Search

This paper presents comparative evaluation of an application of deep convolutional neural networks (dCNNs) to age invariant face recognition. To this end, we use four distinct dCNN models, the AlexNet, VGGNet, GoogLeNet and ResNet. We assess their performance to recognize face images across aging variations, firstly by fine-tuning the models and secondly using them as face feature extractor. We also suggest a novel synthesized aging augmentation technique suitable for age-invariant face recognition using dCNNs. The face recognition experiments are conducted on three challenging FG-NET, MORPH and LAG aging datasets, and results are benchmarked with a simple CNN. The comparative study allows us to answer (i) when and why transfer learning or feature extraction strategies are useful in age-invariant face recognition scenarios, (ii) the potential of aging synthesized augmentation to increase accuracy and (iii) the choice of appropriate feature normalization and distance metrics to be used with deeply learned features. The extensive experiments, and valuable insights presented in this study can be extended to the design of effective age-invariant face recognition algorithms. [ABSTRACT FROM AUTHOR]

Objectives: To determine the association between low birth weight babies and environmental tobacco smoke (ETS) exposure during pregnancy and the factors related to it. Study Design: Case Control study Setting: Obstetrical and Gynecological Units of a Sir Ganga Ram Hospital, Lahore. Period: March 2016 to August 2017. Material & Methods: Babies born to mothers' aged between 20 to 35 years in all four Obstetrical and Gynecological units of a Sir Ganga Ram hospital, Lahore, were included. 150 low birth weight babies (LBW) born at term, fulfilling the exclusion and inclusion criteria were taken as cases and 150 normal weight babies born on same day were taken as controls. Mothers were interviewed on a semi structured and pretested questionnaire covering all variables using non- probability purposive sampling technique and the data was analyzed on SPSS Version 20.0. The difference of means was tested using independent t-test whereas difference of proportions was analyzed by using Pearson Chi -Square test. Odds Ratio of >2 was taken as significant. P value <0.05 was considered statistically significant. In order to adjust the effect of confounder data was stratified. Results: 110/150(73.3%) cases and 48/150(32%) controls were exposed to ETS with odds of low birth weight being 5.84 times higher among exposed mothers. Odds of having low birth weight was 3.55 times for ETS exposure at home than those not having home exposure. Ninety Six cases (64%) and 70(46.7%) controls had no knowledge about passive smoking. Only one third of all respondents had knowledge about Second Hand Smoke(SHS) while 110 exposed to ETS, only 09 had adequate knowledge about passive smoking with a significant difference in the knowledge among mothers of the cases and controls (p value 0.004, T score -2889). Conclusion: Mothers exposed to ETS had 5.84 odds of having LBW babies than unexposed, endorsing the association between the two particular when at home. Illiteracy, illiterate smoker husband and poor knowledge of mother about SHS and its hazards further escalated the situation. More researches are, therefore, required to evaluate the effects of passive smoking exposure in a low income country like Pakistan and steps should be taken to avoid it especially during pregnancy. [ABSTRACT FROM AUTHOR]

Background: Delay in diagnosis of Tuberculosis (TB) and initiation of antitubercular treatment (ATT) contributes to more severe disease manifestations in the individual and higher disease transmission in the community. Objective: To find out the delays in diagnosis and treatment of TB patients and to describe determinants related to these delays. Study Design: Cross-sectional descriptive study. Setting: TB Directly Observed Treatment Short course (DOTS) Center of Jinnah Hospital, Lahore. Period: July to September 2013. Methods: 373 tuberculosis patients attending TB Directly Observed Treatment Short course (DOTS) Center of Jinnah Hospital, Lahore were included using simple random sampling. Delay was then categorized into low delay and high delay depending upon the median of total delay which was of 97 days. Results: It was concluded that the total median diagnostic and treatment delay related to both patients as well as health system was 97 days. Results revealed that about 55.7% patients had high total delay (delay>median) and 44.3% patients had low total delay. Among those patients with high delay, 58.7% were aged above 35 years, 81.4% were females, 96.4% were illiterate, 54.8%were living in rural areas, 76.7% travelled greater than 5 km to reach the nearest health facility and 56.9% patients had more than one health seeking encounters with health care professionals before initial diagnosis. Conclusion: Current study, therefore, highlights the delays in diagnosis, treatment and the determinants of delay showing health system related diagnostic and treatment delay being the main contributor to the total delay. More than half of the patients with TB showed a delay in initiation of treatment. Efforts should be made to minimize health system related delays. Local private practitioners should also be taken on board in combating tuberculosis. [ABSTRACT FROM AUTHOR]

Tuberculosis is one of the major public health problems in Pakistan. Pakistan ranks fifth amongst TB high-burden countries worldwide and accounts for 61% of the TB burden in the WHO Eastern Mediterranean Region. Objectives: To describe knowledge, attitudes, and perceptions among the patients of pulmonary tuberculosis about their disease attending outpatient department at independent university hospital Faisalabad. Study Design: A descriptive cross-sectional study. Period: August to September 2014. Setting: Out patients department at Independent University Hospital Faisalabad. Methods: A total of 197 patients from OPD. Systematic random sampling technique. After an informed consent a pretested questionnaire were used to collect the information from the patient. SPSS-17 were used to calculate the frequencies and percentages. Results: Out of 197 respondents were 55.3% (109) male, 44.7% (88) female and 63% (125) were earning below 10000 rupees per month. As for as knowledge is concerned 50% of the patients viewed that TB is transmitted through contact with the patients having tuberculosis, 83%labelled cough of long duration as commonest sign, 60% viewed that it can be prevented by avoiding contact with patients, 57% labeled healthcare staff as their source of information, 65.9% perceived TB as an infectious disease, 55% stress as cause of disease, 85% as not curable disease. Conclusions: Poor knowledge and false perceptions are the main constrain in the control of TB in Pakistan. [ABSTRACT FROM AUTHOR]