Exploring Negated Entites for Named Entity Recognition in Italian Lung Cancer Clinical Reports.

This paper explores the potential of leveraging electronic health records (EHRs) for personalized health research through the application of artificial intelligence (AI) techniques, specifically Named Entity Recognition (NER). By extracting crucial patient information from clinical texts, including diagnoses, medications, symptoms, and lab tests, AI facilitates the rapid identification of relevant data, paving the way for future care paradigms. The study focuses on Non-small cell lung cancer (NSCLC) in Italian clinical notes, introducing a novel set of 29 clinical entities that include both presence or absence (negation) of relevant information associated with NSCLC. Using a state-of-the-art model pretrained on Italian biomedical texts, we achieve promising results (average F1-score of 80.8%), demonstrating the feasibility of employing AI for extracting biomedical information in the Italian language.

GravityNet for end-to-end small lesion detection.

This paper introduces a novel one-stage end-to-end detector specifically designed to detect small lesions in medical images. Precise localization of small lesions presents challenges due to their appearance and the diverse contextual backgrounds in which they are found. To address this, our approach introduces a new type of pixel-based anchor that dynamically moves towards the targeted lesion for detection. We refer to this new architecture as GravityNet, and the novel anchors as gravity points since they appear to be "attracted" by the lesions. We conducted experiments on two well-established medical problems involving small lesions to evaluate the performance of the proposed approach: microcalcifications detection in digital mammograms and microaneurysms detection in digital fundus images. Our method demonstrates promising results in effectively detecting small lesions in these medical imaging tasks.

Learnable DoG convolutional filters for microcalcification detection.

Difference of Gaussians (DoG) convolutional filters are one of the earliest image processing methods employed for detecting microcalcifications on mammogram images before machine and deep learning methods became widespread. DoG is a blob enhancement filter that consists in subtracting one Gaussian-smoothed version of an image from another less Gaussian-smoothed version of the same image. Smoothing with a Gaussian kernel suppresses high-frequency spatial information, thus DoG can be regarded as a band-pass filter. However, due to their small size and overimposed breast tissue, microcalcifications vary greatly in contrast-to-noise ratio and sharpness. This makes it difficult to find a single DoG configuration that enhances all microcalcifications. In this work, we propose a convolutional network, named DoG-MCNet, where the first layer automatically learns a bank of DoG filters parameterized by their associated standard deviations. We experimentally show that when employed for microcalcification detection, our DoG layer acts as a learnable bank of band-pass preprocessing filters and improves detection performance by 4.86% AUFROC over baseline MCNet and 1.53% AUFROC over state-of-the-art multicontext ensemble of CNNs.

BigNeuron: a resource to benchmark and predict performance of algorithms for automated tracing of neurons in light microscopy datasets.

BigNeuron is an open community bench-testing platform with the goal of setting open standards for accurate and fast automatic neuron tracing. We gathered a diverse set of image volumes across several species that is representative of the data obtained in many neuroscience laboratories interested in neuron tracing. Here, we report generated gold standard manual annotations for a subset of the available imaging datasets and quantified tracing quality for 35 automatic tracing algorithms. The goal of generating such a hand-curated diverse dataset is to advance the development of tracing algorithms and enable generalizable benchmarking. Together with image quality features, we pooled the data in an interactive web application that enables users and developers to perform principal component analysis, t-distributed stochastic neighbor embedding, correlation and clustering, visualization of imaging and tracing data, and benchmarking of automatic tracing algorithms in user-defined data subsets. The image quality metrics explain most of the variance in the data, followed by neuromorphological features related to neuron size. We observed that diverse algorithms can provide complementary information to obtain accurate results and developed a method to iteratively combine methods and generate consensus reconstructions. The consensus trees obtained provide estimates of the neuron structure ground truth that typically outperform single algorithms in noisy datasets. However, specific algorithms may outperform the consensus tree strategy in specific imaging conditions. Finally, to aid users in predicting the most accurate automatic tracing results without manual annotations for comparison, we used support vector machine regression to predict reconstruction quality given an image volume and a set of automatic tracings.

Convolutional Networks and Transformers for Mammography Classification: An Experimental Study.

Convolutional Neural Networks (CNN) have received a large share of research in mammography image analysis due to their capability of extracting hierarchical features directly from raw data. Recently, Vision Transformers are emerging as viable alternative to CNNs in medical imaging, in some cases performing on par or better than their convolutional counterparts. In this work, we conduct an extensive experimental study to compare the most recent CNN and Vision Transformer architectures for whole mammograms classification. We selected, trained and tested 33 different models, 19 convolutional- and 14 transformer-based, on the largest publicly available mammography image database OMI-DB. We also performed an analysis of the performance at eight different image resolutions and considering all the individual lesion categories in isolation (masses, calcifications, focal asymmetries, architectural distortions). Our findings confirm the potential of visual transformers, which performed on par with traditional CNNs like ResNet, but at the same time show a superiority of modern convolutional networks like EfficientNet.

Addressing class imbalance in deep learning for small lesion detection on medical images.

Deep learning methods utilizing Convolutional Neural Networks (CNNs) have led to dramatic advances in automated understanding of medical images. However, in many medical image classification tasks, lesions occupy only a few pixels of the image. This results in a significant class imbalance between lesion and background. From recent literature, it is known that class imbalance may negatively affect the performance of CNN classification. However, very few research exists in the context of lesion detection. In this work, we propose a two-stage deep learning framework able to deal with the high class imbalance encountered during training of small lesion detectors. First, we train a deep cascade (DC) of long sequences of decision trees with an algorithm designed to handle unbalanced data that also drastically reduces the number of background samples reaching the final stage. The remaining samples are fed to a CNN, whose training benefits from both rebalance and hard mining done by the DC. We evaluated DC-CNN on two severely unbalanced classification problems: microcalcification detection and microaneurysm detection. In both cases, DC-CNN outperformed the CNNs trained with commonly used methods for addressing class imbalance such as oversampling, undersampling, hard mining, cost sensitive learning, and one-class classification. The DC-CNN was also ∼10x faster than CNN at test time.

Exploiting Multi-Level Parallelism for Stitching Very Large Microscopy Images.

Due to the limited field of view of the microscopes, acquisitions of macroscopic specimens require many parallel image stacks to cover the whole volume of interest. Overlapping regions are introduced among stacks in order to make it possible automatic alignment by means of a 3D stitching tool. Since state-of-the-art microscopes coupled with chemical clearing procedures can generate 3D images whose size exceeds the Terabyte, parallelization is required to keep stitching time within acceptable limits. In the present paper we discuss how multi-level parallelization reduces the execution times of TeraStitcher, a tool designed to deal with very large images. Two algorithms performing dataset partition for efficient parallelization in a transparent way are presented together with experimental results proving the effectiveness of the approach that achieves a speedup close to 300×, when both coarse- and fine-grained parallelism are exploited. Multi-level parallelization of TeraStitcher led to a significant reduction of processing times with no changes in the user interface, and with no additional effort required for the maintenance of code.

Improving the Automated Detection of Calcifications Using Adaptive Variance Stabilization.

In this paper, we analyze how stabilizing the variance of intensity-dependent quantum noise in digital mammograms can significantly improve the computerized detection of microcalcifications (MCs). These lesions appear on mammograms as tiny deposits of calcium smaller than 20 pixels in diameter. At this scale, high frequency image noise is dominated by quantum noise, which in raw mammograms can be described with a square-root noise model. Under this assumption, we derive an adaptive variance stabilizing transform (VST) that stabilizes the noise to unitary standard deviation in all the images. This is achieved by estimating the noise characteristics from the image at hand. We tested the adaptive VST as a preprocessing stage for four existing computerized MC detection methods on three data sets acquired with mammographic units from different manufacturers. In all the test cases considered, MC detection performance on transformed mammograms was statistically significantly higher than on unprocessed mammograms. Results were also superior in comparison with a "fixed" (nonparametric) VST previously proposed for digital mammograms.

Improving computer-aided detection assistance in breast cancer screening by removal of obviously false-positive findings.

PURPOSE: Computer-aided detection (CADe) systems for mammography screening still mark many false positives. This can cause radiologists to lose confidence in CADe, especially when many false positives are obviously not suspicious to them. In this study, we focus on obvious false positives generated by microcalcification detection algorithms. METHODS: We aim at reducing the number of obvious false-positive findings by adding an additional step in the detection method. In this step, a multiclass machine learning method is implemented in which dedicated classifiers learn to recognize the patterns of obvious false-positive subtypes that occur most frequently. The method is compared to a conventional two-class approach, where all false-positive subtypes are grouped together in one class, and to the baseline CADe system without the new false-positive removal step. The methods are evaluated on an independent dataset containing 1,542 screening examinations of which 80 examinations contain malignant microcalcifications. RESULTS: Analysis showed that the multiclass approach yielded a significantly higher sensitivity compared to the other two methods (P < 0.0002). At one obvious false positive per 100 images, the baseline CADe system detected 61% of the malignant examinations, while the systems with the two-class and multiclass false-positive reduction step detected 73% and 83%, respectively. CONCLUSIONS: Our study showed that by adding the proposed method to a CADe system, the number of obvious false positives can decrease significantly (P < 0.0002).

Bioimage Informatics for Big Data.

Bioimage informatics is a field wherein high-throughput image informatics methods are used to solve challenging scientific problems related to biology and medicine. When the image datasets become larger and more complicated, many conventional image analysis approaches are no longer applicable. Here, we discuss two critical challenges of large-scale bioimage informatics applications, namely, data accessibility and adaptive data analysis. We highlight case studies to show that these challenges can be tackled based on distributed image computing as well as machine learning of image examples in a multidimensional environment.

Label-free near-infrared reflectance microscopy as a complimentary tool for two-photon fluorescence brain imaging.

In vivo two-photon imaging combined with targeted fluorescent indicators is currently extensively used for attaining critical insights into brain functionality and structural plasticity. Additional information might be gained from back-scattered photons from the near-infrared (NIR) laser without introducing any exogenous labelling. Here, we describe a complimentary and versatile approach that, by collecting the reflected NIR light, provides structural details on axons and blood vessels in the brain, both in fixed samples and in live animals under a cranial window. Indeed, by combining NIR reflectance and two-photon imaging of a slice of hippocampus from a Thy1-GFPm mouse, we show the presence of randomly oriented axons intermingled with sparsely fluorescent neuronal processes. The back-scattered photons guide the contextualization of the fluorescence structure within brain atlas thanks to the recognition of characteristic hippocampal structures. Interestingly, NIR reflectance microscopy allowed the label-free detection of axonal elongations over the superficial layers of mouse cortex under a cranial window in vivo. Finally, blood flow can be measured in live preparations, thus validating label free NIR reflectance as a tool for monitoring hemodynamic fluctuations. The prospective versatility of this label-free technique complimentary to two-photon fluorescence microscopy is demonstrated in a mouse model of photothrombotic stroke in which the axonal degeneration and blood flow remodeling can be investigated.

Virtual finger boosts three-dimensional imaging and microsurgery as well as terabyte volume image visualization and analysis.

Three-dimensional (3D) bioimaging, visualization and data analysis are in strong need of powerful 3D exploration techniques. We develop virtual finger (VF) to generate 3D curves, points and regions-of-interest in the 3D space of a volumetric image with a single finger operation, such as a computer mouse stroke, or click or zoom from the 2D-projection plane of an image as visualized with a computer. VF provides efficient methods for acquisition, visualization and analysis of 3D images for roundworm, fruitfly, dragonfly, mouse, rat and human. Specifically, VF enables instant 3D optical zoom-in imaging, 3D free-form optical microsurgery, and 3D visualization and annotation of terabytes of whole-brain image volumes. VF also leads to orders of magnitude better efficiency of automated 3D reconstruction of neurons and similar biostructures over our previous systems. We use VF to generate from images of 1,107 Drosophila GAL4 lines a projectome of a Drosophila brain.

Extensible visualization and analysis for multidimensional images using Vaa3D.

Open-Source 3D Visualization-Assisted Analysis (Vaa3D) is a software platform for the visualization and analysis of large-scale multidimensional images. In this protocol we describe how to use several popular features of Vaa3D, including (i) multidimensional image visualization, (ii) 3D image object generation and quantitative measurement, (iii) 3D image comparison, fusion and management, (iv) visualization of heterogeneous images and respective surface objects and (v) extension of Vaa3D functions using its plug-in interface. We also briefly demonstrate how to integrate these functions for complicated applications of microscopic image visualization and quantitative analysis using three exemplar pipelines, including an automated pipeline for image filtering, segmentation and surface generation; an automated pipeline for 3D image stitching; and an automated pipeline for neuron morphology reconstruction, quantification and comparison. Once a user is familiar with Vaa3D, visualization usually runs in real time and analysis takes less than a few minutes for a simple data set.

Micron-scale resolution optical tomography of entire mouse brains with confocal light sheet microscopy.

Understanding the architecture of mammalian brain at single-cell resolution is one of the key issues of neuroscience. However, mapping neuronal soma and projections throughout the whole brain is still challenging for imaging and data management technologies. Indeed, macroscopic volumes need to be reconstructed with high resolution and contrast in a reasonable time, producing datasets in the TeraByte range. We recently demonstrated an optical method (confocal light sheet microscopy, CLSM) capable of obtaining micron-scale reconstruction of entire mouse brains labeled with enhanced green fluorescent protein (EGFP). Combining light sheet illumination and confocal detection, CLSM allows deep imaging inside macroscopic cleared specimens with high contrast and speed. Here we describe the complete experimental pipeline to obtain comprehensive and human-readable images of entire mouse brains labeled with fluorescent proteins. The clearing and the mounting procedures are described, together with the steps to perform an optical tomography on its whole volume by acquiring many parallel adjacent stacks. We showed the usage of open-source custom-made software tools enabling stitching of the multiple stacks and multi-resolution data navigation. Finally, we illustrated some example of brain maps: the cerebellum from an L7-GFP transgenic mouse, in which all Purkinje cells are selectively labeled, and the whole brain from a thy1-GFP-M mouse, characterized by a random sparse neuronal labeling.

TeraStitcher - a tool for fast automatic 3D-stitching of teravoxel-sized microscopy images.

BACKGROUND: Further advances in modern microscopy are leading to teravoxel-sized tiled 3D images at high resolution, thus increasing the dimension of the stitching problem of at least two orders of magnitude. The existing software solutions do not seem adequate to address the additional requirements arising from these datasets, such as the minimization of memory usage and the need to process just a small portion of data. RESULTS: We propose a free and fully automated 3D Stitching tool designed to match the special requirements coming out of teravoxel-sized tiled microscopy images that is able to stitch them in a reasonable time even on workstations with limited resources. The tool was tested on teravoxel-sized whole mouse brain images with micrometer resolution and it was also compared with the state-of-the-art stitching tools on megavoxel-sized publicy available datasets. This comparison confirmed that the solutions we adopted are suited for stitching very large images and also perform well on datasets with different characteristics. Indeed, some of the algorithms embedded in other stitching tools could be easily integrated in our framework if they turned out to be more effective on other classes of images. To this purpose, we designed a software architecture which separates the strategies that use efficiently memory resources from the algorithms which may depend on the characteristics of the acquired images. CONCLUSIONS: TeraStitcher is a free tool that enables the stitching of Teravoxel-sized tiled microscopy images even on workstations with relatively limited resources of memory (<8 GB) and processing power. It exploits the knowledge of approximate tile positions and uses ad-hoc strategies and algorithms designed for such very large datasets. The produced images can be saved into a multiresolution representation to be efficiently retrieved and processed. We provide TeraStitcher both as standalone application and as plugin of the free software Vaa3D.