Fighting the scanner effect in brain MRI segmentation with a progressive level-of-detail network trained on multi-site data.

Many clinical and research studies of the human brain require accurate structural MRI segmentation. While traditional atlas-based methods can be applied to volumes from any acquisition site, recent deep learning algorithms ensure high accuracy only when tested on data from the same sites exploited in training (i.e., internal data). Performance degradation experienced on external data (i.e., unseen volumes from unseen sites) is due to the inter-site variability in intensity distributions, and to unique artefacts caused by different MR scanner models and acquisition parameters. To mitigate this site-dependency, often referred to as the scanner effect, we propose LOD-Brain, a 3D convolutional neural network with progressive levels-of-detail (LOD), able to segment brain data from any site. Coarser network levels are responsible for learning a robust anatomical prior helpful in identifying brain structures and their locations, while finer levels refine the model to handle site-specific intensity distributions and anatomical variations. We ensure robustness across sites by training the model on an unprecedentedly rich dataset aggregating data from open repositories: almost 27,000 T1w volumes from around 160 acquisition sites, at 1.5 - 3T, from a population spanning from 8 to 90 years old. Extensive tests demonstrate that LOD-Brain produces state-of-the-art results, with no significant difference in performance between internal and external sites, and robust to challenging anatomical variations. Its portability paves the way for large-scale applications across different healthcare institutions, patient populations, and imaging technology manufacturers. Code, model, and demo are available on the project website.

Hierarchical AI enables global interpretation of culture plates in the era of digital microbiology.

Full Laboratory Automation is revolutionizing work habits in an increasing number of clinical microbiology facilities worldwide, generating huge streams of digital images for interpretation. Contextually, deep learning architectures are leading to paradigm shifts in the way computers can assist with difficult visual interpretation tasks in several domains. At the crossroads of these epochal trends, we present a system able to tackle a core task in clinical microbiology, namely the global interpretation of diagnostic bacterial culture plates, including presumptive pathogen identification. This is achieved by decomposing the problem into a hierarchy of complex subtasks and addressing them with a multi-network architecture we call DeepColony. Working on a large stream of clinical data and a complete set of 32 pathogens, the proposed system is capable of effectively assist plate interpretation with a surprising degree of accuracy in the widespread and demanding framework of Urinary Tract Infections. Moreover, thanks to the rich species-related generated information, DeepColony can be used for developing trustworthy clinical decision support services in laboratory automation ecosystems from local to global scale.

CineScale2: a dataset of cinematic camera features in movies.

The position and orientation of the camera in relation to the subject(s) in a movie scene, namely camera "level" and camera "angle", are essential features in the film-making process due to their influence on the viewer's perception of the scene. We provide a database containing camera feature annotations on camera angle and camera level, for about 25,000 image frames. Frames are sampled from a wide range of movies, freely available images, and shots from cinematographic websites, and are annotated on the following five categories - Overhead, High, Neutral, Low, and Dutch - for what concerns camera angle, and on six different classes of camera level: Aerial, Eye, Shoulder, Hip, Knee, and Ground level. This dataset is an extension of the Cinescale dataset [1], which contains movie frames and related annotations regarding shot scale. The CineScale2 database enables AI-driven interpretation of shot scale data and opens to a large set of research activities related to the automatic visual analysis of cinematic material, such as movie stylistic analysis, video recommendation, and media psychology. To these purposes, we also provide the model and the code for building a Convolutional Neural Network (CNN) architecture for automated camera feature recognition. All the material is provided on the the project website; video frames can be also provided upon requests to authors, for research purposes under fair use.

Deep learning to detect significant coronary artery disease from plain chest radiographs AI4CAD.

BACKGROUND: The predictive role of chest radiographs in patients with suspected coronary artery disease (CAD) is underestimated and may benefit from artificial intelligence (AI) applications. OBJECTIVES: To train, test, and validate a deep learning (DL) solution for detecting significant CAD based on chest radiographs. METHODS: Data of patients referred for angina and undergoing chest radiography and coronary angiography were analysed retrospectively. A deep convolutional neural network (DCNN) was designed to detect significant CAD from posteroanterior/anteroposterior chest radiographs. The DCNN was trained for severe CAD binary classification (absence/presence). Coronary angiography reports were the ground truth. Stenosis severity of ≥70% for non-left main vessels and ≥ 50% for left main defined severe CAD. RESULTS: Information of 7728 patients was reviewed. Severe CAD was present in 4091 (53%). Patients were randomly divided for algorithm training (70%; n = 5454) and fine-tuning/model validation (10%; n = 773). Internal clinical validation (model testing) was performed with the remaining patients (20%; n = 1501). At binary logistic regression, DCNN prediction was the strongest severe CAD predictor (p < 0.0001; OR: 1.040; CI: 1.032-1.048). Using a high sensitivity operating cut-point, the DCNN had a sensitivity of 0.90 to detect significant CAD (specificity 0.31; AUC 0.73; 95% CI DeLong, 0.69-0.76). Adding to the AI chest radiograph interpretation angina status improved the prediction (AUC 0.77; 95% CI DeLong, 0.74-0.80). CONCLUSION: AI-read chest radiographs could be used to pre-test significant CAD probability in patients referred for suspected angina. Further studies are required to externally validate our algorithm, develop a clinically applicable tool, and support CAD screening in broader settings.

Assessing Trustworthy AI in Times of COVID-19: Deep Learning for Predicting a Multiregional Score Conveying the Degree of Lung Compromise in COVID-19 Patients.

This article's main contributions are twofold: 1) to demonstrate how to apply the general European Union's High-Level Expert Group's (EU HLEG) guidelines for trustworthy AI in practice for the domain of healthcare and 2) to investigate the research question of what does "trustworthy AI" mean at the time of the COVID-19 pandemic. To this end, we present the results of a post-hoc self-assessment to evaluate the trustworthiness of an AI system for predicting a multiregional score conveying the degree of lung compromise in COVID-19 patients, developed and verified by an interdisciplinary team with members from academia, public hospitals, and industry in time of pandemic. The AI system aims to help radiologists to estimate and communicate the severity of damage in a patient's lung from Chest X-rays. It has been experimentally deployed in the radiology department of the ASST Spedali Civili clinic in Brescia, Italy, since December 2020 during pandemic time. The methodology we have applied for our post-hoc assessment, called Z-Inspection®, uses sociotechnical scenarios to identify ethical, technical, and domain-specific issues in the use of the AI system in the context of the pandemic.

Development of a cadaveric head and neck cancer model and three-dimensional analysis of margins in surgical navigation-aided ablations.

INTRODUCTION: The adequacy of the surgical resection is the main controllable variable that is in the hands of the surgical team. There exists an unmet need to increase the rate of negative margins, particularly in cancers invading the craniofacial area. The study aimed 1) at developing a gross tumor model to be utilized for research, educational, and training purposes and 2) establishing the 3-dimensional relationship between the outer surface of the surgical specimen and tumor surface and test the effect of guiding ablations on cadavers with surgical navigation (SN). MATERIAL AND METHODS: Seven cadaver heads were employed to create 24 craniofacial tumor models. Simulation of tumor resections was performed by 8 surgeons. Fourteen and 10 resections were performed with and without SN-guidance, respectively. Gross specimens underwent computed tomography and 3-dimensional analysis through dedicated software. Task load was assessed through a validated questionnaire. Tumor model reliability was studied based on visual analogue scale rate by surgeons and radiologists. RESULTS: SN reduced the rate of margin involvement, particularly by decreasing the percentage of the gross specimen outer surface involvement in areas uncovered by normal bony structures. The workload of SN-aided ablations was found to be medium-to-somewhat-high. Tumor model reliability was deemed satisfactory except for the extension to bony structures. CONCLUSIONS: A gross tumor model for head and neck cancers involving the craniofacial area was developed and resulted satisfactorily reliable from both a surgical and radiologic standpoint. SN reduced the rate of margin involvement, particularly by improving delineation of bone-uncovered areas.

DenseMatch: a dataset for real-time 3D reconstruction.

We provide a database aimed at real-time quantitative analysis of 3D reconstruction and alignment methods, containing 3140 point clouds from 10 subjects/objects. These scenes are acquired with a high-resolution 3D scanner. It contains depth maps that produce point clouds with more than 500k points on average. This dataset is useful to develop new models and alignment strategies to automatically reconstruct 3D scenes from data acquired with optical scanners or benchmarking purposes.

Effects of remote digital monitoring on oral hygiene of orthodontic patients: a prospective study.

BACKGROUND: Remote digital monitoring during orthodontic treatment can help patients in improving their oral hygiene performance and reducing the number of appointments due to emergency reasons, especially in time of COVID-19 pandemic where non-urgent appointments might be discouraged. METHODS: Thirty patients scheduled to start an orthodontic treatment were divided into two groups of fifteen. Compared to controls, study group patients were provided with scan box and cheek retractor (Dental Monitoring®) and were instructed to take monthly intra-oral scans. Plaque Index (PI), Gingival Index (GI), and White Spot Lesions (WSL) were recorded for both groups at baseline (t0), every month for the first 3 months (t1, t2, t3), and at 6 months (t4). Carious Lesions Onset (CLO) and Emergency Appointments (EA) were also recorded during the observation period. Inter-group differences were assessed with Student's t test and Chi-square test, intra-group differences were assessed with Cochran's Q-test (significance α = 0.05). RESULTS: Study group patients showed a significant improvement in plaque control at t3 (p = 0.010) and t4 (p = 0.039), compared to control group. No significant difference was observed in the number of WSL between the two groups. No cavities were detected in the study group, while five CLO were diagnosed in the control group (p = 0.049). A decreased number of EA was observed in the study group, but the difference was not significant. CONCLUSIONS: Integration of a remote monitoring system during orthodontic treatment was effective in improving plaque control and reducing carious lesions onset. The present findings encourage orthodontists to consider this technology to help maintaining optimal oral health of patients, especially in times of health emergency crisis.

CineScale: A dataset of cinematic shot scale in movies.

We provide a database containing shot scale annotations (i.e., the apparent distance of the camera from the subject of a filmed scene) for more than 792,000 image frames. Frames belong to 124 full movies from the entire filmographies by 6 important directors: Martin Scorsese, Jean-Luc Godard, Béla Tarr, Federico Fellini, Michelangelo Antonioni, and Ingmar Bergman. Each frame, extracted from videos at 1 frame per second, is annotated on the following scale categories: Extreme Close Up (ECU), Close Up (CU), Medium Close Up (MCU), Medium Shot (MS), Medium Long Shot (MLS), Long Shot (LS), Extreme Long Shot (ELS), Foreground Shot (FS), and Insert Shots (IS). Two independent coders annotated all frames from the 124 movies, whilst a third one checked their coding and made decisions in cases of disagreement. The CineScale database enables AI-driven interpretation of shot scale data and opens to a large set of research activities related to the automatic visual analysis of cinematic material, such as the automatic recognition of the director's style, or the unfolding of the relationship between shot scale and the viewers' emotional experience. To these purposes, we also provide the model and the code for building a Convolutional Neural Network (CNN) architecture for automated shot scale recognition. All this material is provided through the project website, where video frames can also be requested to authors, for research purposes under fair use.

BS-Net: Learning COVID-19 pneumonia severity on a large chest X-ray dataset.

In this work we design an end-to-end deep learning architecture for predicting, on Chest X-rays images (CXR), a multi-regional score conveying the degree of lung compromise in COVID-19 patients. Such semi-quantitative scoring system, namely Brixia score, is applied in serial monitoring of such patients, showing significant prognostic value, in one of the hospitals that experienced one of the highest pandemic peaks in Italy. To solve such a challenging visual task, we adopt a weakly supervised learning strategy structured to handle different tasks (segmentation, spatial alignment, and score estimation) trained with a "from-the-part-to-the-whole" procedure involving different datasets. In particular, we exploit a clinical dataset of almost 5,000 CXR annotated images collected in the same hospital. Our BS-Net demonstrates self-attentive behavior and a high degree of accuracy in all processing stages. Through inter-rater agreement tests and a gold standard comparison, we show that our solution outperforms single human annotators in rating accuracy and consistency, thus supporting the possibility of using this tool in contexts of computer-assisted monitoring. Highly resolved (super-pixel level) explainability maps are also generated, with an original technique, to visually help the understanding of the network activity on the lung areas. We also consider other scores proposed in literature and provide a comparison with a recently proposed non-specific approach. We eventually test the performance robustness of our model on an assorted public COVID-19 dataset, for which we also provide Brixia score annotations, observing good direct generalization and fine-tuning capabilities that highlight the portability of BS-Net in other clinical settings. The CXR dataset along with the source code and the trained model are publicly released for research purposes.

Attitude towards Telemonitoring in Orthodontists and Orthodontic Patients.

The purpose of this study was to analyze the attitude of dentists and patients towards the use of Dental MonitoringTM (DM), an orthodontic telemonitoring software. Thus, two different specially prepared specific questionnaires were administered to 80 dentists (40 were general dentists and 40 orthodontists) and 80 orthodontic patients. All dentists judged positively telemonitoring, as 96.25% of them considered telemonitoring indicative of high tech and high-quality treatment; 100% considered it a way to reduce the number of in-office visits; 17.5% agreed on a weekly telemonitoring frequency, 40% on a biweekly, and 42.5% on a lower frequency. Further, 97.5% of patients judged positively telemonitoring; 81.25% of them considered telemonitoring indicative of high-tech treatment; 81.25% declared to be interested in reducing the number of in-office visits through telemonitoring; 27.5% agreed on taking self-picture every week, 57.5% every two weeks, and 15% on a lower frequency. Both patients and dentists positively judged telemonitoring, considering it a technologically advanced tool increasing the perception of quality and accuracy of the treatment. Both groups were interested in reducing the number of in-office visits, although not all of them revealed to be ready to invest more money and time in it.

Automatic classification of mice vocalizations using Machine Learning techniques and Convolutional Neural Networks.

Ultrasonic vocalizations (USVs) analysis is a well-recognized tool to investigate animal communication. It can be used for behavioral phenotyping of murine models of different disorders. The USVs are usually recorded with a microphone sensitive to ultrasound frequencies and they are analyzed by specific software. Different calls typologies exist, and each ultrasonic call can be manually classified, but the qualitative analysis is highly time-consuming. Considering this framework, in this work we proposed and evaluated a set of supervised learning methods for automatic USVs classification. This could represent a sustainable procedure to deeply analyze the ultrasonic communication, other than a standardized analysis. We used manually built datasets obtained by segmenting the USVs audio tracks analyzed with the Avisoft software, and then by labelling each of them into 10 representative classes. For the automatic classification task, we designed a Convolutional Neural Network that was trained receiving as input the spectrogram images associated to the segmented audio files. In addition, we also tested some other supervised learning algorithms, such as Support Vector Machine, Random Forest and Multilayer Perceptrons, exploiting informative numerical features extracted from the spectrograms. The performance showed how considering the whole time/frequency information of the spectrogram leads to significantly higher performance than considering a subset of numerical features. In the authors' opinion, the experimental results may represent a valuable benchmark for future work in this research field.

Spatial-Spectral Evidence of Glare Influence on Hyperspectral Acquisitions.

Glare is an unwanted optical phenomenon which affects imaging systems with optics. This paper presents for the first time a set of hyperspectral image (HSI) acquisitions and measurements to verify how glare affects acquired HSI data in standard conditions. We acquired two ColorCheckers (CCs) in three different lighting conditions, with different backgrounds, different exposure times, and different orientations. The reflectance spectra obtained from the imaging system have been compared to pointwise reference measures obtained with contact spectrophotometers. To assess and identify the influence of glare, we present the Glare Effect (GE) index, which compares the contrast of the grayscale patches of the CC in the hyperspectral images with the contrast of the reference spectra of the same patches. We evaluate, in both spatial and spectral domains, the amount of glare affecting every hyperspectral image in each acquisition scenario, clearly evidencing an unwanted light contribution to the reflectance spectra of each point, which increases especially for darker pixels and pixels close to light sources or bright patches.

CEREBRUM: a fast and fully-volumetric Convolutional Encoder-decodeR for weakly-supervised sEgmentation of BRain strUctures from out-of-the-scanner MRI.

Many functional and structural neuroimaging studies call for accurate morphometric segmentation of different brain structures starting from image intensity values of MRI scans. Current automatic (multi-) atlas-based segmentation strategies often lack accuracy on difficult-to-segment brain structures and, since these methods rely on atlas-to-scan alignment, they may take long processing times. Alternatively, recent methods deploying solutions based on Convolutional Neural Networks (CNNs) are enabling the direct analysis of out-of-the-scanner data. However, current CNN-based solutions partition the test volume into 2D or 3D patches, which are processed independently. This process entails a loss of global contextual information, thereby negatively impacting the segmentation accuracy. In this work, we design and test an optimised end-to-end CNN architecture that makes the exploitation of global spatial information computationally tractable, allowing to process a whole MRI volume at once. We adopt a weakly supervised learning strategy by exploiting a large dataset composed of 947 out-of-the-scanner (3 Tesla T1-weighted 1mm isotropic MP-RAGE 3D sequences) MR Images. The resulting model is able to produce accurate multi-structure segmentation results in only a few seconds. Different quantitative measures demonstrate an improved accuracy of our solution when compared to state-of-the-art techniques. Moreover, through a randomised survey involving expert neuroscientists, we show that subjective judgements favour our solution with respect to widely adopted atlas-based software.

Transfer learning of deep neural network representations for fMRI decoding.

BACKGROUND: Deep neural networks have revolutionised machine learning, with unparalleled performance in object classification. However, in brain imaging (e.g., fMRI), the direct application of Convolutional Neural Networks (CNN) to decoding subject states or perception from imaging data seems impractical given the scarcity of available data. NEW METHOD: In this work we propose a robust method to transfer information from deep learning (DL) features to brain fMRI data with the goal of decoding. By adopting Reduced Rank Regression with Ridge Regularisation we establish a multivariate link between imaging data and the fully connected layer (fc7) of a CNN. We exploit the reconstructed fc7 features by performing an object image classification task on two datasets: one of the largest fMRI databases, taken from different scanners from more than two hundred subjects watching different movie clips, and another with fMRI data taken while watching static images. RESULTS: The fc7 features could be significantly reconstructed from the imaging data, and led to significant decoding performance. COMPARISON WITH EXISTING METHODS: The decoding based on reconstructed fc7 outperformed the decoding based on imaging data alone. CONCLUSION: In this work we show how to improve fMRI-based decoding benefiting from the mapping between functional data and CNN features. The potential advantage of the proposed method is twofold: the extraction of stimuli representations by means of an automatic procedure (unsupervised) and the embedding of high-dimensional neuroimaging data onto a space designed for visual object discrimination, leading to a more manageable space from dimensionality point of view.

Deep Learning Meets Hyperspectral Image Analysis: A Multidisciplinary Review.

Modern hyperspectral imaging systems produce huge datasets potentially conveying a great abundance of information; such a resource, however, poses many challenges in the analysis and interpretation of these data. Deep learning approaches certainly offer a great variety of opportunities for solving classical imaging tasks and also for approaching new stimulating problems in the spatial-spectral domain. This is fundamental in the driving sector of Remote Sensing where hyperspectral technology was born and has mostly developed, but it is perhaps even more true in the multitude of current and evolving application sectors that involve these imaging technologies. The present review develops on two fronts: on the one hand, it is aimed at domain professionals who want to have an updated overview on how hyperspectral acquisition techniques can combine with deep learning architectures to solve specific tasks in different application fields. On the other hand, we want to target the machine learning and computer vision experts by giving them a picture of how deep learning technologies are applied to hyperspectral data from a multidisciplinary perspective. The presence of these two viewpoints and the inclusion of application fields other than Remote Sensing are the original contributions of this review, which also highlights some potentialities and critical issues related to the observed development trends.

Automatic hemolysis identification on aligned dual-lighting images of cultured blood agar plates.

BACKGROUND AND OBJECTIVE: The recent introduction of Full Laboratory Automation systems in clinical microbiology opens to the availability of streams of high definition images representing bacteria culturing plates. This creates new opportunities to support diagnostic decisions through image analysis and interpretation solutions, with an expected high impact on the efficiency of the laboratory workflow and related quality implications. Starting from images acquired under different illumination settings (top-light and back-light), the objective of this work is to design and evaluate a method for the detection and classification of diagnostically relevant hemolysis effects associated with specific bacteria growing on blood agar plates. The presence of hemolysis is an important factor to assess the virulence of pathogens, and is a fundamental sign of the presence of certain types of bacteria. METHODS: We introduce a two-stage approach. Firstly, the implementation of a highly accurate alignment of same-plate image scans, acquired using top-light and back-light illumination, enables the joint spatially coherent exploitation of the available data. Secondly, from each segmented portion of the image containing at least one bacterial colony, specifically designed image features are extracted to feed a SVM classification system, allowing detection and discrimination among different types of hemolysis. RESULTS: The fine alignment solution aligns more than 98.1% images with a residual error of less than 0.13 mm. The hemolysis classification block achieves a 88.3% precision with a recall of 98.6%. CONCLUSIONS: The results collected from different clinical scenarios (urinary infections and throat swab screening) together with accurate error analysis demonstrate the suitability of our system for robust hemolysis detection and classification, which remains feasible even in challenging conditions (low contrast or illumination changes).

Mesh Denoising with (Geo)Metric Fidelity.

Working with noisy meshes and aiming at providing high-fidelity 3D object models without tampering the metric quality of the acquisitions, we propose a mesh denoising technique that, through a normal-diffusion process guided by a curvature saliency map, is able to preserve and emphasize the natural object features, concurrently allowing the introduction of a bound on the maximum distance from the original model. Moreover, both the position of the mesh vertices and the edge orientations are optimized through a tailored geometric-aliasing correction. Thanks to an efficiently parallelized procedure, we are able to process even large models almost instantly with a parameter configuration that does not depend on the scale of the object. An essential survey on mesh denoising is also presented which is functional to the definition of a common framework where to set up our solutions and the related technical and experimental comparisons. The proposed results prove the effectiveness of our method, especially on the challenging target application profiles. Where competing techniques tend to inappropriately recover sharp edges while deforming the surrounding geometry or, on the contrary, to oversmooth shallow features, our method protects and enhances the natural object features and effectively reduces scanning noise on the smooth parts, while guaranteeing the prescribed metric-fidelity to the input model.

Concept and Design of a 3D Printed Support to Assist Hand Scanning for the Realization of Customized Orthosis.

In the rehabilitation field, the use of additive manufacturing techniques to realize customized orthoses is increasingly widespread. Obtaining a 3D model for the 3D printing phase can be done following different methodologies. We consider the creation of personalized upper limb orthoses, also including fingers, starting from the acquisition of the hand geometry through accurate 3D scanning. However, hand scanning procedure presents differences between healthy subjects and patients affected by pathologies that compromise upper limb functionality. In this work, we present the concept and design of a 3D printed support to assist hand scanning of such patients. The device, realized with FDM additive manufacturing techniques in ABS material, allows palmar acquisitions, and its design and test are motivated by the following needs: (1) immobilizing the hand of patients during the palmar scanning to reduce involuntary movements affecting the scanning quality and (2) keeping hands open and in a correct position, especially to contrast the high degree of hypertonicity of spastic subjects. The resulting device can be used indifferently for the right and the left hand; it is provided in four-dimensional sizes and may be also suitable as a palmar support for the acquisition of the dorsal side of the hand.

Hyperspectral image analysis for rapid and accurate discrimination of bacterial infections: A benchmark study.

With the rapid diffusion of Full Laboratory Automation systems, Clinical Microbiology is currently experiencing a new digital revolution. The ability to capture and process large amounts of visual data from microbiological specimen processing enables the definition of completely new objectives. These include the direct identification of pathogens growing on culturing plates, with expected improvements in rapid definition of the right treatment for patients affected by bacterial infections. In this framework, the synergies between light spectroscopy and image analysis, offered by hyperspectral imaging, are of prominent interest. This leads us to assess the feasibility of a reliable and rapid discrimination of pathogens through the classification of their spectral signatures extracted from hyperspectral image acquisitions of bacteria colonies growing on blood agar plates. We designed and implemented the whole data acquisition and processing pipeline and performed a comprehensive comparison among 40 combinations of different data preprocessing and classification techniques. High discrimination performance has been achieved also thanks to improved colony segmentation and spectral signature extraction. Experimental results reveal the high accuracy and suitability of the proposed approach, driving the selection of most suitable and scalable classification pipelines and stimulating clinical validations.