Two-Stage Pedestrian Detection Model Using a New Classification Head for Domain Generalization.

Pedestrian detection based on deep learning methods have reached great success in the past few years with several possible real-world applications including autonomous driving, robotic navigation, and video surveillance. In this work, a new neural network two-stage pedestrian detector with a new custom classification head, adding the triplet loss function to the standard bounding box regression and classification losses, is presented. This aims to improve the domain generalization capabilities of existing pedestrian detectors, by explicitly maximizing inter-class distance and minimizing intra-class distance. Triplet loss is applied to the features generated by the region proposal network, aimed at clustering together pedestrian samples in the features space. We used Faster R-CNN and Cascade R-CNN with the HRNet backbone pre-trained on ImageNet, changing the standard classification head for Faster R-CNN, and changing one of the three heads for Cascade R-CNN. The best results were obtained using a progressive training pipeline, starting from a dataset that is further away from the target domain, and progressively fine-tuning on datasets closer to the target domain. We obtained state-of-the-art results, MR-2 of 9.9, 11.0, and 36.2 for the reasonable, small, and heavy subsets on the CityPersons benchmark with outstanding performance on the heavy subset, the most difficult one.

Development of machine learning models to predict gestational diabetes risk in the first half of pregnancy.

BACKGROUND: Early prediction of Gestational Diabetes Mellitus (GDM) risk is of particular importance as it may enable more efficacious interventions and reduce cumulative injury to mother and fetus. The aim of this study is to develop machine learning (ML) models, for the early prediction of GDM using widely available variables, facilitating early intervention, and making possible to apply the prediction models in places where there is no access to more complex examinations. METHODS: The dataset used in this study includes registries from 1,611 pregnancies. Twelve different ML models and their hyperparameters were optimized to achieve early and high prediction performance of GDM. A data augmentation method was used in training to improve prediction results. Three methods were used to select the most relevant variables for GDM prediction. After training, the models ranked with the highest Area under the Receiver Operating Characteristic Curve (AUCROC), were assessed on the validation set. Models with the best results were assessed in the test set as a measure of generalization performance. RESULTS: Our method allows identifying many possible models for various levels of sensitivity and specificity. Four models achieved a high sensitivity of 0.82, a specificity in the range 0.72-0.74, accuracy between 0.73-0.75, and AUCROC of 0.81. These models required between 7 and 12 input variables. Another possible choice could be a model with sensitivity of 0.89 that requires just 5 variables reaching an accuracy of 0.65, a specificity of 0.62, and AUCROC of 0.82. CONCLUSIONS: The principal findings of our study are: Early prediction of GDM within early stages of pregnancy using regular examinations/exams; the development and optimization of twelve different ML models and their hyperparameters to achieve the highest prediction performance; a novel data augmentation method is proposed to allow reaching excellent GDM prediction results with various models.

Pure platelet-rich plasma and supernatant of calcium-activated P-PRP induce different phenotypes of human macrophages.

AIM: This study aimed to evaluate the effect of two platelet preparations used in the clinic, pure platelet-rich plasma (P-PRP) and the supernatant of calcium-activated P-PRP (S-PRP), on the phenotype of human macrophages. MATERIALS & METHODS: Surface markers and cytokine production of human monocyte-derived macrophages were analyzed after 24 h stimulation with P-PRP or S-PRP. RESULTS: P-PRP and S-PRP present no difference in the expression of CD206, a M2 tissue-repair macrophage-related marker. However, these same macrophages presented different levels of CD163, CD86 as well as different IL-10 secretion capacities after 24 h incubation. CONCLUSION: These platelet preparations do not have an equivalent biological effect over macrophages, which suggest that they may present different clinical regenerative potentials.

Spatially coincident vibrotactile noise improves subthreshold stimulus detection.

Stochastic Resonance (SR) is a phenomenon, mainly present in nonlinear detection systems, in which the addition of certain amount of noise, called optimal noise, has proven to enhance detection performance of subthreshold stimuli. When added noise is present only during the stimulus, an additional enhancement can be reached. This phenomenon was called time Coincidence Enhanced Stochastic Resonance (CESR). The aim of this study was to study the effect of spatially distributed vibrotactile noise in subthreshold stimuli detection. The correct response rates from two different stimuli conditions were compared, using four tactile stimulator systems to excite four different spatial locations on the fingertip. Under two different conditions, the stimuli were present in only one randomly chosen stimulator. For the first condition, all stimulators contain optimal noise level. In the second condition, the optimal noise was present only at the stimulator with the stimulus. SR threshold principle should not produce different correct response rates between the two conditions, since in both cases the noise enables the subthreshold stimulus to go above threshold. The stimulus signal used was a rectangular displacement controlled pulse that lasted 300ms within a 1.5s attention interval, applied to the exploratory zone of the index finger of 13 human subjects. For all subjects it was found that detection rates were better (p<0.0003) when noise was spatially coincident with the stimulus, compared to the condition in which noise was present simultaneously in all the stimulators. According to our literature review this is the first report of SR being influenced by the spatial location of the noise. These results were not found previously reported, so represent the discovery of a new phenomenon. We call this phenomenon Spatial-Coincidence-Enhanced Stochastic Resonance (SCESR). As results show, the optimal noise level is dependent on the relative position between stimulus and noise.

Sleep-spindle identification on EEG signals from polysomnographie recordings using correntropy.

Sleep spindles (SSs) are characteristic electroencephalographic (EEG) waveforms of sleep stages N2 and N3. One of the main problems associated with SS detection is the high number of false positives. In this paper we propose a new periodogram based on correntropy to detect SSs and enhance their characterization. Correntropy is a generalized correlation, under the information theoretic learning framework. A non-negative matrix factorization decomposition of correntropy allows us to obtain a new periodogram, which shows an improved resolution capability compared to the conventional power spectrum density. Preliminary results show that the proposed method obtained a sensitivity rate of 0.868 with a false positive rate of 0.121.

Dendritic cell chimerism in oral mucosa of transplanted patients affected by graft-versus-host disease.

OBJECTIVE: Graft-versus-host disease (GVHD) is one of the main complications after haematopoietic stem cell transplantation. Clinical features of GVHD include either an acute (aGVHD) or a chronic (cGVHD) condition that affects locations such as the oral mucosa. While the involvement of the host's dendritic cells (DCs) has been demonstrated in aGVHD, the origin (donor/host) and mechanisms underlying oral cGVHD have not been completely elucidated. In this study, we intend to determine the origin of DCs present in mucosal tissue biopsies from the oral cavity of transplanted patients affected by cGVHD. METHODS: We purified DCs, from oral biopsies of three patients with cGVHD, through immunobeads and subsequently performed DNA extraction. The origin of the obtained DCs was determined by PCR amplification of 13 informative short tandem repeat (STR) alleles. We also characterised the DCs phenotype and the inflammatory infiltrate from biopsies of two patients by immunohistochemistry. RESULTS: Clinical and histological features of the biopsies were concordant with oral cGVHD. We identified CD11c-, CD207- and CD1a-positive cells in the epithelium and beneath the basal layer. Purification of DCs from the mucosa of patients affected by post-transplantation cGVHD was >95%. PCR-STR data analysis of DCs DNA showed that 100% of analysed cells were of donor origin in all of the evaluated patients. CONCLUSION: Our results demonstrate that resident DCs isolated from the oral tissue of allotransplanted patients affected by cGVHD are originated from the donor. Further research will clarify the role of DCs in the development and/or severity of oral cGVHD.

An accurate skull stripping method based on simplex meshes and histogram analysis for magnetic resonance images.

Skull stripping methods are designed to eliminate the non-brain tissue in magnetic resonance (MR) brain images. Removal of non-brain tissues is a fundamental step in enabling the processing of brain MR images. The aim of this study is to develop an automatic accurate skull stripping method based on deformable models and histogram analysis. A rough-segmentation step is used to find the optimal starting point for the deformation and is based on thresholds and morphological operators. Thresholds are computed using comparisons with an atlas, and modeling by Gaussians. The deformable model is based on a simplex mesh and its deformation is controlled by the image local gray levels and the information obtained on the gray level modeling of the rough-segmentation. Our Simplex Mesh and Histogram Analysis Skull Stripping (SMHASS) method was tested on the following international databases commonly used in scientific articles: BrainWeb, Internet Brain Segmentation Repository (IBSR), and Segmentation Validation Engine (SVE). A comparison was performed against three of the best skull stripping methods previously published: Brain Extraction Tool (BET), Brain Surface Extractor (BSE), and Hybrid Watershed Algorithm (HWA). Performance was measured using the Jaccard index (J) and Dice coefficient (κ). Our method showed the best performance and differences were statistically significant (p<0.05): J=0.904 and κ=0.950 on BrainWeb; J=0.905 and κ=0.950 on IBSR; J=0.946 and κ=0.972 on SVE.

Automated detection of rapid eye movements in children.

We present an automated multiple-step tool to identify Rapid Eye Movements (REMs) in the polysomnogram, based on modeling expert criteria. It begins by identifying the polysomnogram segments compatible with REMs presence. On these segments, high-energy REMs are identified. Then, vicinity zones around those REMs are defined, and lesser-energy REMs are sought in these vicinities. This strategy has the advantage that it can detect lesser-energy REMs without increasing much the false positive detections. Signal processing, feature extraction, and fuzzy logic tools are used to achieve the goal. The tool was trained and validated on a database consisting of 20 all-night polysomnogram recordings (160 hr) of healthy ten-year-old children. Preliminary results on the validation set show 85.5% sensitivity and a false positive rate of 16.2%. Our tool works on complete polysomnogram recordings, without the need of preprocessing, prior knowledge of the hypnogram, or noise-free segments selection.

Registration of renal SPECT and 2.5D US images.

Image registration is the process of transforming different image data sets of an object into the same coordinate system. This is a relevant task in the field of medical imaging; one of its objectives is to combine information from different imaging modalities. The main goal of this study is the registration of renal SPECT (Single Photon Emission Computerized Tomography) images and a sparse set of ultrasound slices (2.5D US), combining functional and anatomical information. Registration is performed after kidney segmentation in both image types. The SPECT segmentation is achieved using a deformable model based on a simplex mesh. The 2.5D US image segmentation is carried out in each of the 2D slices employing a deformable contour and Gabor filters to capture multi-scale image features. Moreover, a renal medulla detection method was developed to improve the US segmentation. A nonlinear optimization algorithm is used for the registration. In this process, movements caused by patient breathing during US image acquisition are also corrected. Only a few reports describe registration between SPECT images and a sparse set of US slices of the kidney, and they usually employ an optical localizer, unlike our method, that performs movement correction using information only from the SPECT and US images. Moreover, it does not require simultaneous acquisition of both image types. The registration method and both segmentations were evaluated separately. The SPECT segmentation was evaluated qualitatively by medical experts, obtaining a score of 5 over a scale from 1 to 5, where 5 represents a perfect segmentation. The 2.5D US segmentation was evaluated quantitatively, by comparing our method with an expert manual segmentation, and obtaining an average error of 3.3mm. The registration was evaluated quantitatively and qualitatively. Quantitatively the distance between the manual segmentation of the US images and the model extracted from the SPECT image was measured, obtaining an average distance of 1.07 pixels on 7 exams. The qualitative evaluation was carried out by a group of physicians who assessed the perceived clinical usefulness of the image registration, rating each registration on a scale from 1 to 5. The average score obtained was 4.1, i.e. relevantly useful for medical purposes.

A critical experimental study of the classical tactile threshold theory.

BACKGROUND: The tactile sense is being used in a variety of applications involving tactile human-machine interfaces. In a significant number of publications the classical threshold concept plays a central role in modelling and explaining psychophysical experimental results such as in stochastic resonance (SR) phenomena. In SR, noise enhances detection of sub-threshold stimuli and the phenomenon is explained stating that the required amplitude to exceed the sensory threshold barrier can be reached by adding noise to a sub-threshold stimulus. We designed an experiment to test the validity of the classical vibrotactile threshold. Using a second choice experiment, we show that individuals can order sensorial events below the level known as the classical threshold. If the observer's sensorial system is not activated by stimuli below the threshold, then a second choice could not be above the chance level. Nevertheless, our experimental results are above that chance level contradicting the definition of the classical tactile threshold. RESULTS: We performed a three alternative forced choice detection experiment on 6 subjects asking them first and second choices. In each trial, only one of the intervals contained a stimulus and the others contained only noise. According to the classical threshold assumptions, a correct second choice response corresponds to a guess attempt with a statistical frequency of 50%. Results show an average of 67.35% (STD = 1.41%) for the second choice response that is not explained by the classical threshold definition. Additionally, for low stimulus amplitudes, second choice correct detection is above chance level for any detectability level. CONCLUSIONS: Using a second choice experiment, we show that individuals can order sensorial events below the level known as a classical threshold. If the observer's sensorial system is not activated by stimuli below the threshold, then a second choice could not be above the chance level. Nevertheless, our experimental results are above that chance level. Therefore, if detection exists below the classical threshold level, then the model to explain the SR phenomenon or any other tactile perception phenomena based on the psychophysical classical threshold is not valid. We conclude that a more suitable model of the tactile sensory system is needed.

Automated sleep-spindle detection in healthy children polysomnograms.

We present a new methodology to detect and characterize sleep spindles (SSs), based on the nonlinear algorithms, empirical-mode decomposition, and Hilbert-Huang transform, which provide adequate temporal and frequency resolutions in the electroencephalographic analysis. In addition, the application of fuzzy logic allows to emulate expert's procedures. Additionally, we built a database of 56 all-night polysomnographic recordings from children for training and testing, which is among the largest annotated databases published on the subject. The database was split into training (27 recordings), validation (10 recordings), and testing (19 recordings) datasets. The SS events were marked by sleep experts using visual inspection, and these marks were used as golden standard. The overall SS detection performance on the testing dataset of continuous all-night sleep recordings was 88.2% sensitivity, 89.7% specificity, and 11.9% false-positive (FP) rate. Considering only non-REM sleep stage 2, the results showed 92.2% sensitivity, 90.1% specificity, and 8.9% FP rate. In general, our system presents enhanced results when compared with most systems found in the literature, thus improving SS detection precision significantly without the need of hypnogram information.

Normalized mutual information feature selection.

A filter method of feature selection based on mutual information, called normalized mutual information feature selection (NMIFS), is presented. NMIFS is an enhancement over Battiti's MIFS, MIFS-U, and mRMR methods. The average normalized mutual information is proposed as a measure of redundancy among features. NMIFS outperformed MIFS, MIFS-U, and mRMR on several artificial and benchmark data sets without requiring a user-defined parameter. In addition, NMIFS is combined with a genetic algorithm to form a hybrid filter/wrapper method called GAMIFS. This includes an initialization procedure and a mutation operator based on NMIFS to speed up the convergence of the genetic algorithm. GAMIFS overcomes the limitations of incremental search algorithms that are unable to find dependencies between groups of features.

Coincidence-enhanced stochastic resonance: experimental evidence challenges the psychophysical theory behind stochastic resonance.

Stochastic resonance (SR) is the counterintuitive phenomenon in which noise enhances detection of sub-threshold stimuli. The SR psychophysical threshold theory establishes that the required amplitude to exceed the sensory threshold barrier can be reached by adding noise to a sub-threshold stimulus. The aim of this study was to test the SR theory by comparing detection results from two different randomly-presented stimulus conditions. In the first condition, optimal noise was present during the whole attention interval; in the second, the optimal noise was restricted to the same time interval as the stimulus. SR threshold theory predicts no difference between the two conditions because noise helps the sub-threshold stimulus to reach threshold in both cases. The psychophysical experimental method used a 300 ms rectangular force pulse as a stimulus within an attention interval of 1.5 s, applied to the index finger of six human subjects in the two distinct conditions. For all subjects we show that in the condition in which the noise was present only when synchronized with the stimulus, detection was better (p<0.05) than in the condition in which the noise was delivered throughout the attention interval. These results provide the first direct evidence that SR threshold theory is incomplete and that a new phenomenon has been identified, which we call Coincidence-Enhanced Stochastic Resonance (CESR). We propose that CESR might occur because subject uncertainty is reduced when noise points at the same temporal window as the stimulus.

Extracting fuzzy rules from polysomnographic recordings for infant sleep classification.

A neuro-fuzzy classifier (NFC) of sleep-wake states and stages has been developed for healthy infants of ages 6 mo and onward. The NFC takes five input patterns previously identified on 20-s epochs from polysomnographic recordings and assigns them to one out of five possible classes: Wakefulness, REM-Sleep, Non-REM Sleep Stage 1, Stage 2, and Stage 3-4. The definite criterion for a sleep state or stage to be established is duration of at least 1 min. The data set consisted of a total of 14 continuous recordings of naturally occurring naps (average duration: 143 +/- 39 min), corresponding to a total of 6021 epochs. They were divided in a training, a validation and a test set with 7, 2, and 5 recordings, respectively. During supervised training, the system determined the fuzzy concepts associated to the inputs and the rules required for performing the classification, extracting knowledge from the training set, and pruning nonrelevant rules. Results on an independent test set achieved 83.9 +/- 0.4% of expert agreement. The fuzzy rules obtained from the training examples without a priori information showed a high level of coincidence with the crisp rules stated by the experts, which are based on internationally accepted criteria. These results show that the NFC can be a valuable tool for implementing an automated sleep-wake classification system.

Genetic polymorphism at position-308 in the promoter region of the tumornecrosis factor (TNF): implications of its allelic distribution on susceptibility or resistance to diseases in the Chilean population

Several single-nucleotide polymorphisms have been identified in the human TNF gene promoter. The polymorphism at position-308 (TNF-308), which involves substituting G for A and designing the TNF2 allele, leads to a higher rate of TNF gene transcription than the wild-type TNF1 allele in in vitro expression studies. It has also been linked to increased susceptibility to a variety of illnesses. Using PCR-RFLP analysis we detected significant differences in the TNF-308 genotypes of Chilean and other populations. We conclude that there is a gradient in the distribution of the TNF2 allele according to ethnicity; we have also hypothesized that populations bearing a higher proportion of the TNF2 allele may have an increased predisposition toward or incidence of several chronic metabolic, degenerative, inflammatory and autoimmune diseases.