No abstract

Obstructive sleep apnea (OSA) and obstructive sleep apnea syndrome (OSAS) are subsets of sleep-disordered breathing. Awareness about OSA and its consequences among the general public as well as the majority of primary care physicians across India is poor. This necessitated the development of the Indian initiative on obstructive sleep apnea (INOSA) guidelines under the auspices of Department of Health Research, Ministry of Health and Family Welfare, Government of India. OSA is the occurrence of an average five or more episodes of obstructive respiratory events per hour of sleep with either sleep-related symptoms or co-morbidities or ≥15 such episodes without any sleep-related symptoms or co-morbidities. OSAS is defined as OSA associated with daytime symptoms, most often excessive sleepiness. Patients undergoing routine health check-up with snoring, daytime sleepiness, obesity, hypertension, motor vehicular accidents, and high-risk cases should undergo a comprehensive sleep evaluation. Medical examiners evaluating drivers, air pilots, railway drivers, and heavy machinery workers should be educated about OSA and should comprehensively evaluate applicants for OSA. Those suspected to have OSA on comprehensive sleep evaluation should be referred for a sleep study. Supervised overnight polysomnography is the “gold standard” for evaluation of OSA. Positive airway pressure (PAP) therapy is the mainstay of treatment of OSA. Oral appliances (OA) are indicated for use in patients with mild to moderate OSA who prefer OA to PAP, or who do not respond to PAP or who fail treatment attempts with PAP or behavioral measures. Surgical treatment is recommended in patients who have failed or are intolerant to PAP therapy.

No abstract

Obstructive sleep apnoea (OSA) and obstructive sleep apnoea syndrome (OSAS) are subsets of sleep-disordered breathing. Awareness about OSA and its consequences amongst the general public as well as the majority of primary care physcians across India is poor. This necessiated the development of the INdian initiative on Obstructive sleep apnoea (INOSA) guidelines under the auspices of Department of Health Research, Ministry of Health & Family Welfare, Government of India. OSA is the occurrence of an average five or more episodes of obstructive respiratory events per hour of sleep with either sleep related symptoms or co-morbidities or ≥ 15 such episodes without any sleep related symptoms or co-morbidities. OSAS is defined as OSA associated with daytime symptoms, most often excessive sleepiness. Patients undergoing routine health check-up with snoring, daytime sleepiness, obesity, hypertension, motor vehicular accidents and high risk cases should undergo a comprehensive sleep evaluation. Medical examiners evaluating drivers, air pilots, railway drivers and heavy machinery workers should be educated about OSA and should comprehensively evaluate applicants for OSA. Those suspected to have OSA on comprehensive sleep evaluation should be referred for a sleep study. Supervised overnight polysomnography (PSG) is the “gold standard” for evaluation of OSA. Positive airway pressure (PAP) therapy is the mainstay of treatment of OSA. Oral appliances are indicated for use in patients with mild to moderate OSA who prefer oral appliances to PAP, or who do not respond to PAP or who fail treatment attempts with PAP or behavioural measures. Surgical treatment is recommended in patients who have failed or are intolerant to PAP therapy.

No abstract

Figure 1: We propose a framework for semi-supervised training of conditional GANs, which uses much fewer labels than traditionally required. Here we train a semantic image synthesis network using our framework with just 5 labeled pairs (shown on the right), and around 29000 unpaired images. Synthesized images and corresponding input semantic maps from the test set are shown on the left. Even with just 5 labelled pairs, the network is able to synthesize high quality results, while accurately respecting the semantic layout.

No abstract

Graph Neural Networks (GNNs) exhibit excellent performance when graphs have strong homophily property, i.e. connected nodes have the same labels. However, they perform poorly on heterophilic graphs. Several approaches address the issue of heterophily by proposing models that adapt the graph by optimizing task-specific loss function using labelled data. These adaptations are made either via attention or by attenuating or enhancing various low-frequency/high-frequency signals, as needed for the task at hand. More recent approaches adapt the eigenvalues of the graph. One important interpretation of this adaptation is that these models select/weigh the eigenvectors of the graph. Based on this interpretation, we present an eigendecomposition based approach and propose EIGENNETWORK models that improve the performance of GNNs on heterophilic graphs. Performance improvement is achieved by learning flexible graph adaptation functions that modulate the eigenvalues of the graph. Regularization of these functions via parameter sharing helps to improve the performance even more. Our approach achieves up to 11% improvement in performance over the state-of-the-art methods on heterophilic graphs.Preprint. Under review.