Radiomics Boosts Deep Learning Model for IPMN Classification.

Intraductal Papillary Mucinous Neoplasm (IPMN) cysts are pre-malignant pancreas lesions, and they can progress into pancreatic cancer. Therefore, detecting and stratifying their risk level is of ultimate importance for effective treatment planning and disease control. However, this is a highly challenging task because of the diverse and irregular shape, texture, and size of the IPMN cysts as well as the pancreas. In this study, we propose a novel computer-aided diagnosis pipeline for IPMN risk classification from multi-contrast MRI scans. Our proposed analysis framework includes an efficient volumetric self-adapting segmentation strategy for pancreas delineation, followed by a newly designed deep learning-based classification scheme with a radiomics-based predictive approach. We test our proposed decision-fusion model in multi-center data sets of 246 multi-contrast MRI scans and obtain superior performance to the state of the art (SOTA) in this field. Our ablation studies demonstrate the significance of both radiomics and deep learning modules for achieving the new SOTA performance compared to international guidelines and published studies (81.9% vs 61.3% in accuracy). Our findings have important implications for clinical decision-making. In a series of rigorous experiments on multi-center data sets (246 MRI scans from five centers), we achieved unprecedented performance (81.9% accuracy). The code is available upon publication.

Fully automatic segmentation of the mandible based on convolutional neural networks (CNNs).

OBJECTIVES: To evaluate the accuracy of automatic deep learning-based method for fully automatic segmentation of the mandible from CBCTs. SETTING AND SAMPLE POPULATION: CBCT-derived mandible fully automatic segmentation. METHODS: Forty CBCT scans from healthy patients (20 females and 20 males, mean age 23.37 ± 3.34) were collected, and a manual mandible segmentation was carried out by using Mimics software. Twenty CBCT scans were randomly selected and used for training the artificial intelligence model file. The remaining 20 CBCT segmentation masks were used to test the accuracy of the CNN automatic method by comparing the segmentation volumes of the 3D models obtained with automatic and manual segmentations. The accuracy of the CNN-based method was also assessed by using the DICE Score coefficient (DSC) and by the surface-to-surface matching technique. The intraclass correlation coefficient (ICC) and Dahlberg's formula were used respectively to test the intra-observer reliability and method error. Independent Student's t test was used for between-groups volumetric comparison. RESULTS: Measurements were highly correlated with an ICC value of 0.937, while the method error was 0.24 mm3 . A difference of 0.71 (±0.49) cm3 was found between the methodologies, but it was not statistically significant (P > .05). The matching percentage detected was 90.35% (±1.88) (tolerance 0.5 mm) and 96.32% ± 1.97% (tolerance 1.0 mm). The differences, measured as DSC in percentage, between the assessments done with both methods were, respectively, 2.8% and 3.1%. CONCLUSION: The tested deep learning CNN-based technology is accurate and performs as well as an experienced image reader but at much higher speed, which is of significant clinical relevance.

An explainable AI system for automated COVID-19 assessment and lesion categorization from CT-scans.

COVID-19 infection caused by SARS-CoV-2 pathogen has been a catastrophic pandemic outbreak all over the world, with exponential increasing of confirmed cases and, unfortunately, deaths. In this work we propose an AI-powered pipeline, based on the deep-learning paradigm, for automated COVID-19 detection and lesion categorization from CT scans. We first propose a new segmentation module aimed at automatically identifying lung parenchyma and lobes. Next, we combine the segmentation network with classification networks for COVID-19 identification and lesion categorization. We compare the model's classification results with those obtained by three expert radiologists on a dataset of 166 CT scans. Results showed a sensitivity of 90.3% and a specificity of 93.5% for COVID-19 detection, at least on par with those yielded by the expert radiologists, and an average lesion categorization accuracy of about 84%. Moreover, a significant role is played by prior lung and lobe segmentation, that allowed us to enhance classification performance by over 6 percent points. The interpretation of the trained AI models reveals that the most significant areas for supporting the decision on COVID-19 identification are consistent with the lesions clinically associated to the virus, i.e., crazy paving, consolidation and ground glass. This means that the artificial models are able to discriminate a positive patient from a negative one (both controls and patients with interstitial pneumonia tested negative to COVID) by evaluating the presence of those lesions into CT scans. Finally, the AI models are integrated into a user-friendly GUI to support AI explainability for radiologists, which is publicly available at http://perceivelab.com/covid-ai. The whole AI system is unique since, to the best of our knowledge, it is the first AI-based software, publicly available, that attempts to explain to radiologists what information is used by AI methods for making decisions and that proactively involves them in the decision loop to further improve the COVID-19 understanding.

Deep Neuro-Vision Embedded Architecture for Safety Assessment in Perceptive Advanced Driver Assistance Systems: The Pedestrian Tracking System Use-Case.

In recent years, the automotive field has been changed by the accelerated rise of new technologies. Specifically, autonomous driving has revolutionized the car manufacturer's approach to design the advanced systems compliant to vehicle environments. As a result, there is a growing demand for the development of intelligent technology in order to make modern vehicles safer and smarter. The impact of such technologies has led to the development of the so-called Advanced Driver Assistance Systems (ADAS), suitable to maintain control of the vehicle in order to avoid potentially dangerous situations while driving. Several studies confirmed that an inadequate driver's physiological condition could compromise the ability to drive safely. For this reason, assessing the car driver's physiological status has become one of the primary targets of the automotive research and development. Although a large number of efforts has been made by researchers to design safety-assessment applications based on the detection of physiological signals, embedding them into a car environment represents a challenging task. These mentioned implications triggered the development of this study in which we proposed an innovative pipeline, that through a combined less invasive Neuro-Visual approach, is able to reconstruct the car driver's physiological status. Specifically, the proposed contribution refers to the sampling and processing of the driver PhotoPlethysmoGraphic (PPG) signal. A parallel enhanced low frame-rate motion magnification algorithm is used to reconstruct such features of the driver's PhotoPlethysmoGraphic (PPG) data when that signal is no longer available from the native embedded sensor platform. A parallel monitoring of the driver's blood pressure levels from the PPG signal as well as the driver's eyes dynamics completes the reconstruction of the driver's physiological status. The proposed pipeline has been tested in one of the major investigated automotive scenarios i.e., the detection and monitoring of pedestrians while driving (pedestrian tracking). The collected performance results confirmed the effectiveness of the proposed approach.

Fully automatic segmentation of sinonasal cavity and pharyngeal airway based on convolutional neural networks.

INTRODUCTION: This study aimed to test the accuracy of a new automatic deep learning-based approach on the basis of convolutional neural networks (CNN) for fully automatic segmentation of the sinonasal cavity and the pharyngeal airway from cone-beam computed tomography (CBCT) scans. METHODS: Forty CBCT scans from healthy patients (20 women and 20 men; mean age, 23.37 ± 3.34 years) were collected, and manual segmentation of the sinonasal cavity and pharyngeal subregions were carried out by using Mimics software (version 20.0; Materialise, Leuven, Belgium). Twenty CBCT scans from the total sample were randomly selected and used for training the artificial intelligence model file. The remaining 20 CBCT segmentation masks were used to test the accuracy of the CNN fully automatic method by comparing the segmentation volumes of the 3-dimensional models obtained with automatic and manual segmentations. The accuracy of the CNN-based method was also assessed by using the Dice score coefficient and by the surface-to-surface matching technique. The intraclass correlation coefficient and Dahlberg's formula were used to test the intraobserver reliability and method error, respectively. Independent Student t test was used for between-groups volumetric comparison. RESULTS: Measurements were highly correlated with an intraclass correlation coefficient value of 0.921, whereas the method error was 0.31 mm3. A mean difference of 1.93 ± 0.73 cm3 was found between the methodologies, but it was not statistically significant (P >0.05). The mean matching percentage detected was 85.35 ± 2.59 (tolerance 0.5 mm) and 93.44 ± 2.54 (tolerance 1.0 mm). The differences, measured as the Dice score coefficient in percentage, between the assessments done with both methods were 3.3% and 5.8%, respectively. CONCLUSIONS: The new deep learning-based method for automated segmentation of the sinonasal cavity and the pharyngeal airway in CBCT scans is accurate and performs equally well as an experienced image reader.

Information Bottleneck Attribution for Visual Explanations of Diagnosis and Prognosis.

Visual explanation methods have an important role in the prognosis of the patients where the annotated data is limited or unavailable. There have been several attempts to use gradient-based attribution methods to localize pathology from medical scans without using segmentation labels. This research direction has been impeded by the lack of robustness and reliability. These methods are highly sensitive to the network parameters. In this study, we introduce a robust visual explanation method to address this problem for medical applications. We provide an innovative visual explanation algorithm for general purpose and as an example application we demonstrate its effectiveness for quantifying lesions in the lungs caused by the Covid-19 with high accuracy and robustness without using dense segmentation labels. This approach overcomes the drawbacks of commonly used Grad-CAM and its extended versions. The premise behind our proposed strategy is that the information flow is minimized while ensuring the classifier prediction stays similar. Our findings indicate that the bottleneck condition provides a more stable severity estimation than the similar attribution methods. The source code will be publicly available upon publication.

Hierarchical 3D Feature Learning for Pancreas Segmentation.

We propose a novel 3D fully convolutional deep network for automated pancreas segmentation from both MRI and CT scans. More specifically, the proposed model consists of a 3D encoder that learns to extract volume features at different scales; features taken at different points of the encoder hierarchy are then sent to multiple 3D decoders that individually predict intermediate segmentation maps. Finally, all segmentation maps are combined to obtain a unique detailed segmentation mask. We test our model on both CT and MRI imaging data: the publicly available NIH Pancreas-CT dataset (consisting of 82 contrast-enhanced CTs) and a private MRI dataset (consisting of 40 MRI scans). Experimental results show that our model outperforms existing methods on CT pancreas segmentation, obtaining an average Dice score of about 88%, and yields promising segmentation performance on a very challenging MRI data set (average Dice score is about 77%). Additional control experiments demonstrate that the achieved performance is due to the combination of our 3D fully-convolutional deep network and the hierarchical representation decoding, thus substantiating our architectural design.

Decoding Brain Representations by Multimodal Learning of Neural Activity and Visual Features.

This work presents a novel method of exploring human brain-visual representations, with a view towards replicating these processes in machines. The core idea is to learn plausible computational and biological representations by correlating human neural activity and natural images. Thus, we first propose a model, EEG-ChannelNet, to learn a brain manifold for EEG classification. After verifying that visual information can be extracted from EEG data, we introduce a multimodal approach that uses deep image and EEG encoders, trained in a siamese configuration, for learning a joint manifold that maximizes a compatibility measure between visual features and brain representations. We then carry out image classification and saliency detection on the learned manifold. Performance analyses show that our approach satisfactorily decodes visual information from neural signals. This, in turn, can be used to effectively supervise the training of deep learning models, as demonstrated by the high performance of image classification and saliency detection on out-of-training classes. The obtained results show that the learned brain-visual features lead to improved performance and simultaneously bring deep models more in line with cognitive neuroscience work related to visual perception and attention.

Deep Learning Provides Exceptional Accuracy to ECoG-Based Functional Language Mapping for Epilepsy Surgery.

The success of surgical resection in epilepsy patients depends on preserving functionally critical brain regions, while removing pathological tissues. Being the gold standard, electro-cortical stimulation mapping (ESM) helps surgeons in localizing the function of eloquent cortex through electrical stimulation of electrodes placed directly on the cortical brain surface. Due to the potential hazards of ESM, including increased risk of provoked seizures, electrocorticography based functional mapping (ECoG-FM) was introduced as a safer alternative approach. However, ECoG-FM has a low success rate when compared to the ESM. In this study, we address this critical limitation by developing a new algorithm based on deep learning for ECoG-FM and thereby we achieve an accuracy comparable to ESM in identifying eloquent language cortex. In our experiments, with 11 epilepsy patients who underwent presurgical evaluation (through deep learning-based signal analysis on 637 electrodes), our proposed algorithm obtained an accuracy of 83.05% in identifying language regions, an exceptional 23% improvement with respect to the conventional ECoG-FM analysis (∼60%). Our findings have demonstrated, for the first time, that deep learning powered ECoG-FM can serve as a stand-alone modality and avoid likely hazards of the ESM in epilepsy surgery. Hence, reducing the potential for developing post-surgical morbidity in the language function.

MASK-RL: Multiagent Video Object Segmentation Framework Through Reinforcement Learning.

Integrating human-provided location priors into video object segmentation has been shown to be an effective strategy to enhance performance, but their application at large scale is unfeasible. Gamification can help reduce the annotation burden, but it still requires user involvement. We propose a video object segmentation framework that leverages the combined advantages of user feedback for segmentation and gamification strategy by simulating multiple game players through a reinforcement learning (RL) model that reproduces human ability to pinpoint moving objects and using the simulated feedback to drive the decisions of a fully convolutional deep segmentation network. Experimental results on the DAVIS-17 benchmark show that: 1) including user-provided prior, even if not precise, yields high performance; 2) our RL agent replicates satisfactorily the same variability of humans in identifying spatiotemporal salient objects; and 3) employing artificially generated priors in an unsupervised video object segmentation model reaches state-of-the-art performance.

Three-dimensional analysis of mandibular functional units in adult patients with unilateral posterior crossbite: A cone beam study with the use of mirroring and surface-to-surface matching techniques.

OBJECTIVES: To use three-dimensional (3D) mirroring and surface-to-surface techniques to determine any differences in mandibular functional unit shape and morphology between the crossbite side and non-crossbite side in adult patients with posterior unilateral crossbite who had not received any corrective treatment for malocclusion. MATERIALS AND METHODS: Cone-beam computed tomography (CBCT) records from 24 consecutive adult white patients (mean age, 27.5 years; range 22.6-39.7 years; 14 women and 10 men) seeking treatment for maxillary transverse deficiency were assessed in this study. The control group comprised CBCT scans from age- and sex-matched patients. Segmentation masks were generated to obtain 3D surface mesh models of the mandibles and analyze the six skeletal functional units, which were further analyzed with reverse engineering software. RESULTS: Statistically significant differences in the mean surface distance when comparing the study sample and the control sample were found at the condylar process, mandibular ramus, angular process (P ≤ .0001), and alveolar process (P ≤ .01); no statistically significant differences were found for the coronoid process, the chin, and the mandibular body (P ≥ .5). CONCLUSIONS: The condylar, angular, and alveolar processes plus the mandibular ramus appear to play a more dominant role than did the body, the coronoid, and the chin units in the asymmetry of the mandible in patients with unilateral crossbite.

Online comprehension across different semantic categories in preschool children with autism spectrum disorder.

BACKGROUND: Word comprehension across semantic categories is a key area of language development. Using online automated eye-tracking technology to reduce response demands during a word comprehension test may be advantageous in children with autism spectrum disorder (ASD). OBJECTIVES: To measure online accuracy of word recognition across eleven semantic categories in preschool children with ASD and in typically developing (TD) children matched for gender and developmental age. METHODS: Using eye-tracker methodology we measured the relative number of fixations on a target image as compared to a foil of the same category shown simultaneously on screen. This online accuracy measure was considered a measure of word understanding. We tested the relationship between online accuracy and offline word recognition and the effects of clinical variables on online accuracy. Twenty-four children with ASD and 21 TD control children underwent the eye-tracking task. RESULTS: On average, children with ASD were significantly less accurate at fixating on the target image than the TD children. After multiple comparison correction, no significant differences were found across the eleven semantic categories of the experiment between preschool children with ASD and younger TD children matched for developmental age. The ASD group showed higher intragroup variability consistent with greater variation in vocabulary growth rates. Direct effects of non-verbal cognitive levels, vocabulary levels and gesture productions on online word recognition in both groups support a dimensional view of language abilities in ASD. CONCLUSIONS: Online measures of word comprehension across different semantic categories show higher interindividual variability in children with ASD and may be useful for objectively monitor gains on targeted language interventions.

Gamifying Video Object Segmentation.

Video object segmentation can be considered as one of the most challenging computer vision problems. Indeed, so far, no existing solution is able to effectively deal with the peculiarities of real-world videos, especially in cases of articulated motion and object occlusions; limitations that appear more evident when we compare the performance of automated methods with the human one. However, manually segmenting objects in videos is largely impractical as it requires a lot of time and concentration. To address this problem, in this paper we propose an interactive video object segmentation method, which exploits, on one hand, the capability of humans to identify correctly objects in visual scenes, and on the other hand, the collective human brainpower to solve challenging and large-scale tasks. In particular, our method relies on a game with a purpose to collect human inputs on object locations, followed by an accurate segmentation phase achieved by optimizing an energy function encoding spatial and temporal constraints between object regions as well as human-provided location priors. Performance analysis carried out on complex video benchmarks, and exploiting data provided by over 60 users, demonstrated that our method shows a better trade-off between annotation times and segmentation accuracy than interactive video annotation and automated video object segmentation approaches.

Correlation between Motor Cortex Excitability Changes and Cognitive Impairment in Vascular Depression: Pathophysiological Insights from a Longitudinal TMS Study.

Background. Transcranial magnetic stimulation (TMS) highlighted functional changes in dementia, whereas there are few data in patients with vascular cognitive impairment-no dementia (VCI-ND). Similarly, little is known about the neurophysiological impact of vascular depression (VD) on deterioration of cognitive functions. We test whether depression might affect not only cognition but also specific cortical circuits in subcortical vascular disease. Methods. Sixteen VCI-ND and 11 VD patients, age-matched with 15 controls, underwent a clinical-cognitive, neuroimaging, and TMS assessment. After approximately two years, all participants were prospectively reevaluated. Results. At baseline, a significant more pronounced intracortical facilitation (ICF) was found in VCI-ND patients. Reevaluation revealed an increase of the global excitability in both VCI-ND and VD subjects. At follow-up, the ICF of VCI-ND becomes similar to the other groups. Only VD patients showed cognitive deterioration. Conclusions. Unlike VD, the hyperfacilitation found at baseline in VCI-ND patients suggests enhanced glutamatergic neurotransmission that might contribute to the preservation of cognitive functioning. The hyperexcitability observed at follow-up in both groups of patients also indicates functional changes in glutamatergic neurotransmission. The mechanisms enhancing the risk of dementia in VD might be related either to subcortical vascular lesions or to the lack of compensatory functional cortical changes.

Modeling skeletal bone development with hidden Markov models.

This paper presents a tool for automatic assessment of skeletal bone age according to a modified version of the Tanner and Whitehouse (TW2) clinical method. The tool is able to provide an accurate bone age assessment in the range 0-6 years by processing epiphysial/metaphysial ROIs with image-processing techniques, and assigning TW2 stage to each ROI by means of hidden Markov models. The system was evaluated on a set of 360 X-rays (180 for males and 180 for females) achieving a high success rate in bone age evaluation (mean error rate of 0.41±0.33 years comparable to human error) as well as outperforming other effective methods. The paper also describes the graphical user interface of the tool, which is also released, thus to support and speed up clinicians' practices when dealing with bone age assessment.

Parallel stochastic systems biology in the cloud.

The stochastic modelling of biological systems, coupled with Monte Carlo simulation of models, is an increasingly popular technique in bioinformatics. The simulation-analysis workflow may result computationally expensive reducing the interactivity required in the model tuning. In this work, we advocate the high-level software design as a vehicle for building efficient and portable parallel simulators for the cloud. In particular, the Calculus of Wrapped Components (CWC) simulator for systems biology, which is designed according to the FastFlow pattern-based approach, is presented and discussed. Thanks to the FastFlow framework, the CWC simulator is designed as a high-level workflow that can simulate CWC models, merge simulation results and statistically analyse them in a single parallel workflow in the cloud. To improve interactivity, successive phases are pipelined in such a way that the workflow begins to output a stream of analysis results immediately after simulation is started. Performance and effectiveness of the CWC simulator are validated on the Amazon Elastic Compute Cloud.

Different patterns of cortical excitability in major depression and vascular depression: a transcranial magnetic stimulation study.

BACKGROUND: Clinical and functional studies consider major depression (MD) and vascular depression (VD) as different neurobiological processes. Hypoexcitability of the left frontal cortex to transcranial magnetic stimulation (TMS) is frequently reported in MD, whereas little is known about the effects of TMS in VD. Thus, we aimed to assess and compare motor cortex excitability in patients with VD and MD. METHODS: Eleven VD patients, 11 recurrent drug-resistant MD patients, and 11 healthy controls underwent clinical, neuropsychological and neuroimaging evaluations in addition to bilateral resting motor threshold, cortical silent period, and paired-pulse TMS curves of intracortical excitability. All patients continued on psychotropic drugs, which were unchanged throughout the study. RESULTS: Scores on one of the tests evaluating frontal lobe abilities (Stroop Color-Word interference test) were worse in patients compared with controls. The resting motor threshold in patients with MD was significantly higher in the left hemisphere compared with the right (p < 0.05), and compared with the VD patients and controls. The cortical silent period was bilaterally prolonged in MD patients compared with VD patients and controls, with a statistically significant difference in the left hemisphere (p < 0.01). No differences were observed in the paired-pulse curves between patients and controls. CONCLUSIONS: This study showed distinctive patterns of motor cortex excitability between late-onset depression with subcortical vascular disease and early-onset recurrent drug resistant MD. The data provide a TMS model of the different processes underlying VD and MD. Additionally, our results support the "Vascular depression hypothesis" at the neurophysiological level, and confirm the inter-hemispheric asymmetry to TMS in patients with MD. We were unable to support previous findings of impaired intracortical inhibitory mechanisms to TMS in patients with MD, although a drug-induced effect on our results cannot be excluded. This study may aid the understanding of the pathogenetic differences underlying the clinical spectrum of depressive disorders.

Preserved transcallosal inhibition to transcranial magnetic stimulation in nondemented elderly patients with leukoaraiosis.

Structural corpus callosum (CC) changes in patients with leukoaraiosis (LA) are significantly associated with cognitive and motor impairment. The aim of this study is to investigate the transcallosal fibers functioning by means of transcranial magnetic stimulation (TMS) in elderly patients with LA. The resting motor threshold (rMT), the motor-evoked potentials (MEPs), and the controlateral (cSP) and ipsilateral silent periods (iSP) were determined using single-pulse TMS in 15 patients and 15 age-matched controls. The neuropsychological profile and the vascular burden at brain magnetic resonance imaging (MRI) were concomitantly explored. Patients reported abnormal scores at tests evaluating executive control functions. No significant difference was found in TMS measures of intra- and intercortical excitability. No CC lesion was evident at MRI. Transcallosal inhibitory mechanisms to TMS seem to be spared in LA patients, a finding which is in line with neuroimaging features and suggests a functional integrity of the CC despite the ischemic interruption of corticosubcortical loops implicated in cognition and behavior. The observed neurophysiological finding differs from that reported in degenerative dementia, even in the preclinical or early stage. In our group of patients, the pure extent of LA is more related to impairment of frontal lobe abilities rather than functional callosal changes.

TMS follow-up study in patients with vascular cognitive impairment-no dementia.

Vascular cognitive impairment-no dementia (VCI-ND) is a condition at risk for future dementia and should be the target of preventive strategies. Recently, an enhanced intracortical facilitation observed in VCI-ND patients was proposed as a candidate neurophysiological marker of the disease process. The aim of this study was to monitor the excitability of the motor cortex and the functioning of excitatory/inhibitory intracortical circuits in patients with VCI-ND after a follow-up period of approximately 2 years, in order to pick out early markers of disease progression into dementia. Nine patients and 9 age-matched controls were re-evaluated for single and paired pulse TMS measures of cortical excitability, as well as for neuropsycological and functional assessment. Compared to the first evaluation, patients showed a decrease of the median resting motor threshold (rMT). Patients exhibited a significant worsening at Stroop Color-Word Test Interference scores without substantial functional impairment. Our study represents the first evidence of a decrease of rMT in VCI-ND patients during the progression of cognitive impairment. This result might be considered an index of motor cortex plasticity and interpreted as a compensatory mechanism for the loss of motor cortex neurons.