Dataset Information

Evaluating cellularity and structural connectivity on whole brain slides using a custom-made digital pathology pipeline.

ABSTRACT: AIMS:In brain research, the histopathological examination of coronar whole-brain slides provides important insights into spatial disease characteristics. Regarding brain tumor research, this enables visualization of tumor heterogeneity, infiltration patterns and the relationship with the surrounding brain parenchyma. The precise correlation between radiological imaging and post-mortem brains is of special interest. NEW METHOD:We developed a wide-field slide scanner, comprising a microscope, a high-precision remotely controllable x-y-stage, a camera and a computer workstation, for automatically scanning uncommonly large formats. We analyzed whole brain slides of three patients and constructed cellularity heatmaps and fiber tract maps using a custom-made image processing pipeline. RESULTS:The obtained cellularity heatmaps allow for distinguishing compact tumor (5714?±?1786 cells/mm², mean?±?standard deviation) from white matter (3581?±?828 cells/mm²) and grey matter (2473?±?716 cells/mm²). Compared to magnetic resonance imaging, the proposed histopathological work-up (i) reveals a larger zone of tumor infiltration around the compact tumor areas and (ii) shows how pre-existing tracts are displaced by the tumor bulk. Moreover, we highlight differences in the histological tumor growth pattern of two different radiological progression subtypes. COMPARISON WITH EXISTING METHOD(S):Compared to existing (commercial) solutions, our slide scanning solution is adaptable and cost-efficient. Moreover, we showcase potential clinical applications by mapping whole brain histology to magnetic resonance imaging. CONCLUSIONS:We herein provide instructions on how to (i) construct a custom-built slide scanner capable of scanning arbitrary slide formats, (ii) automatically evaluate the cell density and (iii) perform fiber tracking on whole brain slides.

SUBMITTER: Roetzer T

PROVIDER: S-EPMC6269083 | biostudies-other | 2019 Jan

REPOSITORIES: biostudies-other

ACCESS DATA

Publications

Evaluating cellularity and structural connectivity on whole brain slides using a custom-made digital pathology pipeline.

Roetzer Thomas T Leskovar Konrad K Peter Nadine N Furtner Julia J Muck Martina M Augustin Marco M Lichtenegger Antonia A Nowosielski Martha M Hainfellner Johannes A JA Baumann Bernhard B Woehrer Adelheid A

Journal of neuroscience methods 20181024

<h4>Aims</h4>In brain research, the histopathological examination of coronar whole-brain slides provides important insights into spatial disease characteristics. Regarding brain tumor research, this enables visualization of tumor heterogeneity, infiltration patterns and the relationship with the surrounding brain parenchyma. The precise correlation between radiological imaging and post-mortem brains is of special interest.<h4>New method</h4>We developed a wide-field slide scanner, comprising a m ...[more]

PMID: 30365968

Similar Datasets

Project description:We sought to evaluate visual performance and satisfaction with ready-made spectacles (RMS) in Chinese school-aged children with uncorrected refractive error.Randomized, double-blind, clinical trial.Junior high school students from urban Guangzhou, China, aged approximately 12 to 15 years with > or =1 diopter (D) of uncorrected spherical equivalent (SE) refractive error. Students were excluded with > or =2.00 D astigmatism, > or =2 D myopic anisometropia, and > or =1 D hyperopic anisometropia and ocular disease affecting vision.Refractive error was determined by cycloplegic subjective refraction. Students were randomly assigned to receive RMS or custom spectacles (CS) and assessed after 1 month of use. We required 175 students to complete in each arm to be able to measure a 15% difference in compliance.Compliance with spectacles lens wear, patterns of use, vision, symptoms, and perceived value.Screening identified 965 of 4607 (20.9%) students with reduced distance vision; 212 of the 965 (22.0%) refused evaluation and 187 of the 965 (20.8%) had <1 D of SE refractive error. Sixty-one (6.3%) were referred for further evaluation and the remaining 495 (51.3%) participated. Social, demographic, and ocular parameters were similar in the 2 groups. Average SE refractive error was -2.57+/-1.31 (mean value +/- standard deviation [SD]). Spectacle vision (Snellen acuity, mean +/- SD) was worse with RMS in the eye with lower SE (20/25(-0.5)+/-0.9 lines vs 20/25(+1)+/-0.7 lines; P = 0.004) and higher SE (20/25(-2)+/-1.2 lines vs 20/25(+1)+/-0.8; P<0.001). There were no differences (P>0.05) in the rate of use (94.3% vs 92.2%), wearing to the 1-month visit (46.9% vs 51.5%), planned use (93.3% vs 93.7%), value (89.5% vs 91.7% "moderate or high value or most valued possession"), or symptoms (blur, 21.1% vs 19.4% [P = 0.8] and other symptoms [P>0.2]).Although visual acuity was better with CS, no difference was found in acceptability in this population of students with predominantly simple myopic refractive error. This study supports the use of RMS in a school-based refractive services program, saving costs and improving the logistics of service delivery.

Project description:For a method to be widely adopted in medical research or clinical practice, it needs to be reproducible so that clinicians and regulators can have confidence in its use. Machine learning and deep learning have a particular set of challenges around reproducibility. Small differences in the settings or the data used for training a model can lead to large differences in the outcomes of experiments. In this work, three top-performing algorithms from the Camelyon grand challenges are reproduced using only information presented in the associated papers and the results are then compared to those reported. Seemingly minor details were found to be critical to performance and yet their importance is difficult to appreciate until the actual reproduction is attempted. We observed that authors generally describe the key technical aspects of their models well but fail to maintain the same reporting standards when it comes to data preprocessing which is essential to reproducibility. As an important contribution of the present study and its findings, we introduce a reproducibility checklist that tabulates information that needs to be reported in histopathology ML-based work in order to make it reproducible. Author summary For a method to be used a lot in medical research or clinical practice, it needs to be able to be reproduced so that people can trust it. Machine learning and deep learning have some challenges around this. For example, small changes in how a model is trained can lead to significant changes in the results of experiments. This makes it essential that researchers report how they do things in enough detail for the results of their experiments to be reproducible. In this work, we looked at three different algorithms used for digital pathology image analysis. We tried to reproduce them using only the information reported in their papers. We confirmed that even minor details could be essential. Authors often do not report all the details needed to reproduce their work. We also created a checklist of things that need to be reported to help other researchers make their work reproducible.

Dataset Information

Evaluating cellularity and structural connectivity on whole brain slides using a custom-made digital pathology pipeline.

Publications

Evaluating cellularity and structural connectivity on whole brain slides using a custom-made digital pathology pipeline.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets