To demonstrate the identification of corneal diseases using a novel deep learning algorithm. A novel hierarchical deep learning network, which is composed of a family of multi-task multi-label learning classifiers representing different levels of eye diseases derived from a predefined hierarchical eye disease taxonomy was designed. Next, we proposed a multi-level eye disease-guided loss function to learn the fine-grained variability of eye diseases features. The proposed algorithm was trained end-to-end directly using 5,325 ocular surface images from a retrospective dataset. Finally, the algorithm’s performance was tested against 10 ophthalmologists in a prospective clinic-based dataset with 510 outpatients newly enrolled with diseases of infectious keratitis, non-infectious keratitis, corneal dystrophy or degeneration, and corneal neoplasm. The area under the ROC curve of the algorithm for each corneal disease type was over 0.910 and in general it had sensitivity and specificity similar to or better than the average values of all ophthalmologists. Confusion matrices revealed similarities in misclassification between human experts and the algorithm. In addition, our algorithm outperformed over all four previous reported methods in identified corneal diseases. The proposed algorithm may be useful for computer-assisted corneal disease diagnosis.

Background Conversational agents (CAs) have been developed in outpatient departments to improve physician-patient communication efficiency. As end users, patients’ continuance intention is essential for the sustainable development of CAs. Objective The aim of this study was to facilitate the successful usage of CAs by identifying key factors influencing patients’ continuance intention and proposing corresponding managerial implications. Methods This study proposed an extended expectation-confirmation model and empirically tested the model via a cross-sectional field survey. The questionnaire included demographic characteristics, multiple-item scales, and an optional open-ended question on patients’ specific expectations for CAs. Partial least squares structural equation modeling was applied to assess the model and hypotheses. The qualitative data were analyzed via thematic analysis. Results A total of 172 completed questionaries were received, with a 100% (172/172) response rate. The proposed model explained 75.5% of the variance in continuance intention. Both satisfaction (β=.68; P<.001) and perceived usefulness (β=.221; P=.004) were significant predictors of continuance intention. Patients' extent of confirmation significantly and positively affected both perceived usefulness (β=.817; P<.001) and satisfaction (β=.61; P<.001). Contrary to expectations, perceived ease of use had no significant impact on perceived usefulness (β=.048; P=.37), satisfaction (β=−.004; P=.63), and continuance intention (β=.026; P=.91). The following three themes were extracted from the 74 answers to the open-ended question: personalized interaction, effective utilization, and clear illustrations. Conclusions This study identified key factors influencing patients’ continuance intention toward CAs. Satisfaction and perceived usefulness were significant predictors of continuance intention (P<.001 and P<.004, respectively) and were significantly affected by patients’ extent of confirmation (P<.001 and P<.001, respectively). Developing a better understanding of patients’ continuance intention can help administrators figure out how to facilitate the effective implementation of CAs. Efforts should be made toward improving the aspects that patients reasonably expect CAs to have, which include personalized interactions, effective utilization, and clear illustrations.

BACKGROUND Conversational agents(CAs) have been developed in outpatient departments to improve doctor-patient communication efficiency. As end users, patients’ continuance intention is essential for the sustainable development of the agents. OBJECTIVE The aim of this study was to identify key factors influencing patients’ continuance intention towards CAs and provide corresponding optimization strategies. METHODS This study proposed an extended expectation-confirmation model and empirically tested the model via a cross-sectional field survey at Shanghai eye and ENT hospital outpatient department. Structured interviews were conducted and analyzed through manual deductive coding to better understand the confirmation of patients’ initial expectations. Partial least squares structural equation modelling was applied to assess the model and hypotheses. RESULTS A total of 172 completed questionaries were received (100% response rate). The proposed model explained 75.5% of the variance of continuance intention. Satisfaction (β = .68; P<.001) and perceived usefulness (β = .221; P=.004) were both significant predictors of continuance intention. Patients' extent of confirmation significantly and positively affected both perceived usefulness (β = .817; P<.001) and satisfaction (β = .61; P<.001). Contrary to expectations, perceived ease of use had no significant impact on perceived usefulness (β = .048; P=.367), satisfaction (β =- .004; P=.633) and continuance intention (β = .026; P=.91).74 of 172 patients participated in the structured interviews. The answers were categorized into three aspects, personalized interaction (mentioned 50 times), effective utilization (mentioned 37 times) and publicity on benefits (mentioned 5 times). CONCLUSIONS Implementing CA systems is not once and for all. By identifying key factors influencing patients’ continuance intention, this study provides a feedback channel for the administrators to gain insight into patients’ experience and thoughts. To maintain patients’ continuance intention and promote the sustainable development of CAs, efforts should be made towards maximizing patients’ satisfaction and perceived usefulness. Feasible measures include understanding patients’ expectations after their actual use and optimizing CA systems accordingly.