An automated system for detecting and measuring nailfold capillaries.

Nailfold capillaroscopy is an established qualitative technique in the assessment of patients displaying Raynaud's phenomenon. We describe a fully automated system for extracting quantitative biomarkers from capillaroscopy images, using a layered machine learning approach. On an unseen set of 455 images, the system detects and locates individual capillaries as well as human experts, and makes measurements of vessel morphology that reveal statistically significant differences between patients with (relatively benign) primary Raynaud's phenomenon, and those with potentially life-threatening systemic sclerosis.

Movement variability in stroke patients and controls performing two upper limb functional tasks: a new assessment methodology.

BACKGROUND: In the evaluation of upper limb impairment post stroke there remains a gap between detailed kinematic analyses with expensive motion capturing systems and common clinical assessment tests. In particular, although many clinical tests evaluate the performance of functional tasks, metrics to characterise upper limb kinematics are generally not applicable to such tasks and very limited in scope. This paper reports on a novel, user-friendly methodology that allows for the assessment of both signal magnitude and timing variability in upper limb movement trajectories during functional task performance. In order to demonstrate the technique, we report on a study in which the variability in timing and signal magnitude of data collected during the performance of two functional tasks is compared between a group of subjects with stroke and a group of individually matched control subjects. METHODS: We employ dynamic time warping for curve registration to quantify two aspects of movement variability: 1) variability of the timing of the accelerometer signals' characteristics and 2) variability of the signals' magnitude. Six stroke patients and six matched controls performed several trials of a unilateral ('drinking') and a bilateral ('moving a plate') functional task on two different days, approximately 1 month apart. Group differences for the two variability metrics were investigated on both days. RESULTS: For 'drinking from a glass' significant group differences were obtained on both days for the timing variability of the acceleration signals' characteristics (p = 0.002 and p = 0.008 for test and retest, respectively); all stroke patients showed increased signal timing variability as compared to their corresponding control subject. 'Moving a plate' provided less distinct group differences. CONCLUSION: This initial application establishes that movement variability metrics, as determined by our methodology, appear different in stroke patients as compared to matched controls during unilateral task performance ('drinking'). Use of a user-friendly, inexpensive accelerometer makes this methodology feasible for routine clinical evaluations. We are encouraged to perform larger studies to further investigate the metrics' usefulness when quantifying levels of impairment.

An instance-based algorithm with auxiliary similarity information for the estimation of gait kinematics from wearable sensors.

Wearable human movement measurement systems are increasingly popular as a means of capturing human movement data in real-world situations. Previous work has attempted to estimate segment kinematics during walking from foot acceleration and angular velocity data. In this paper, we propose a novel neural network [GRNN with Auxiliary Similarity Information (GASI)] that estimates joint kinematics by taking account of proximity and gait trajectory slope information through adaptive weighting. Furthermore, multiple kernel bandwidth parameters are used that can adapt to the local data density. To demonstrate the value of the GASI algorithm, hip, knee, and ankle joint motions are estimated from acceleration and angular velocity data for the foot and shank, collected using commercially available wearable sensors. Reference hip, knee, and ankle kinematic data were obtained using externally mounted reflective markers and infrared cameras for subjects while they walked at different speeds. The results provide further evidence that a neural net approach to the estimation of joint kinematics is feasible and shows promise, but other practical issues must be addressed before this approach is mature enough for clinical implementation. Furthermore, they demonstrate the utility of the new GASI algorithm for making estimates from continuous periodic data that include noise and a significant level of variability.

Artificial neural network prediction using accelerometers to control upper limb FES during reaching and grasping following stroke.

This work investigates arm acceleration as a control signal for functional electrical stimulation (FES) of the upper limb during reaching and grasping. We segment the reach and grasp motion into phases and present an artificial neural network (ANN) approach that estimates the phase of the reaching cycle from accelerometer signals. We then select the stimulator command that maximizes successful triggering without unnecessary risk to the patient's safety. Our results suggest that the algorithm successfully generalizes between sessions and patients but is less successful at generalizing between different motions.