METHODS: We propose to use 3D Axial-Attention, which requires a fraction of the computing power of a regular Non-Local network (i.e., self-attention). Unlike a regular Non-Local network, the 3D Axial-Attention network applies the attention operation to each axis separately. Additionally, we solve the invariant position problem of the Non-Local network by proposing to add 3D positional encoding to shared embeddings.
RESULTS: We validated the proposed method on 442 benign nodules and 406 malignant nodules, extracted from the public LIDC-IDRI dataset by following a rigorous experimental setup using only nodules annotated by at least three radiologists. Our results show that the 3D Axial-Attention model achieves state-of-the-art performance on all evaluation metrics, including AUC and Accuracy.
CONCLUSIONS: The proposed model provides full 3D attention, whereby every element (i.e., pixel) in the 3D volume space attends to every other element in the nodule effectively. Thus, the 3D Axial-Attention network can be used in all layers without the need for local filters. The experimental results show the importance of full 3D attention for classifying lung nodules.
OBJECTIVE: This study aimed to evaluate the acceptance of the ChildSafe smartphone app intervention by parents/guardians.
METHODS: This study was conducted using a qualitative exploratory approach on selected participants of the ChildSafe intervention app study. A total of 27 semistructured in-depth interviews were carried out among parents or guardians who have at least one child between the age of 0 and 59 months in the area of Sungai Buloh, Selangor, between November 2017 and March 2018. Interview questions were developed from the consolidated framework for implementation research (CFIR). Interviews were recorded, transcribed verbatim, and data were thematically analyzed guided by CFIR.
RESULTS: The study revealed users' perception on usability, feasibility, and acceptability toward the ChildSafe app. Three CFIR domains were identified: intervention characteristics, inner setting, and characteristics of individuals. A total of 5 constructs were revealed under intervention characteristics: evidence strength and quality, relative advantage, adaptability, trialability, and design quality and packaging; 2 under inner setting: implementation climate and readiness for implementation; and 4 under characteristics of individuals: knowledge and beliefs about the intervention, self-efficacy, individual stage of change, and other personal attributes. In general, participants felt the app is extremely useful and effective, easy to use, and purposeful in achieving home safety assessment via reminders. The app replaces the need for participants to search for information on home safety and dangers, as the app itself was designed as a tool to assess for this specific purpose. Even at the nascent stage and despite its limitations, the app has prompted users to consider and make changes around their own home. However, future versions of the app should be expanded to make it more attractive to users as it lacks interactive feedback and additional features.
CONCLUSIONS: Parents/guardians are accepting the use of the ChildSafe app to prevent child injury at home. However, further expansion and improvements are needed to increase the acceptability of this app by parents/guardians.