Maciej Plocharski

Classification of Alzheimer's disease from MRI using sulcal morphology

Alzheimer’s disease (AD), an age-related progressive neurodegenerative disorder, is the most common cause of dementia. It is characterised by abnormal neuroanatomical changes in the brain, some of which can be difficult to distinguish from the alterations caused by normal aging. Sulcal morphology is affected by AD atrophy, indicates significant differences between cognitively normal (CN) and AD subjects, and proves to be a potential AD biomarker. 210 subjects (100 CN, 110 AD) were acquired from the ADNI database. 120 sulci were extracted per subject using BrainVISA sulcal identification pipeline. Mean curvature, surface area and volume were calculated for each sulcus, parameterized by a 3D mesh, and used as AD/CN classification features. 184 subjects were correctly classified (AD=98, CN=86), producing an accuracy of 88%, sensitivity of 89%, specificity of 86%, based on 33 features. Results indicate that sulcal morphology, when based on specific features, could be a valuable AD biomarker.

DeepQSM - Using Deep Learning to Solve the Dipole Inversion for MRI Susceptibility Mapping

Quantitative susceptibility mapping (QSM) aims to extract the magnetic susceptibility of tissue from magnetic resonance imaging (MRI) phase measurements. The mapping of magnetic susceptibility in vivo has gained broad interest in several fields of science and medicine because it yields relevant information on biological tissue properties, predominantly myelin, iron and calcium. Thereby, QSM can also reveal pathological changes of these key components in devastating diseases such as Parkinson’s disease, Multiple Sclerosis, or hepatic iron overload. As QSM requires the solution of an ill-posed field-to-source-inversion, current techniques utilize manual optimization of regularization parameters to balance between smoothing, artifacts and quantification accuracy. We trained a fully convolutional deep neural network - DeepQSM - to invert the magnetic dipole kernel convolution. This network is capable of solving the ill-posed field-to-source inversion on real-world in vivo MRI phase data without the need for manual parameter tuning, which proves that this network has generalized the underlying mathematical principle of the dipole inversion. We demonstrate that DeepQSM’s susceptibility maps enable identification of deep brain substructures that are not visible in MRI phase data and provide information on their respective magnetic tissue properties. We illustrate DeepQSM’s clinical relevance in a patient with multiple sclerosis showing its sensitivity to white matter lesions. In summary, DeepQSM can be used to determine the composition of myelin sheets of nerve fibers in the brain, and to assess quantitative information on iron homeostasis and its dysregulation, and will subsequently contribute to a better understanding of these biological processes in health and disease.

Cervical flexion and extension includes anti-directional cervical joint motion in healthy adults

BACKGROUND CONTEXT Anti-directional cervical joint motion has previously been demonstrated. However, quantitative studies of anti-directional and pro-directional cervical flexion and extension motions have not been published. PURPOSE This study aimed for a quantitative assessment of directional and anti-directional cervical joint motion in healthy subjects. STUDY DESIGN An observational study was carried out. PATIENTS SAMPLE Eighteen healthy subjects comprised the study sample. OUTCOME MEASURES Anti-directional and pro-directional cervical flexion and extension motion from each cervical joint in degrees were the outcome measures. METHODS Fluoroscopy videos of cervical flexion and extension motions (from neutral to end-range) were acquired from 18 healthy subjects. The videos were divided into 10% epochs of C0/C7 range of motion (ROM). The pro-directional and anti-directional motions in each 10% epoch were extracted, and the ratios of anti-directional motions with respect to the pro-directional motions (0%=no anti-directional movement) were calculated for joints and 10% epochs. RESULTS The flexion and extension ROM for C0/C7 were 51.9°±9.3° and 57.2°±12.2°. The anti-directional motions of flexion and extension ROM constituted 42.8%±9.7% and 41.2%±8.2% of the respective pro-directional movements. For flexion, the first three joints (C0/C1, C1/C2, C2/C3) demonstrated larger ratios compared with the last three joints (C4/C5, C5/C6, C6/C7) (p<.03). For extension, C1/C2 and C2/C3 ratios were larger compared with C0/C1, C4/C5, and C5/C6 (p<.03). Comparisons between flexion and extension motions showed larger C0/C1 ratio but smaller C5/C6 and C6/C7 ratios in extension (p<.05). CONCLUSIONS This is the first report of quantified anti-directional cervical flexion and extension motion. The anti-directional motion is approximately 40% of the pro-directional motion. The results document that large proportions of anti-directional cervical flexion and extension motions were normal.

Prediction of Alzheimer’s Disease in Mild Cognitive Impairment Using Sulcal Morphology and Cortical Thickness

Mild cognitive impairment (MCI) is an intermediate condition between healthy ageing and dementia. The amnestic MCI is often a high risk factor for subsequent Alzheimer’s disease (AD) conversion. Some MCI patients never develop AD (MCI non-converters, or MCInc), but some do progress to AD (MCI converters, or MCIc). The purpose of this study was to predict future AD-conversion in patients with MCI using machine learning with sulcal morphology and cortical thickness measures as classification features. 32 sulci per subject were extracted from 1.5T T1-weighted ADNI database MRI scans of 90 MCIc and 104 MCInc subjects. We computed sulcal morphology features and cortical thickness measurements for support vector machine classification to identify structural patterns distinguishing future AD conversions. The linear kernel classifier trained with these features was able to predict 87.0% of MCI subjects as future converters, (89.7% sensitivity, 84.4% specificity, 0.94 AUC), using 10-fold cross-validation. These results using sulcal and cortical features are superior to the state-of-the-art methods. The most discriminating predictive features were observed in the temporal and frontal lobes in the left hemispheres, and in the entorhinal cortices, which is consistent with literature. However, we also observed structural changes in the cingulate and calcarine cortices, suggesting that the limbic and occipital lobe atrophy may be linked to AD conversion.

Motion analysis of the cervical spine during extension and flexion: Reliability of the vertebral marking procedure

Cervical spine motion analysis using videofluoroscopy is currently a technique without a gold standard. We demonstrate the reliability of a rigid and reliable analysis methodology for cervical motion using videofluoroscopic images, representing the entire range of motion during flexion and extension, from the neutral position to the end-range in the sagittal plane. Two researchers with radiography and vertebral marking expertise, and two inexperienced researchers with 10 hours of training manually marked anatomical structures on fluoroscopic images in a procedure designed to control for vertebral rotation around the mid-plane axis. The average marking error across examiners and images was -0.12∘ (standard deviation: 0.88°), and the intraexaminer error ranged from -1.00∘ to 1.61° (standard deviation range: 0.27°-1.19°). Our method demonstrated lower errors compared to the higher resolution X-ray studies, and proved that vertebral marking can be performed by persons with no experience in radiographic image analysis.

Repeatability of cervical joint flexion and extension within and between days

Objective: The purpose of this study was to investigate within‐ and between‐day repeatability of free and unrestricted healthy cervical flexion and extension motion when assessing dynamic cervical spine motion. Methods: Fluoroscopy videos of 2 repeated cervical flexion and 2 repeated extension motions were examined for within‐day repeatability (20‐second interval) for 18 participants (6 females) and between‐day repeatability (1‐week interval) for 15 participants (6 females). The dynamic cervical motions were free and unrestricted from neutral to end range. The flexion videos and extension videos were evenly divided into 10% epochs of the C0‐to‐C7 range of motion. Within‐day and between‐day repeatability of joint motion angles (all 7 joints and epochs, respectively) was tested in a repeated‐measures analysis of variance. Joint motion angle differences between repetitions were calculated for each epoch and joint (7 joints), and these joint motion angle differences between within‐day and between‐day repetitions were tested in mixed‐model analysis of variance. Results: For all joints and epochs, respectively, no significant differences were found in joint motion angle between within‐day or between‐day repetitions. There were no significant effects of joint motion angle differences between within‐day and between‐day repetitions. The average within‐day joint motion angle differences across all joints and epochs were 0.00° ± 2.98° and 0.00° ± 3.05° for flexion and extension, respectively. The average between‐day joint motion angle differences were 0.02° ± 2.56° and 0.05° ± 2.40° for flexion and extension, respectively. Conclusions: This is the first study to report the within‐day and between‐day joint motion angle differences of repeated cervical flexion and extension. This study supports the idea that cervical joints repeat their motion accurately.

Semi-automatic method for intervertebral kinematics measurement in the cervical spine

Cervical spine injuries, such as whiplash or disk herniation, are a worldwide health problem. Digital videofluoroscopic is often used to examine spine movement by means of manual identification and marking of vertebral landmarks, which is a complicated and time-consuming process. The aim of our study was to develop a fast, semi-automatic cervical vertebrae tracking method to accurately calculate the inter vertebral rotation in C1–C7 vertebrae. We manually defined templates for each cervical vertebra, so that these templates would be automatically tracked throughout neck movement. Subjects performed extension and flexion in the sagittal plane, which was recorded with digital videofluoroscopy. We implemented cross-correlation and Kalman filters for template tracking, and validated our method by comparing our results with a manual method, where a trained clinician manually marked the vertebrae. Our method provided higher intraobserver repeatability for C2/C3 to C6/C7 segments. Accordingly, the intraobserver repeatability was also comparable to other methods developed to track the lumbar vertebrae.

Building a high-resolution in vivo minimum deformation average model of the human hippocampus

Objective Minimum deformation averaging (MDA) procedures exploit the information contained in inter-individual variations to generate high-resolution, high-contrast models through iterative model building. However, MDA models built from different image contrasts reside in disparate spaces and their complementary information cannot be utilized easily. The aim of this work was to develop an algorithm for the non-linear alignment of two MDA models with different contrasts to create a high-resolution in vivo model of the human hippocampus with a spatial resolution of 300 μm. Methods A Turbo Spin Echo MDA model covering the hippocampus was contrast matched to a whole-brain MP2RAGE MDA model and aligned using cross-correlation and non-linear transformation. The contrast matching algorithm followed a global voxel location-based approach to estimate the relationship between intensity values of the two models. The performance of the algorithm was evaluated by comparing it to a non-linear registration obtained using mutual information without contrast matching. The complimentary information from both contrasts was then utilized in an automated hippocampal subfield segmentation pipeline. Results The contrast of the Turbo Spin Echo MDA model could successfully be matched to the MP2RAGE MDA model. Registration using cross correlation provided more accurate alignment of the models compared to a mutual information based approach. The segmentation using ASHS resulted in hippocampal subfield delineations that resembled the tissue boundaries observed in the Turbo Spin Echo MDA model. Conclusion The developed contrast matching algorithm facilitated the creation of a high-resolution multi-modal in vivo MDA model of the human hippocampus. This model can be used to improve algorithms for hippocampal subfield segmentation and could potentially support the early detection of neurodegenerative diseases.

Mechanisms contributing to reduced knee stiffness during movement

The ability to modulate the mechanical properties of our limbs contributes to our ability to interact with the physical world in a consistent and predictable manner. An individual joint’s contributions to whole limb mechanics can be quantified by its joint impedance, which characterizes the torque generated about a joint in response to external perturbations of position. A number of studies have estimated joint impedance during movement and have shown that it can be much lower than it is during posture. However, the mechanisms contributing to these differences remain unknown partly because conditions known to affect impedance, including muscle activation and joint angles, have not been carefully controlled across studies. The goal of this study was to contrast knee impedance during continuous volitional movements with that during maintained postures spanning a similar range of joint angles and muscle activations and to explore physiological mechanisms likely to contribute to the observed differences. We found that knee impedance was substantially lower during movement than during matched postural tasks, even for matched muscle activations. At times, the impedance during movement was even lower than that measured during isometric tasks with no volitional muscle activity. These decreases in impedance could be attributed in part to reduced stretch reflexes during movement and to an effect of movement itself on reducing knee impedance.