Rule-based deep learning method for prognosis of neonatal hypoxic-ischemic encephalopathy by using susceptibility weighted image analysis.

OBJECTIVE: Susceptibility weighted imaging (SWI) of neonatal hypoxic-ischemic brain injury can provide assistance in the prognosis of neonatal hypoxic-ischemic encephalopathy (HIE). We propose a convolutional neural network model to classify SWI images with HIE. MATERIALS AND METHODS: Due to the lack of a large dataset, transfer learning method with fine-tuning a pre-trained ResNet 50 is introduced. We randomly select 11 datasets from patients with normal neurology outcomes (n = 31) and patients with abnormal neurology outcomes (n = 11) at 24 months of age to avoid bias in classification due to any imbalance in the data. RESULTS: We develop a rule-based system to improve the classification performance, with an accuracy of 0.93 ± 0.09. We also compute heatmaps produced by the Grad-CAM technique to analyze which areas of SWI images contributed more to the classification patients with abnormal neurology outcome. CONCLUSION: Such regions that are important in the classification accuracy can interpret the relationship between the brain regions affected by hypoxic-ischemic and neurodevelopmental outcomes of infants with HIE at the age of 2 years.

Machine learning as an adjunct to expert observation in classification of radiographic knee osteoarthritis: findings from the Hertfordshire Cohort Study.

BACKGROUND: Osteoarthritis is the most prevalent type of arthritis. Many approaches exist for characterising radiographic knee OA, including machine learning (ML). AIMS: To examine Kellgren and Lawrence (K&L) scores from ML and expert observation, minimum joint space and osteophyte in relation to pain and function. METHODS: Participants from the Hertfordshire Cohort Study, comprising individuals born in Hertfordshire from 1931 to 1939, were analysed. Radiographs were assessed by clinicians and ML (convolutional neural networks) for K&L scoring. Medial minimum joint space and osteophyte area were ascertained using the knee OA computer-aided diagnosis (KOACAD) program. The Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) was administered. Receiver operating characteristic analysis was implemented for minimum joint space, osteophyte, and observer- and ML-derived K&L scores in relation to pain (WOMAC pain score > 0) and impaired function (WOMAC function score > 0). RESULTS: 359 participants (aged 71-80) were analysed. Among both sexes, discriminative capacity regarding pain and function was fairly high for observer-derived K&L scores [area under curve (AUC): 0.65 (95% CI 0.57, 0.72) to 0.70 (0.63, 0.77)]; results were similar among women for ML-derived K&L scores. Discriminative capacity was moderate among men for minimum joint space in relation to pain [0.60 (0.51, 0.67)] and function [0.62 (0.54, 0.69)]. AUC < 0.60 for other sex-specific associations. DISCUSSION: Observer-derived K&L scores had higher discriminative capacity regarding pain and function compared to minimum joint space and osteophyte. Among women, discriminative capacity was similar for observer- and ML-derived K&L scores. CONCLUSION: ML as an adjunct to expert observation for K&L scoring may be beneficial due to the efficiency and objectivity of ML.

Automatic veins analysis of susceptibility weighted image in hypoxic-ischaemic encephalopathy.

BACKGROUND AND OBJECTIVE: The purpose of this study is to evaluate venous vascular structure and distribution as prognostic indicators of developmental outcomes for infants with neonatal hypoxic-ischaemic encephalopathy (HIE) by detecting and analysing ridges representing vessels on susceptibility-weighted magnetic resonance images (SWIs). METHODS: Forty-two infants with neonatal HIE underwent SWI in the neonatal period and neurodevelopmental assessment at age 2 years. Normalised histograms of the width, intensity, length and Hessian eigenvalues extracted from the ridge analysis of each patient's SWI are applied as feature vectors to feed into supervised classifiers such as the kNN and random forest (RF) classifiers to predict their neurodevelopmental outcomes. Here we also propose a supervised classifier for automatic prognosis of automated detection of SWI signs of HIE. Our classifier proposed in this paper demonstrates a superior performance in HIE prognosis for the datasets associated with cognitive and motor outcomes and it also enables to determination of brain regions which have been affected by hypoxia-ischaemia by extracting appropriate features from SWI images. RESULTS: The feature vectors containing width, intensity, length, and eigenvalue show a promising classification accuracy of 78.67% ± 2.58Linear regression, polynomial regression, and support vector regression (SVR) models predicted outcomes and the lower mean relative errors (MRE) for motor and cognitive outcomes are 0.088 ± 0.073 and 0.101 ± 0.11 respectively. CONCLUSION: The features derived from the vascular ridges improve the prognostic value of SWI in HIE. Our findings suggest that it is possible to predict neurological, motor, and cognitive outcomes by numerical analysis of neonatal SW images and to identify brain regions on SWI affected by hypoxia-ischaemia.

Machine learning applied to HR-pQCT images improves fracture discrimination provided by DXA and clinical risk factors.

BACKGROUND: Traditional analysis of High Resolution peripheral Quantitative Computed Tomography (HR-pQCT) images results in a multitude of cortical and trabecular parameters which would be potentially cumbersome to interpret for clinicians compared to user-friendly tools utilising clinical parameters. A computer vision approach (by which the entire scan is 'read' by a computer algorithm) to ascertain fracture risk, would be far simpler. We therefore investigated whether a computer vision and machine learning technique could improve upon selected clinical parameters in assessing fracture risk. METHODS: Participants of the Hertfordshire Cohort Study (HCS) attended research visits at which height and weight were measured; fracture history was determined via self-report and vertebral fracture assessment. Bone microarchitecture was assessed via HR-pQCT scans of the non-dominant distal tibia (Scanco XtremeCT), and bone mineral density measurement and lateral vertebral assessment were performed using dual-energy X-ray absorptiometry (DXA) (Lunar Prodigy Advanced). Images were cropped, pre-processed and texture analysis was performed using a three-dimensional local binary pattern method. These image data, together with age, sex, height, weight, BMI, dietary calcium and femoral neck BMD, were used in a random-forest classification algorithm. Receiver operating characteristic (ROC) analysis was used to compare fracture risk identification methods. RESULTS: Overall, 180 males and 165 females were included in this study with a mean age of approximately 76 years and 97 (28 %) participants had sustained a previous fracture. Using clinical risk factors alone resulted in an area under the curve (AUC) of 0.70 (95 % CI: 0.56-0.84), which improved to 0.71 (0.57-0.85) with the addition of DXA-measured BMD. The addition of HR-pQCT image data to the machine learning classifier with clinical risk factors and DXA-measured BMD as inputs led to an improved AUC of 0.90 (0.83-0.96) with a sensitivity of 0.83 and specificity of 0.74. CONCLUSION: These results suggest that using a three-dimensional computer vision method to HR-pQCT scanning may enhance the identification of those at risk of fracture beyond that afforded by clinical risk factors and DXA-measured BMD. This approach has the potential to make the information offered by HR-pQCT more accessible (and therefore) applicable to healthcare professionals in the clinic if the technology becomes more widely available.

A general framework in single and multi-modality registration for lung imaging analysis using statistical prior shapes.

BACKGROUND AND OBJECTIVE: A fusion of multi-slice computed tomography (MSCT) and single photon emission computed tomography (SPECT) represents a powerful tool for chronic obstructive pulmonary disease (COPD) analysis. In this paper, a novel and high-performance MSCT/SPECT non-rigid registration algorithm is proposed to accurately map the lung lobe information onto the functional imaging. Such a fusion can then be used to guide lung volume reduction surgery. METHODS: The multi-modality fusion method proposed here is developed by a multi-channel technique which performs registration from MSCT scan to ventilation and perfusion SPECT scans simultaneously. Furthermore, a novel function with less parameters is also proposed to avoid the adjustment of the weighting parameter and to achieve a better performance in comparison with the exisitng methods in the literature. RESULTS: A lung imaging dataset from a hospital and a synthetic dataset created by software are employed to validate single- and multi-modality registration results. Our method is demonstrated to achieve the improvements in terms of registration accuracy and stability by up to 23% and 54% respectively. Our multi-channel technique proposed here is also proved to obtain improved registration accuracy in comparison with single-channel method. CONCLUSIONS: The fusion of lung lobes onto SPECT imaging is achievable by accurate MSCT/SPECT alignment. It can also be used to perform lobar lung activity analysis for COPD diagnosis and treatment.

Human female attractiveness: waveform analysis of body shape.

Two putative cues to female physical attractiveness are body mass index (BMI) and shape (particularly the waist-hip ratio or WHR). To determine the relative importance of these cues we asked 23 male and 23 female undergraduates to rate a set of 60 pictures of real women's bodies in front-view for attractiveness. In our set of images, the relative ranges of BMI and WHR favoured WHR. We based these ranges on a sample of 457 women. We did not limit the WHR range, although we kept the BMI range to 0.5 s.d. either side of the sample means. As a result, WHR averaged 1.65 s.d. either side of its sample mean. However, even with these advantages, WHR was less important than BMI as a predictor of attractiveness ratings for bodies. BMI is far more strongly correlated with ratings of attractiveness than WHR (BMI approximately 0.5, WHR approximately 0.2). To further explore the relative importance of BMI and WHR, we deliberately chose a subset of these images that demonstrated an inverse correlation of BMI and WHR (i.e. a group in which as images get heavier they also become more curvaceous). If WHR is the most important determinant of attractiveness, then the more curvaceous (but higher BMI) images should be judged most attractive. However, if BMI is a better predictor, then the opposite should be true. We found that the more curvaceous (but higher BMI) images were judged least attractive, thereby inverting the expected rating pattern. This strongly suggests that viewers' judgements were influenced more by BMI than WHR. Finally, it is possible that body shape is an important cue to attractiveness, but that simple ratios (such as WHR) are not adequately capturing it. Therefore, we treated the outline of the torso as a waveform and carried out a set of waveform analyses on it to allow us to quantify body shape and correlate it with attractiveness. The waveform analyses address the complexity of the whole torso shape, and reveal innate properties of the torso shape and not shape elements based on prior decisions about arbitrary physical features. Our analyses decompose the waveform into objective quantified elements whose importance in predicting attractiveness can then be tested. All of the components that were good descriptors of body shape were weakly correlated with attractiveness. Our results suggest that BMI is a stronger predictor of attractiveness than WHR.

Longer fixation duration while viewing face images.

The spatio-temporal properties of saccadic eye movements can be influenced by the cognitive demand and the characteristics of the observed scene. Probably due to its crucial role in social communication, it is argued that face perception may involve different cognitive processes compared with non-face object or scene perception. In this study, we investigated whether and how face and natural scene images can influence the patterns of visuomotor activity. We recorded monkeys' saccadic eye movements as they freely viewed monkey face and natural scene images. The face and natural scene images attracted similar number of fixations, but viewing of faces was accompanied by longer fixations compared with natural scenes. These longer fixations were dependent on the context of facial features. The duration of fixations directed at facial contours decreased when the face images were scrambled, and increased at the later stage of normal face viewing. The results suggest that face and natural scene images can generate different patterns of visuomotor activity. The extra fixation duration on faces may be correlated with the detailed analysis of facial features.

Centre-surround interactions in response to natural scene stimulation in the primary visual cortex.

Centre-surround interaction in the primary visual cortex (area V1) has been studied extensively using artificial, abstract stimulus patterns, such as bars, gratings and simple texture patterns. In this experiment, we extend the study of centre-surround interaction by using natural scene images. We systematically varied the contrast of natural image surrounds presented outside the classical receptive field (CRF), and recorded neuronal response to a natural image patch presented within the CRF in area V1 of awake, fixating macaques. For the majority of neurons (67 out of 111), the natural image surrounds profoundly modulated, mainly by suppressing, neuronal responses to CRF images. These modulatory effects started at the earliest stage of neuronal responses, and often depended on the contrast and higher-order structures of the surrounds. For 47 out of 67 neurons, randomising the phases of the Fourier spectrum of the natural image surround diminished the centre-surround interaction. Our results suggest that the centre-surround interaction in area V1 can be extended to natural vision, and is sensitive to the higher-order structures of natural scene images, such as image contours.

How do monkeys view faces?--A study of eye movements.

Face perception plays a crucial role in primate social communication. We have investigated the pattern of eye movements produced by rhesus monkeys (Macaca mulatta) as they viewed images of faces. Eye positions were recorded accurately using implanted eye coils, while neutral upright, inverted and scrambled images of monkey and human faces were presented on a computer screen. The monkeys exhibited a similar eye scan pattern while viewing familiar and unfamiliar monkey face images, or while viewing monkey and human face images. No differences were observed in the distribution of viewing times, number of fixations, time into the trial of first saccade to local facial features, and the temporal and spatial characteristics of viewing patterns across the facial images. However, there was a greater probability of re-fixation of the eye region of unfamiliar faces during the first few seconds of the trial suggesting that the eyes are important for the initial encoding of identity. Indeed, the highest fixation density was found in the eye region of all the face images. The viewing duration and the number of fixations per image decreased when inverted or scrambled faces were presented. The eye region in these modified images remained the primary area of fixation. However, the number of fixations directed to the eyes decreased monotonically from the upright images through the inverted versions to the scrambled face images. Nonetheless, the eyes remain the most salient facial substructure regardless of the arrangement of other features, although the extent of salience which they attain may depend both on the low level properties of the eyes and on the global arrangement of facial features.