Left ventricle functional analysis from coronary CT angiography

Coronary CT angiography is widely used in clinical practice for the assessment of coronary artery disease. Several studies have shown that the same exam can also be used to assess left ventricle (LV) function. LV function is usually evaluated using just the data from end-systolic and end-diastolic phases even though coronary CT angiography (CTA) provides data concerning multiple cardiac phases, along the cardiac cycle. This unused wealth of data, mostly due to its complexity and the lack of proper tools, has still to be explored in order to assess if further insight is possible regarding regional LV functional analysis. Furthermore, different parameters can be computed to characterize LV function and while some are well known by clinicians others still need to be evaluated concerning their value in clinical scenarios. The work presented in this thesis covers two steps towards extended use of CTA data: LV segmentation and functional analysis. A new semi-automatic segmentation method is presented to obtain LV data for all cardiac phases available in a CTA exam and a 3D editing tool was designed to allow users to fine tune the segmentations. Regarding segmentation evaluation, a methodology is proposed in order to help choose the similarity metrics to be used to compare segmentations. This methodology allows the detection of redundant measures that can be discarded. The evaluation was performed with the help of three experienced radiographers yielding low intraand inter-observer variability. In order to allow exploring the segmented data, several parameters characterizing global and regional LV function are computed for the available cardiac phases. The data thus obtained is shown using a set of visualizations allowing synchronized visual exploration. The main purpose is to provide means for clinicians to explore the data and gather insight over their meaning, as well as their correlation with each other and with diagnosis outcomes. Finally, an interactive method is proposed to help clinicians assess myocardial perfusion by providing automatic assignment of lesions, detected by clinicians, to a myocardial segment. This new approach has obtained positive feedback from clinicians and is not only an improvement over their current assessment method but also an important first step towards systematic validation of automatic myocardial perfusion assessment measures.

Neuromuscular adaptations in football athletes with prior history of hamstring strain injury

Background: Hamstring strain injuries (HSI) are one of the most common injuries in a wide variety of running-sports, resulting in a considerable loss of competition and training time. One of the most problematic consequences regarding HSI is the recurrence rate and its non-decrease over the past decades, despite increasing evidence. Recent studies also found several maladaptations post-HSI probably due to neuromuscular inhibition and it has been proposed that these adaptations post-injury may contribute as risk factors for the injury-reinjury cycle and high recurrence rates. Furthermore it has been recently proposed not to disregard the inter-relationship between these adaptations and risk-factors post-injury in order to better understand the mechanisms of this complex injury. Objective: To determine, analyze and correlate neuromuscular adaptations in amateur football players with prior history of HSI per comparison to uninjured athletes in similar conditions. Methodology: Every participant was subjected to isokinetic concentric (60 and 240deg.sec) and eccentric (30 and 120deg.sec¯¹) testing, and peak torque, angle of peak torque and hamstrings to quadriceps (H:Q) conventional ratios were measured, myoelectrical activity of Bicep Femoris (BF) and Medial Hamstrings (MH) were also measured during isokinetic eccentric testing at both velocities and muscle activation percentages were calculated at 30, 50 and 100ms after onset of contraction. Furthermore active and passive knee extension, knee joint position sense (JPS) test, triple-hop distance (THD) test and core stability (flexors and extensors endurance, right and left side bridge test) were used and correlated. Results: Seventeen players have participated in this study: 10 athletes with prior history of HSI, composing the Hamstring injury group (HG) and 7 athletes without prior severe injuries as control group (CG). We found statistical significant differences between HG injured and uninjured sides in the BF myoelectrical activity at almost all times in both velocities and between HG injured and CG non-dominant sides at 100ms in eccentric 120deg.sec¯¹ velocity (p<.05). We found no differences in MH activity. Regarding proprioception we found differences between the HG injured and uninjured sides (p=.027). We found no differences in the rest of used tests. However, significant correlation between myoelectrical activation at 100ms in 120deg.sec¯¹ testing and JPS with initial position at 90º (r-.372; p=0.031) was found, as well as between isokinetic H:Q ratio at 240deg.sec and THD score (r=-.345; p=.045). Conclusion: We found significant differences that support previous research regarding neuromuscular adaptations and BF inhibition post-HSI. Moreover, to our knowledge, this was the first study that found correlation between these adaptations, and may open a door to new perspectives and future studies.