons License CC-BY, found here:https://go-cart.io/faq.</p
