Search

Today, there is a visible reduction in productivity of common crops like rice and wheat in agricultural sector. At the same time, millet crops have emerged as an alternative to counter global food security issues because of their adaptability and nutritional value that can combat malnutrition. Also, because of its low moisture requirement and tolerance to extreme climatic conditions, it is perceived as a stable cereal. But its cultivation is affected by diseases like rust and blast, thereby causing harm to the farming economy. Conventional operational methods of disease detection require regular manual intervention which are also costly. The correctness of expert suggestions is also under scrutiny. Thus, a sustainable, robust and low cost modern approach for millet crop monitoring and disease detection is required. This research aims to develop a smart and sustainable framework by integrating Internet of things (IoT) and deep learning (DL). In the presented framework, a sensory module based automated crop health data gathering system with an improved deep learning-assisted intelligent disease recognition model is developed for the millet crop. Sensors collect data from millet farmland transfer the crop data readings to the cloud server for storage and Raspberry Pi for further detection. Any abnormality in reading leads to an alert notification communicated to the farmer. A hybrid predictive model, Customized Convolutional neural network (Customized-CNN) model works with the Raspberry Pi to predict the presence of blast and rust disease symptoms in millet. The proposed sustainable model is implemented, and it generates productive outcomes. The recorded accuracy, precision, recall, and f-score values with the customized CNN model were 98.8%, 98.2%, 97.4%, and 97.7%, respectively. The training and testing delays were only 67 s and 88 s, respectively. A sample of yearly sensor readings was generated to show the accuracy and reliability of the model’s data collection. Also, the model proved to be scalable, as it gave noteworthy performance with diverse crops. The evaluation shows the reliability of the model, and it can be used by farming societies to enhance millet cereal yield in a cost-effective way.

BACKGROUND: Attention-deficit/hyperactivity disorder (ADHD) is a common neurodevelopmental condition in children. Although ADHD is treatable, barriers remain to engagement in treatment, especially among socioeconomically disadvantaged and racial and ethnic minority families. Our goal was to examine the process by which families engage in ADHD treatment and to identify targets for an intervention to improve engagement in care. METHODS: We conducted in-depth semistructured qualitative interviews with 41 parents of diverse youth aged 3 to 17 years old in treatment of ADHD at an urban safety net hospital. Parents were asked about their journey through diagnosis and treatment, community attitudes about ADHD, and other factors influencing treatment access and decision-making. Transcripts were analyzed by using thematic analysis. RESULTS: Of children with ADHD, 69.2% were male, 57.7% were Black or African American, and 38.5% were of Hispanic, Latino, or Spanish origin. Parents were 92.7% female, were 75.6% English speaking, and had a median income of $20 000. Parents described 6 stages to the process of engaging in care for their childs ADHD, which unfolded like a developmental process: (1) normalization and hesitation, (2) fear and stigmatization, (3) action and advocacy, (4) communication and navigation, (5) care and validation, and (6) preparation and transition. Barriers often occurred at points of stage mismatch between parents and providers and/or systems. Difficulty resolving an earlier stage interfered with the progression through subsequent stages. CONCLUSIONS: The 6 stages framework could be used to develop new strategies to measure engagement and to design family-centered interventions to facilitate engagement in ADHD treatment.