Guidelines for Developing Automated Quality Control Procedures for Brain Magnetic Resonance Images Acquired in Multi-Centre Clinical Trials

Automated quality control (QC) procedures are critical for efficiently obtaining precise quantitative brain imaging-based metrics of in vivo brain pathology. This is especially important for multi-centre clinical trials of therapeutics for neurological diseases, in which brain imaging-based metrics may be used to quantify therapeutic efficacy. While there are many different types of brain imaging methods (e.g. computed tomography, magnetic resonance imaging, positron emission tomography, etc.) that have been used to quantify different aspects of in vivo pathology (e.g. presence of tumours, brain atrophy, hydrocephalus, abnormalities in blood vessels or the extravasation of blood, the depletion of receptors available for the binding of an injected substance, abnormal brain metabolism, etc.), this Chapter will focus on the automated QC procedures required to use magnetic resonance (MR) images (MRI) to yield imaging-based metrics of in vivo brain tissue pathology. Magnetic resonance imaging is a powerful non-invasive technology that can provide in vivo images sensitive to normal and pathological brain tissue. Important strengths of MR imaging include its superior grey-matter (GM)/ white-matter (WM) tissue contrast, sensitivity to WM pathology and clinical feasibility of relatively high-resolution whole-brain imaging. In conventional brain MRI, the signal intensities arise from the different relaxation characteristics of protons in water molecules present in different brain environments following radio-frequency (RF) excitation when the brain is in a magnetic field. MRI acquisition sequences vary the timing and duration of RF excitation pulses and magnetic field gradients, yielding different contrasts (termed MRI modalities) that can highlight different aspects of brain anatomy and pathology. This is illustrated in Fig. 1 using 4 conventional imaging modalities, T1-weighed (T1w) and T1w 5 min after intravenous injection of a gadolinium (Gd) contrast agent (T1w+Gd), T2-weighted (T2w), proton density weighted (PDw), and fluid attenuated inversion recovery (FLAIR) image, which were all acquired from a patient with multiple sclerosis (MS), a neurological disease that affects the brain and spinal cord. The T1w image most clearly differentiates brain GM, WM and