After the devastating 2010 flood in Pakistan, an early warning system (EWS) for river floods has been established in the Lai basin passing through the twin cities of Islamabad and Rawalpindi. Inequalities in society are amplified at the time of disasters, and EWS that are people-centred proved more effective in communicating risk and saving people. This article undertakes a gender analysis of Pakistan's EWS for each of the four pillars of people-centred EWS in order to highlight gendered and classed vulnerabilities to flood. Focus group discussions and key informant interviews were conducted with members of relevant institutions and communities in four neighbourhoods across the length of the Lai basin for understanding how gendered vulnerability impacts the acquisition of the risk messages, how congruent are the risk messages in EWS with gendered risk perception, and to what extent formal EWS enable or hinder behavioural responses to the risk messages. The EWS in the capital of Pakistan comes up short on all the criteria for a people-centred gender-sensitive EWS. Technocratic approach, lack of citizens' involvement and communication gap between the official jargonistic early warning messages and communities at risk are the major obstacles. Despite the establishment of ad hoc cells for addressing gender issues, gender is hardly operationalised and does not go beyond a token recognition

Due to the successful application of machine learning techniques in several fields, automated diagnosis system in healthcare has been increasing at a high rate. The aim of the study is to propose an automated skin cancer diagnosis and triaging model and to explore the impact of integrating the clinical features in the diagnosis and enhance the outcomes achieved by the literature study. We used an ensemble-learning framework, consisting of the EfficientNetB3 deep learning model for skin lesion analysis and Extreme Gradient Boosting (XGB) for clinical data. The study used PAD-UFES-20 data set consisting of six unbalanced categories of skin cancer. To overcome the data imbalance, we used data augmentation. Experiments were conducted using skin lesion merely and the combination of skin lesion and clinical data. We found that integration of clinical data with skin lesions enhances automated diagnosis accuracy. Moreover, the proposed model outperformed the results achieved by the previous study for the PAD-UFES-20 data set with an accuracy of 0.78, precision of 0.89, recall of 0.86, and F1 of 0.88. In conclusion, the study provides an improved automated diagnosis system to aid the healthcare professional and patients for skin cancer diagnosis and remote triaging.

No abstract

No abstract

Floods after monsoon rains are frequent disasters that affect millions of lives in Pakistan. Human lives are lost, agriculture economies are destroyed, and livestock animals, houses, fruit farms, and crops are lost which are the major livelihoods of thousands of people in Punjab. Each year there are heavy rains in the monsoon season and, due to global warming, there is the rapid melting of snow in northern glaciers; these factors subsequently cause floods. There is also loss of life due to the spread of waterborne diseases and snake bites. Flood monitoring provides early detection of a flood and the calculation of its intensity, which results in reduced human life losses and economic losses. Most casualties are caused by the lack of timely real-time, authentic information about the high-risk areas, and flood intensity, speed, and direction. Therefore, the proposed approach is centered on formal modeling and verification of safety and liveness properties of flood monitoring perceivers. Each flood perceiver has several sensors. It requires the collection of information starting from the flood perceiver, observer, and environmental forecast. This information is processed to determine the flood intensity level. We have developed a CP-Nets’ formal model and model-checked it. We have verified the safety and liveness properties of correctness by exhaustive verification of the system using model-based proof obligations (Event-B method using Rodin). Our objective in this research is to propose a correct, reliable, and efficient flood warning, monitoring, and rescue (WMR) SoS based on formal methods. We have used formal modeling and model-checking based on state-of-the-art hierarchical CP-Nets supported by exhaustive formal proof obligations of Event-B.