On the Composition and Limitations of Publicly Available COVID-19 X-Ray Imaging Datasets

E-print/Working paper (2020)

 Machine learning based methods for diagnosis and progression prediction of COVID-19 from imaging data have gained significant attention in the last months, in particular by the use of deep learning ... [more ▼]

 Machine learning based methods for diagnosis and progression prediction of COVID-19 from imaging data have gained significant attention in the last months, in particular by the use of deep learning models. In this context hundreds of models where proposed with the majority of them trained on public datasets. Data scarcity, mismatch between training and target population, group imbalance, and lack of documentation are important sources of bias, hindering the applicability of these models to real-world clinical practice. Considering that datasets are an essential part of model building and evaluation, a deeper understanding of the current landscape is needed. This paper presents an overview of the currently public available COVID-19 chest X-ray datasets. Each dataset is briefly described and potential strength, limitations and interactions between datasets are identified. In particular, some key properties of current datasets that could be potential sources of bias, impairing models trained on them are pointed out. These descriptions are useful for model building on those datasets, to choose the best dataset according the model goal, to take into account the specific limitations to avoid reporting overconfident benchmark results, and to discuss their impact on the generalisation capabilities in a specific clinical setting. [less ▲]

Assessing suppression strategies against epidemicoutbreaks like COVID-19: the SPQEIR model

E-print/Working paper (2020)

The current COVID-19 outbreak represents a most serious challenge for societies worldwide. It isendangering the health of millions of people, and resulting in severe socioeconomic challenges dueto lock ... [more ▼]

The current COVID-19 outbreak represents a most serious challenge for societies worldwide. It isendangering the health of millions of people, and resulting in severe socioeconomic challenges dueto lock-down measures. Governments worldwide aim to devise exit strategies to revive the economywhile keeping the pandemic under control. The problem is that the effects of distinct measures arenot well quantified. This paper compares several suppression approaches and potential exit strategiesusing a new extended epidemic SEIR model. It concludes that while rapid and strong lock-down isan effective pandemic suppression measure, a combination of other strategies such as social distanc-ing, active protection and removal can achieve similar suppression synergistically. This quantitativeunderstanding will support the establishment of mid- and long-term interventions. Finally, the paperprovides an online tool that allows researchers and decision makers to interactively simulate diversescenarios with our model. [less ▲]

Deep Learning Quality Control for High-Throughput High-Content Screening Microscopy Images

Poster (2019, October 10)

Automation of biological image analysis is essential to boost biomedical research. The study of complex diseases such as neurodegenerative diseases calls for big amounts of data to build models towards ... [more ▼]

Automation of biological image analysis is essential to boost biomedical research. The study of complex diseases such as neurodegenerative diseases calls for big amounts of data to build models towards precision medicine. Such data acquisition is feasible in the context of high-throughput high-content screening (HTHCS) in which the quality of the results relays on the accuracy of image analysis. Deep learning (DL) yields great performance in image analysis tasks especially with big amounts of data such as the produced in HTHCS contexts. Such DL and HTHCS strength is also their biggest weakness since DL solutions are highly sensitive to bad quality datasets. Hence, accurate Quality Control (QC) for microscopy HTHCS becomes an essential step to obtain reliable pipelines for HTHCS analysis. Usually, artifacts found on these platforms are the consequence of out-of-focus and undesirable density variations. The importance of accurate outlier detection becomes essential for both the training process of generic ML solutions (i.e. segmentation or classification) and the QC of the input data such solution will predict on. Moreover, during the QC of the input dataset, we aim not only to discard unsuitable images but to report the user on the quality of its dataset giving the user the choice to keep or discard the bad images. To build the QC solution we employed fluorescent microscopy images of rosella biosensor generated in the HTHCS platform. A total of 15 planes ranging from -6z to +7z steps to the two optimum planes. We evaluated 27 known focus measure operators and concluded that they have low sensitivity in noisy conditions. We propose a CNN solution which predicts the focus error based on the distance to the optimal plane, outperforming the evaluated focus operators. This QC allows for better results in cell segmentation models based on U-Net architecture as well as promising improvements in image classification tasks. [less ▲]

PATH-29. POTENTIAL OF RAMAN SPECTROSCOPY IN ONCOLOGICAL NEUROSURGERY

Poster (2019)

Raman spectroscopy (RS) has gained increasing interest for the analysis of biological tissues within the recent years. It is a label-free, non-destructive method providing insights in biochemical ... [more ▼]

Raman spectroscopy (RS) has gained increasing interest for the analysis of biological tissues within the recent years. It is a label-free, non-destructive method providing insights in biochemical properties of tumor cells. It is possible to compare RS signals with histological properties of identical tissue parts. Therefore, RS bears promising potentials in neurosurgical neurooncology. On one hand, it could potentially be used for both intraoperative tumor diagnostics and resection control. On the other hand, it could provide important knowledge on tumor biochemistry and used for a subclassification of tumors with a potential impact on personalized therapy approaches. Within our group, we analyzed over 3000 measurement points in different brain tumors ex vivo with a robotized RS system and correlated the spectral curves with histopathological results. We separated and subclassified the data by AI-based methods. Additionally, we compared the latter results with those of a handheld probe, which is potentially navigatable for in vivo, intraoperative applications. We could demonstrate, that it is possible to separate distinct tumor groups only based on RS signals, especially by using computer-based signal analysis. Furthermore, we could demonstrate the differences of the spectra of deep-frozen and formalin-fixed tissues versus non-fixed tissues. Based on our results, we will highlight the potentials of RS for intraoperative neurosurgical application in resection control for brain tumors, as well as we will focus on the potentials for brain tumor diagnostics based purely on this method or by using it as an adjunct. Those methods bear additional potentials in the field of personalized chemotherapy approaches. [less ▲]

Using automated electrode localization to guide stimulation management in DBS

in Annals of Clinical and Translational Neurology (2018), 0(0),

Abstract Deep Brain Stimulation requires extensive postoperative testing of stimulation parameters to achieve optimal outcomes. Testing is typically not guided by neuroanatomical information on electrode ... [more ▼]

Abstract Deep Brain Stimulation requires extensive postoperative testing of stimulation parameters to achieve optimal outcomes. Testing is typically not guided by neuroanatomical information on electrode contact locations. To address this, we present an automated reconstruction of electrode locations relative to the treatment target, the subthalamic nucleus, comparing different targeting methods: atlas‐, manual‐, or tractography‐based subthalamic nucleus segmentation. We found that most electrode contacts chosen to deliver stimulation were closest or second closest to the atlas‐based subthalamic nucleus target. We suggest that information on each electrode contact's location, which can be obtained using atlas‐based methods, might guide clinicians during postoperative stimulation testing. [less ▲]

Lead-DBS v2: Towards a comprehensive pipeline for deep brain stimulation imaging.

in NeuroImage (2018)

Deep brain stimulation (DBS) is a highly efficacious treatment option for movement disorders and a growing number of other indications are investigated in clinical trials. To ensure optimal treatment ... [more ▼]

Deep brain stimulation (DBS) is a highly efficacious treatment option for movement disorders and a growing number of other indications are investigated in clinical trials. To ensure optimal treatment outcome, exact electrode placement is required. Moreover, to analyze the relationship between electrode location and clinical results, a precise reconstruction of electrode placement is required, posing specific challenges to the field of neuroimaging. Since 2014 the open source toolbox Lead-DBS is available, which aims at facilitating this process. The tool has since become a popular platform for DBS imaging. With support of a broad community of researchers worldwide, methods have been continuously updated and complemented by new tools for tasks such as multispectral nonlinear registration, structural/functional connectivity analyses, brain shift correction, reconstruction of microelectrode recordings and orientation detection of segmented DBS leads. The rapid development and emergence of these methods in DBS data analysis require us to revisit and revise the pipelines introduced in the original methods publication. Here we demonstrate the updated DBS and connectome pipelines of Lead-DBS using a single patient example with state-of-the-art high-field imaging as well as a retrospective cohort of patients scanned in a typical clinical setting at 1.5T. Imaging data of the 3T example patient is co-registered using five algorithms and nonlinearly warped into template space using ten approaches for comparative purposes. After reconstruction of DBS electrodes (which is possible using three methods and a specific refinement tool), the volume of tissue activated is calculated for two DBS settings using four distinct models and various parameters. Finally, four whole-brain tractography algorithms are applied to the patient's preoperative diffusion MRI data and structural as well as functional connectivity between the stimulation volume and other brain areas are estimated using a total of eight approaches and datasets. In addition, we demonstrate impact of selected preprocessing strategies on the retrospective sample of 51 PD patients. We compare the amount of variance in clinical improvement that can be explained by the computer model depending on the method of choice. This work represents a multi-institutional collaborative effort to develop a comprehensive, open source pipeline for DBS imaging and connectomics, which has already empowered several studies, and may facilitate a variety of future studies in the field. [less ▲]

Integration of sparse electrophysiological measurements with preoperative MRI using 3D surface estimation in deep brain stimulation surgery

in Webster, Robert; Fei, Baowei (Eds.) Medical Imaging 2017: Image-Guided Procedures, Robotic Interventions, and Modeling (2017, February 14)

Intraoperative microelectrode recordings (MER) have been used for several decades to guide neurosurgeons during the implantation of Deep Brain Stimulation (DBS) electrodes, especially when targeting the ... [more ▼]

Intraoperative microelectrode recordings (MER) have been used for several decades to guide neurosurgeons during the implantation of Deep Brain Stimulation (DBS) electrodes, especially when targeting the subthalamic nucleus (STN) to suppress the symptoms of Parkinson’s Disease. The standard approach is to use an array of up to five MER electrodes in a fixed configuration. Interpretation of the recorded signals yields a spatiallyvery sparse set of information about the morphology of the respective brain structures in the targeted area. However, no aid is currently available for surgeons to intraoperatively integrate this information with other data available on the patient’s individual morphology (e.g. MR imaging data used for surgical planning). This integration might allow surgeons to better determine the most probable position of the electrodes within the target structure during surgery. This paper suggests a method for reconstructing a surface patch from the sparse MER dataset utilizing additional a-priori knowledge about the geometrical configuration of the measurement electrodes. The conventional representation of MER measurements as intervals of target region/non-target region is therefore transformed into an equivalent boundary set representation, allowing efficient point-based calculations. Subsequently, the problem is to integrate the resulting patch with a preoperative model of the target structure, which can be formulated as registration problem minimizing a distance measure between the two surfaces. When restricting this registration procedure to translations, which is reasonable given certain geometric considerations, the problem can be solved globally by employing an exhaustive search with arbitrary precision in polynomial time. The proposed method is demonstrated using bilateral STN/Substantia Nigra segmentation data from preoperative MRIs of 17 Patients with simulated MER electrode placement. When using simulated data of heavily perturbed electrodes and subsequent MER measuremen [less ▲]

A solution for Multi-Alignment by Transformation Synchronisation

in The proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2015)

The alignment of a set of objects by means of transformations plays an important role in computer vision. Whilst the case for only two objects can be solved globally, when multiple objects are considered ... [more ▼]

The alignment of a set of objects by means of transformations plays an important role in computer vision. Whilst the case for only two objects can be solved globally, when multiple objects are considered usually iterative methods are used. In practice the iterative methods perform well if the relative transformations between any pair of objects are free of noise. However, if only noisy relative transformations are available (e.g. due to missing data or wrong correspondences) the iterative methods may fail. Based on the observation that the underlying noise-free transformations lie in the null space of a matrix that can directly be obtained from pairwise alignments, this paper presents a novel method for the synchronisation of pairwise transformations such that they are globally consistent. Simulations demonstrate that for a high amount of noise, a large proportion of missing data and even for wrong correspondence assignments the method delivers encouraging results. [less ▲]

Assessment of Electrode Displacement and Deformation with Respect to Pre-Operative Planning in Deep Brain Stimulation

in Handels, Heinz; Deserno, Thomas Martin; Meinzer, Hans-Peter (Eds.) et al Bildverarbeitung für die Medizin 2015 (2015)

The post-operative validation of deep brain stimulation electrode displacement and deformation is an important task towards improved DBS targeting. In this paper a method is proposed to align models of ... [more ▼]

The post-operative validation of deep brain stimulation electrode displacement and deformation is an important task towards improved DBS targeting. In this paper a method is proposed to align models of deep brain stimulation electrodes that are automatically extracted from post-operative CT imaging in a common coordinate system utilizing the planning data as reference. This enables the assessment of electrode displacement and deformation over the whole length of the trajectory with respect to the pre-operative planning. Accordingly, it enables the estimation of plan deviations in the surgical process as well as cross-patient statistics on electrode deformation, e.g. the bending induced by brain-shift. [less ▲]

Improving the Consistency of Manual Deep Brain Structure Segmentations by Combining Variational Interpolation, Simultaneous Multi-Modality Visualisation and Histogram Equilisation

in Biomedizinische Technik. Biomedical Engineering (2014), 59(1), 131-134