Effect of remote ischaemic preconditioning in cardiac dysfunction and end-organ injury following cardiac surgery with cardiopulmonary bypass in children: A translational approach investigating clinical outcome and myocardial molecular biology

Congenital heart disease (CHD) is the most common birth defect, causing an important rate of morbidity and mortality. Treatment of CHD requires surgical correction in a significant percentage of cases which exposes patients to cardiac and end organ injury. Cardiac surgical procedures often require the utilisation of cardiopulmonary bypass (CPB), a system that replaces heart and lungs function by diverting circulation into an external circuit. The use of CPB can initiate potent inflammatory responses, in addition a proportion of procedures require a period of aortic cross clamp during which the heart is rendered ischaemic and is exposed to injury. High O2 concentrations are used during cardiac procedures and when circulation is re-established to the heart which had adjusted metabolically to ischaemia, further injury is caused in a process known as ischaemic reperfusion injury (IRI). Several strategies are in place in order to protect the heart during surgery, however injury is still caused, having detrimental effects in patients at short and long term. Remote ischaemic preconditioning (RIPC) is a technique proposed as a potential cardioprotective measure. It consists of exposing a remote tissue bed to brief episodes of ischaemia prior to surgery in order to activate protective pathways that would act during CPB, ischaemia and reperfusion. This study aimed to assess RIPC in paediatric patients requiring CHD surgical correction with a translational approach, integrating clinical outcome, marker analysis, cardiac function parameters and molecular mechanisms within the cardiac tissue. A prospective, single blinded, randomized, controlled trial was conducted applying a RIPC protocol to randomised patients through episodes of limb ischaemia on the day before surgery which was repeated right before the surgery started, after anaesthesia induction. Blood samples were obtained before surgery and at three post-operative time points from venous lines, additional pre and post-bypass blood samples were obtained from the right atrium. Myocardial tissue was resected during the ischaemic period of surgery. Echocardiographic images were obtained before the surgery started after anaesthetic induction and the day after surgery, images were stored for later off line analysis. PICU surveillance data was collected including ventilation parameters, inotrope use, standard laboratory analysis and six hourly blood gas analysis. Pre and post-operative quantitation of markers in blood specimens included cardiac troponin I (cTnI) and B-type natriuretic peptide (BNP), inflammatory mediators including interleukins IL-6, IL-8, IL-10, tumour necrosis factor (TNF-α), and the adhesion molecules ICAM-1 and VCAM-1; the renal marker Cystatin C and the cardiovascular markers asymmetric dymethylarginine (ADMA) and symmetric dymethylarginine (SDMA). Nitric oxide (NO) metabolites and cyclic guanosine monophosphate (cGMP) were measured before and after bypass. Myocardial tissue was processed at baseline and after incubation at hyperoxic concentration during four hours in order to mimic surgical conditions. Expression of genes involved in IRI and RIPC pathways was analysed including heat shock proteins (HSPs), toll like receptors (TLRs), transcription factors nuclear factor κ-B (NF- κ-B) and hypoxia inducible factor 1 (HIF-1). The participation of hydrogen sulfide enzymatic genes, apelin and its receptor were explored. There was no significant difference according to group allocation in any of the echocardiographic parameters. There was a tendency for higher cTnI values and inotropic score in control patients post-operatively, however this was not statistically significant. BNP presented no significant difference according to group allocation. Inflammatory parameters tended to be higher in the control group, however only TNF- α was significantly higher. There was no difference in levels of Cystatin C, NO metabolites, cGMP, ADMA or SDMA. RIPC patients required shorter PICU stay, all other clinical and laboratory analysis presented no difference related to the intervention. Gene expression analysis revealed interesting patterns before and after incubation. HSP-60 presented a lower expression at baseline in tissue corresponding to RIPC patients, no other differences were found. This study provided with valuable descriptive information on previously known and newly explored parameters in the study population. Demographic characteristics and the presence of cyanosis before surgery influenced patterns of activity in several parameters, numerous indicators were linked to the degree of injury suffered by the myocardium. RIPC did not reduce markers of cardiac injury or improved echocardiographic parameters and it did not have an effect on end organ function; some effects were seen in inflammatory responses and gene expression analysis. Nevertheless, an important clinical outcome indicator, PICU length of stay was reduced suggesting benefit from the intervention. Larger studies with more statistical power could determine if the tendency of lower injury and inflammatory markers linked to RIPC is real. The present results mostly support findings of larger multicentre trials which have reported no cardiac benefit from RIPC in paediatric cardiac surgery.

The role of context in image annotation and recommendation

With the rise of smart phones, lifelogging devices (e.g. Google Glass) and popularity of image sharing websites (e.g. Flickr), users are capturing and sharing every aspect of their life online producing a wealth of visual content. Of these uploaded images, the majority are poorly annotated or exist in complete semantic isolation making the process of building retrieval systems difficult as one must firstly understand the meaning of an image in order to retrieve it. To alleviate this problem, many image sharing websites offer manual annotation tools which allow the user to “tag” their photos, however, these techniques are laborious and as a result have been poorly adopted; Sigurbjörnsson and van Zwol (2008) showed that 64% of images uploaded to Flickr are annotated with < 4 tags. Due to this, an entire body of research has focused on the automatic annotation of images (Hanbury, 2008; Smeulders et al., 2000; Zhang et al., 2012a) where one attempts to bridge the semantic gap between an image’s appearance and meaning e.g. the objects present. Despite two decades of research the semantic gap still largely exists and as a result automatic annotation models often offer unsatisfactory performance for industrial implementation. Further, these techniques can only annotate what they see, thus ignoring the “bigger picture” surrounding an image (e.g. its location, the event, the people present etc). Much work has therefore focused on building photo tag recommendation (PTR) methods which aid the user in the annotation process by suggesting tags related to those already present. These works have mainly focused on computing relationships between tags based on historical images e.g. that NY and timessquare co-exist in many images and are therefore highly correlated. However, tags are inherently noisy, sparse and ill-defined often resulting in poor PTR accuracy e.g. does NY refer to New York or New Year? This thesis proposes the exploitation of an image’s context which, unlike textual evidences, is always present, in order to alleviate this ambiguity in the tag recommendation process. Specifically we exploit the “what, who, where, when and how” of the image capture process in order to complement textual evidences in various photo tag recommendation and retrieval scenarios. In part II, we combine text, content-based (e.g. # of faces present) and contextual (e.g. day-of-the-week taken) signals for tag recommendation purposes, achieving up to a 75% improvement to precision@5 in comparison to a text-only TF-IDF baseline. We then consider external knowledge sources (i.e. Wikipedia & Twitter) as an alternative to (slower moving) Flickr in order to build recommendation models on, showing that similar accuracy could be achieved on these faster moving, yet entirely textual, datasets. In part II, we also highlight the merits of diversifying tag recommendation lists before discussing at length various problems with existing automatic image annotation and photo tag recommendation evaluation collections. In part III, we propose three new image retrieval scenarios, namely “visual event summarisation”, “image popularity prediction” and “lifelog summarisation”. In the first scenario, we attempt to produce a rank of relevant and diverse images for various news events by (i) removing irrelevant images such memes and visual duplicates (ii) before semantically clustering images based on the tweets in which they were originally posted. Using this approach, we were able to achieve over 50% precision for images in the top 5 ranks. In the second retrieval scenario, we show that by combining contextual and content-based features from images, we are able to predict if it will become “popular” (or not) with 74% accuracy, using an SVM classifier. Finally, in chapter 9 we employ blur detection and perceptual-hash clustering in order to remove noisy images from lifelogs, before combining visual and geo-temporal signals in order to capture a user’s “key moments” within their day. We believe that the results of this thesis show an important step towards building effective image retrieval models when there lacks sufficient textual content (i.e. a cold start).