Privacy Preserving Technology using Federated Learning and Blockchain in protecting against Adversarial Attacks for Retinal Imaging.

PURPOSE: Collaboration provides valuable data for robust artificial intelligence (AI) model development. Federated learning (FL) is a privacy enhancing technology that allows collaboration while respecting privacy via the development of models without raw data transfer. However state-of-the-art FL models still face challenges in non-independent and identically distributed (non-i.i.d.) health care settings and remain susceptible to privacy breaches. We propose a novel FL framework coupled with blockchain technology to address these challenges. DESIGN: Retrospective multicohort study SUBJECTS AND METHODS: 27,145 images from Singapore, China and Taiwan were used to design a novel FL aggregation method for the detection of myopic macular degeneration (MMD) from fundus photographs and macular disease from optical coherence tomography (OCT) scans in feature distribution skew and label distribution imbalance scenarios. We further performed adversarial attacks (label flipping and clean label). As proof of concept, blockchain was incorporated into FL to demonstrate secure transfer of model updates across collaborating sites. MAIN OUTCOME MEASURES: We evaluated our FL model performance in MMD and OCT classification and compared our model against state-of the-art FL and centralized models. RESULTS: Our FL model showed robust performance with areas under the receiving operating characteristic curves (AUC) of 0.868±0.009 for MMD detection and 0.970±0.012 for OCT macular disease classification. In label flipping attack, our FL model had an AUC of 0.861±0.019, similar to the centralized model (AUC 0.856± 0.015) and higher than other FL models (AUC 0.578-0.819) In clean label attack, our FL model had an AUC of 0.878±0.006 which was comparable to the centralized model (AUC 0.878±0.001) and superior to other state-of-the-art FL models with AUC of 0.529-0.838. Simulation showed that the additional time with blockchain in one global epoch was around 5 seconds. The addition of blockchain to the FL framework was feasible with a minimal impact on model development time. CONCLUSIONS: Our proposed FL algorithm overcomes the shortcoming of the traditional FL in non i.i.d. situations and remains robust to against adversarial attacks. The integration of blockchain adds further security during the transfer of model updates. Blockchain-enabled FL can be a trusted platform for collaborative health AI research.

Artificial Intelligence-Based Disease Activity Monitoring to Personalized Neovascular Age-Related Macular Degeneration Treatment: A Feasibility Study.

Purpose: To evaluate the performance of a disease activity (DA) model developed to detect DA in participants with neovascular age-related macular degeneration (nAMD). Design: Post hoc analysis. Participants: Patient dataset from the phase III HAWK and HARRIER (H&H) studies. Methods: An artificial intelligence (AI)-based DA model was developed to generate a DA score based on measurements of OCT images and other parameters collected from H&H study participants. Disease activity assessments were classified into 3 categories based on the extent of agreement between the DA model's scores and the H&H investigators' decisions: agreement ("easy"), disagreement ("noisy"), and close to the decision boundary ("difficult"). Then, a panel of 10 international retina specialists ("panelists") reviewed a sample of DA assessments of these 3 categories that contributed to the training of the final DA model. A panelists' majority vote on the reviewed cases was used to evaluate the accuracy, sensitivity, and specificity of the DA model. Main Outcome Measures: The DA model's performance in detecting DA compared with the DA assessments made by the investigators and panelists' majority vote. Results: A total of 4472 OCT DA assessments were used to develop the model; of these, panelists reviewed 425, categorized as "easy" (17.2%), "noisy" (20.5%), and "difficult" (62.4%). False-positive and false negative rates of the DA model's assessments decreased after changing the assessment in some cases reviewed by the panelists and retraining the DA model. Overall, the DA model achieved 80% accuracy. For "easy" cases, the DA model reached 96% accuracy and performed as well as the investigators (96% accuracy) and panelists (90% accuracy). For "noisy" cases, the DA model performed similarly to panelists and outperformed the investigators (84%, 86%, and 16% accuracies, respectively). The DA model also outperformed the investigators for "difficult" cases (74% and 53% accuracies, respectively) but underperformed the panelists (86% accuracy) owing to lower specificity. Subretinal and intraretinal fluids were the main clinical parameters driving the DA assessments made by the panelists. Conclusions: These results demonstrate the potential of using an AI-based DA model to optimize treatment decisions in the clinical setting and in detecting and monitoring DA in patients with nAMD. Financial Disclosures: Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

Awareness and attitudes of elderly Southeast Asian adults towards telehealth during the COVID-19 pandemic: a qualitative study.

Introduction: We aimed to understand the awareness and attitudes of elderly Southeast Asians towards telehealth services during the coronavirus disease 2019 (COVID-19) pandemic in this study. Methods: In this qualitative study, 78 individuals from Singapore (51.3% female, mean age 73.0 ± 7.6 years) were interviewed via telephone between 13 May 2020 and 9 June 2020 during Singapore's first COVID-19 'circuit breaker'. Participants were asked to describe their understanding of telehealth, their experience of and willingness to utilise these services, and the barriers and facilitators underlying their decision. Transcripts were analysed using thematic analysis, guided by the United Theory of Acceptance Use of Technology framework. Results: Of the 78 participants, 24 (30.8%) were able to describe the range of telehealth services available and 15 (19.2%) had previously utilised these services. Conversely, 14 (17.9%) participants thought that telehealth comprised solely home medication delivery and 50 (51.3%) participants did not know about telehealth. Despite the advantages offered by telehealth services, participants preferred in-person consultations due to a perceived lack of human interaction and accuracy of diagnoses, poor digital literacy and a lack of access to telehealth-capable devices. Conclusion: Our results showed poor overall awareness of the range of telehealth services available among elderly Asian individuals, with many harbouring erroneous views regarding their use. These data suggest that public health education campaigns are needed to improve awareness of and correct negative perceptions towards telehealth services in elderly Asians.

Logistic regression was as good as machine learning for predicting major chronic diseases.

OBJECTIVE: To evaluate the performance of machine learning (ML) algorithms and to compare them with logistic regression for the prediction of risk of cardiovascular diseases (CVDs), chronic kidney disease (CKD), diabetes (DM), and hypertension (HTN) and in a prospective cohort study using simple clinical predictors. STUDY DESIGN AND SETTING: We conducted analyses in a population-based cohort study in Asian adults (n = 6,762). Five different ML models were considered-single-hidden-layer neural network, support vector machine, random forest, gradient boosting machine, and k-nearest neighbor-and were compared with standard logistic regression. RESULTS: The incidences at 6 years of CVD, CKD, DM, and HTN cases were 4.0%, 7.0%, 9.2%, and 34.6%, respectively. Logistic regression reached the highest area under the receiver operating characteristic curve for CKD (0.905 [0.88, 0.93]) and DM (0.768 [0.73, 0.81]) predictions. For CVD and HTN, the best models were neural network (0.753 [0.70, 0.81]) and support vector machine (0.780 [0.747, 0.812]), respectively. However, the differences with logistic regression were small (less than 1%) and nonsignificant. Logistic regression, gradient boosting machine, and neural network were systematically ranked among the best models. CONCLUSION: Logistic regression yields as good performance as ML models to predict the risk of major chronic diseases with low incidence and simple clinical predictors.

Deep learning for detecting retinal detachment and discerning macular status using ultra-widefield fundus images.

Retinal detachment can lead to severe visual loss if not treated timely. The early diagnosis of retinal detachment can improve the rate of successful reattachment and the visual results, especially before macular involvement. Manual retinal detachment screening is time-consuming and labour-intensive, which is difficult for large-scale clinical applications. In this study, we developed a cascaded deep learning system based on the ultra-widefield fundus images for automated retinal detachment detection and macula-on/off retinal detachment discerning. The performance of this system is reliable and comparable to an experienced ophthalmologist. In addition, this system can automatically provide guidance to patients regarding appropriate preoperative posturing to reduce retinal detachment progression and the urgency of retinal detachment repair. The implementation of this system on a global scale may drastically reduce the extent of vision impairment resulting from retinal detachment by providing timely identification and referral.