Novel methods and applications in Computer Aided Medical Diagnosis

计算机辅助医学诊断（mCAD）是一个快速发展的研究领域，提供了改进的诊断方法。 为临床医生提供诊断和定量图像分析解决方案，有助于提高特异性和诊断 信心mCAD系统可应用于所有成像模式，以实现准确的病变检测、治疗 计划和连续监测患者。特征提取和分析是信号分析的核心要素， mCAD中遇到的图像处理系统。成功设计这样一个系统的人机交互的关键取决于一个强大的特征提取和分类系统，这是我们的主要研究课题。我们的目标是开发一个计算机辅助医疗决策支持系统， 儿科患者的疾病识别和分类。儿科患者通常有各种常见和不常见的疾病，如代谢性脑疾病（MBD）。与成人患者人群相比，儿科患者也代表了诊断成像的独特挑战，这是由于其 更小的尺寸、解剖变异性和改变的运动。 我们提出了一种新的计算机辅助方法，以协助放射科医生在早期和准确的诊断， 儿科代谢性脑疾病据我们所知，我们的方法开创了医疗领域的新纪元。 诊断和计算机辅助支持系统， 不同的成像模式，以提高诊断准确性，其中每种模式本身可能并不总是足够的。此外，虽然我们提出的方法侧重于代谢性脑疾病的目的，本赠款申请，所产生的方法将显着有助于生物医学信号和图像分析和模式识别的进步可扩展到其他疾病。我们提出的mCAD方案在智能系统和信息技术领域也具有巨大的潜力。我们正在与多伦多的SickKids的诊断成像部门合作。硕士和博士生将提供一个独特的机会，在这个多学科的环境中合作。

Computer aided medical diagnosis (mCAD) is a rapidly developing field of investigation offering improved diagnosis and quantitative image analysis solutions for clinicians, helping to increase specificity and diagnostic confidence. mCAD systems can be applied to all imaging modalities for accurate lesion detection, treatment planning and serial monitoring of the patients. Feature extraction and analysis is a core element of signal and image processing systems encountered in mCAD. The key to successful design of such a system for human-machine interaction depends on a robust feature extraction and classification systems which is the main topics of our investigation. Our goal is to develop a computer aided medical decision support system for disease identification and classification in pediatric patients. Pediatric patients often have a variety of usual and uncommon disorders not encountered in adults such as metabolic brain diseases (MBD). Compared to the adult patient population, pediatric patients also represent unique challenges for diagnostic imaging due to their smaller size, anatomical variability and altered motion. We propose a novel computer aided methodology to assist radiologists in early and accurate diagnosis of pediatric metabolic brain diseases. To our knowledge, our approach introduces a new era in medical diagnostics and computer aided support systems by infusing image and signal information obtained from two different imaging modalities for enhanced diagnosis accuracy, where each modality by itself may not always be sufficient. Additionally, while our proposed approach focuses on metabolic brain disease for the purpose of this grant application, the resultant methodologies will significantly contribute to the advancement of biomedical signal and image analysis and pattern recognition extensible to other disorders. Our proposed mCAD scheme could also have immense potential in the intelligent systems and information technology sectors. We are collaborating with the Department of Diagnostic Imaging at SickKids in Toronto. The MSc and PhD trainees will be provided with a unique opportunity to collaborate in this multidisciplinary environment.