Receive a weekly summary and discussion of the top papers of the week by leading researchers in the field.

General General

AI Moral Enhancement: Upgrading the Socio-Technical System of Moral Engagement.

In Science and engineering ethics

Several proposals for moral enhancement would use AI to augment (auxiliary enhancement) or even supplant (exhaustive enhancement) human moral reasoning or judgment. Exhaustive enhancement proposals conceive AI as some self-contained oracle whose superiority to our own moral abilities is manifest in its ability to reliably deliver the 'right' answers to all our moral problems. We think this is a mistaken way to frame the project, as it presumes that we already know many things that we are still in the process of working out, and reflecting on this fact reveals challenges even for auxiliary proposals that eschew the oracular approach. We argue there is nonetheless a substantial role that 'AI mentors' could play in our moral education and training. Expanding on the idea of an AI Socratic Interlocutor, we propose a modular system of multiple AI interlocutors with their own distinct points of view reflecting their training in a diversity of concrete wisdom traditions. This approach minimizes any risk of moral disengagement, while the existence of multiple modules from a diversity of traditions ensures pluralism is preserved. We conclude with reflections on how all this relates to the broader notion of moral transcendence implicated in the project of AI moral enhancement, contending it is precisely the whole concrete socio-technical system of moral engagement that we need to model if we are to pursue moral enhancement.

Volkman Richard, Gabriels Katleen

2023-Mar-23

AI socratic interlocutor, Artificial intelligence, Artificial moral agent, Moral enhancement

General General

Predicting epileptic seizures based on EEG signals using spatial depth features of a 3D-2D hybrid CNN.

In Medical & biological engineering & computing ; h5-index 32.0

Epilepsy is a recurrent chronic brain disease that affects nearly 75 million people around the world. Therefore, the ability to reliably predict epileptic seizures would be instrumental for implementing interventions to reduce brain injury and improve patients' quality of life. In addition to classical machine learning algorithms and feature engineering methods, the use of electroencephalography (EEG) to predict seizures has gradually become a mainstream trend. Here, we propose a patient-specific method to predict epileptic seizures based on EEG data acquired using spatial depth features of a three-dimensional-two-dimensional hybrid convolutional neural network (3D-2D HyCNN) model. This method facilitates the acquisition of abundant and reliable deep features from multi-channel EEG signals. We first developed a reliable data preprocessing method to reconstruct time-series EEG signals into 3D feature images. Then, the 3D-2D HyCNN model was used to extract correlation features between multiple channels of EEG signals, which are automatically exploited by the network to improve seizure prediction. The method achieved accuracy of 98.43% and 93.11%, sensitivity of 98.58% and 90.98%, and specificity of 96.86% and 92.39% on the CHB-MIT Scalp EEG dataset and the American Epilepsy Society Epilepsy Prediction Challenge dataset, respectively. The results revealed that the new algorithm is reliable. Graphical Abstract A new patient-specific epilepsy prediction approach.

Qi Nan, Piao Yan, Yu Peng, Tan Baolin

2023-Mar-23

3D-2D hybrid CNN, EEG, Epilepsy, Seizure prediction

Surgery Surgery

Effectiveness of high-flow nasal cannulae compared with noninvasive positive-pressure ventilation in preventing reintubation in patients receiving prolonged mechanical ventilation.

In Scientific reports ; h5-index 158.0

Many intensive care unit patients who undergo endotracheal extubation experience extubation failure and require reintubation. Because of the high mortality rate associated with reintubation, postextubation respiratory management is crucial, especially for high-risk populations. We conducted the present study to compare the effectiveness of oxygen therapy administered using high-flow nasal cannulae (HFNC) and noninvasive positive pressure ventilation (NIPPV) in preventing reintubation among patients receiving prolonged mechanical ventilation (PMV). This single-center, prospective, unblinded randomized controlled trial was at the respiratory care center (RCC). Participants were randomized to an HFNC group or an NIPPV group (20 patients in each) and received noninvasive respiratory support (NRS) administered using their assigned method. The primary outcome was reintubation within7 days after extubation. None of the patients in the NIPPV group required reintubation, whereas 5 (25%) of the patients in the HFNC group required reintubation (P = 0.047). The 90-day mortality rates of the NIPPV and HFNC groups (four patients [20%] vs. two patients [10%], respectively) did not differ significantly. No significant differences in length of RCC stay, length of hospital stay, time to liberation from NRS, and ventilator-free days at 28-day were identified. The time to event outcome analysis also revealed that the risk of reintubation in the HFNC group was higher than that in the NIPPV group (P = 0.018). Although HFNC is becoming increasingly common as a form of postextubation NRS, HFNC may not be as effective as NIPPV in preventing reintubation among patients who have been receiving PMV for at least 2 weeks. Additional studies evaluating HFNC as an alternative to NIPPV for patients receiving PMV are warranted.ClinicalTrial.gov ID: NCT04564859; IRB number: 20160901R.Trial registration: ClinicalTrial.gov ( https://clinicaltrials.gov/ct2/show/NCT04564859 ).

Tseng Chi-Wei, Chao Ke-Yun, Wu Hsiu-Li, Lin Chen-Chun, Hsu Han-Shui

2023-Mar-22

General General

Machine Ethics: Do Androids Dream of Being Good People?

In Science and engineering ethics

Is ethics a computable function? Can machines learn ethics like humans do? If teaching consists in no more than programming, training, indoctrinating… and if ethics is merely following a code of conduct, then yes, we can teach ethics to algorithmic machines. But if ethics is not merely about following a code of conduct or about imitating the behavior of others, then an approach based on computing outcomes, and on the reduction of ethics to the compilation and application of a set of rules, either a priori or learned, misses the point. Our intention is not to solve the technical problem of machine ethics, but to learn something about human ethics, and its rationality, by reflecting on the ethics that can and should be implemented in machines. Any machine ethics implementation will have to face a number of fundamental or conceptual problems, which in the end refer to philosophical questions, such as: what is a human being (or more generally, what is a worthy being); what is human intentional acting; and how are intentional actions and their consequences morally evaluated. We are convinced that a proper understanding of ethical issues in AI can teach us something valuable about ourselves, and what it means to lead a free and responsible ethical life, that is, being good people beyond merely "following a moral code". In the end we believe that rationality must be seen to involve more than just computing, and that value rationality is beyond numbers. Such an understanding is a required step to recovering a renewed rationality of ethics, one that is urgently needed in our highly technified society.

Génova Gonzalo, Moreno Valentín, González M Rosario

2023-Mar-23

Artificial intelligence, Computability, Intentional action, Machine ethics, Moral codes of conduct, Rationality of ethics

General General

A device-independent method for the colorimetric quantification on microfluidic sensors using a color adaptation algorithm.

In Mikrochimica acta

A general and adaptable method is proposed to reliably extract quantitative information from smartphone images of microfluidic sensors. By analyzing and processing the color information of selected standard substances, the influence of light conditions, device differences, and human factors could be significantly reduced. Machine learning and multivariate fitting methods were proved to be effective for chroma correction, and a key element was the training of sample size and the fitting form, respectively. A custom APP was developed and validated using a high-sensitivity chromium ion quantification paper chip. The average chroma deviations under different conditions were reduced by more than 75% in RGB color space, and the concentration test error was reduced by more than half compared with the commonly used method. The proposed approach could be a beneficial supplement to existing and potential colorimetry-based detection methods.

Feng Junjie, Jiang Huiyun, Jin Yan, Rong Shenghui, Wang Shiqiang, Wang Haozhi, Wang Lin, Xu Wei, Sun Bing

2023-Mar-23

Colorimetry, Image processing, Microfluidic sensors, Paper-based analytical devices, Smartphone

Public Health Public Health

Understanding Public Attitudes and Willingness to Share Commercial Data for Health Research: Survey Study in the United Kingdom.

In JMIR public health and surveillance

BACKGROUND : Health research using commercial data is increasing. The evidence on public acceptability and sociodemographic characteristics of individuals willing to share commercial data for health research is scarce.

OBJECTIVE : This survey study investigates the willingness to share commercial data for health research in the United Kingdom with 3 different organizations (government, private, and academic institutions), 5 different data types (internet, shopping, wearable devices, smartphones, and social media), and 10 different invitation methods to recruit participants for research studies with a focus on sociodemographic characteristics and psychological predictors.

METHODS : We conducted a web-based survey using quota sampling based on age distribution in the United Kingdom in July 2020 (N=1534). Chi-squared tests tested differences by sociodemographic characteristics, and adjusted ordered logistic regressions tested associations with trust, perceived importance of privacy, worry about data misuse and perceived risks, and perceived benefits of data sharing. The results are shown as percentages, adjusted odds ratios, and 95% CIs.

RESULTS : Overall, 61.1% (937/1534) of participants were willing to share their data with the government and 61% (936/1534) of participants were willing to share their data with academic research institutions compared with 43.1% (661/1534) who were willing to share their data with private organizations. The willingness to share varied between specific types of data-51.8% (794/1534) for loyalty cards, 35.2% (540/1534) for internet search history, 32% (491/1534) for smartphone data, 31.8% (488/1534) for wearable device data, and 30.4% (467/1534) for social media data. Increasing age was consistently and negatively associated with all the outcomes. Trust was positively associated with willingness to share commercial data, whereas worry about data misuse and the perceived importance of privacy were negatively associated with willingness to share commercial data. The perceived risk of sharing data was positively associated with willingness to share when the participants considered all the specific data types but not with the organizations. The participants favored postal research invitations over digital research invitations.

CONCLUSIONS : This UK-based survey study shows that willingness to share commercial data for health research varies; however, researchers should focus on effectively communicating their data practices to minimize concerns about data misuse and improve public trust in data science. The results of this study can be further used as a guide to consider methods to improve recruitment strategies in health-related research and to improve response rates and participant retention.

Hirst Yasemin, Stoffel Sandro T, Brewer Hannah R, Timotijevic Lada, Raats Monique M, Flanagan James M

2023-Mar-23

acceptability, commercial data, data, data donation, data sharing, digital, health, loyalty cards, mobile phone, participant recruitment, public, sociodemographic factors