Machine learning framework incorporating expert knowledge in tissue image annotation

Kromp, Florian; Ambros, Inge M.; Weiss, Tamara; Bogen, Dominik; Dodig, Helena; Berneder, Maria; Gerber, Teresa; Taschner‐Mandl, Sabine; Ambros, Peter F.; Hanbury, Allan

doi:10.1109/icpr.2016.7899657

Cited by 7 publications

(5 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…c). Since determination of SC purity by manual counting is time consuming, we used a novel software developed by our group (Kromp et al, ) which measures single cell features of IF‐images after semi‐automated image acquisition. Only 500–1,000 cells are required for Image Scatter‐Plot (ISP) analysis, which visualizes cells according to their cellular features.…”

Section: Resultsmentioning

confidence: 99%

See 1 more Smart Citation

Proteomics and transcriptomics of peripheral nerve tissue and cells unravel new aspects of the human Schwann cell repair phenotype

Weiss

Taschner‐Mandl

Bileck

et al. 2016

Glia

Self Cite

128

View full text Add to dashboard Cite

The remarkable feature of Schwann cells (SCs) to transform into a repair phenotype turned the spotlight on this powerful cell type. SCs provide the regenerative environment for axonal re-growth after peripheral nerve injury (PNI) and play a vital role in differentiation of neuroblastic tumors into a benign subtype of neuroblastoma, a tumor originating from neural crestderived neuroblasts. Hence, understanding their mode-of-action is of utmost interest for new approaches in regenerative medicine, but also for neuroblastoma therapy. However, literature on human SCs is scarce and it is unknown to which extent human SC cultures reflect the SC repair phenotype developing after PNI in patients. We performed high-resolution proteome profiling and RNA-sequencing on highly enriched human SC and fibroblast cultures, control and ex vivo degenerated nerve explants to identify novel molecules and functional processes active in repair SCs. In fact, we found cultured SCs and degenerated nerves to share a similar repair SC-associated expression signature, including the upregulation of JUN, as well as two prominent functions, i.e., myelin debris clearance and antigen presentation via MHCII. In addition to myelin degradation, cultured SCs were capable of actively taking up cell-extrinsic components in functional phagocytosis and co-cultivation assays. Moreover, in cultured SCs and degenerated nerve tissue MHCII was upregulated at the cellular level along with high expression of chemoattractants and co-inhibitory rather than -stimulatory molecules. These results demonstrate human SC cultures to execute an inherent program of nerve repair and support two novel repair SC functions, debris clearance via phagocytosisrelated mechanisms and type II immune-regulation.

show abstract

Section: Resultsmentioning

confidence: 99%

“…The visualization of single cells (cropped nuclear images) arranged according to chosen features in a two‐dimensional plot is provided as and referred to as Image Scatter‐Plot (ISP). Hierarchical gating strategies comparable to flow cytometry analysis were applied to the ISP (Kromp et al, ).…”

Section: Methodsmentioning

confidence: 99%

Proteomics and transcriptomics of peripheral nerve tissue and cells unravel new aspects of the human Schwann cell repair phenotype

Weiss

Taschner‐Mandl

Bileck

et al. 2016

Glia

Self Cite

128

View full text Add to dashboard Cite

show abstract

“…ogists trained by a disease expert. To accelerate the time consuming process of image annotation, a machine learning-based framework (MLF) was utilized supporting the process of annotation by learning characteristics of annotation in multiple steps 24 . The MLF annotations result in a coarse annotation of nuclear contours and have to be refined to serve as ground truth annotation.…”

Section: Ground Truth Annotation Nuclei Image Annotation Was Performmentioning

confidence: 99%

An annotated fluorescence image dataset for training nuclear segmentation methods

et al. 2020

Self Cite

View full text Add to dashboard Cite

Fully-automated nuclear image segmentation is the prerequisite to ensure statistically significant, quantitative analyses of tissue preparations,applied in digital pathology or quantitative microscopy. The design of segmentation methods that work independently of the tissue type or preparation is complex, due to variations in nuclear morphology, staining intensity, cell density and nuclei aggregations. Machine learning-based segmentation methods can overcome these challenges, however high quality expert-annotated images are required for training. Currently, the limited number of annotated fluorescence image datasets publicly available do not cover a broad range of tissues and preparations. We present a comprehensive, annotated dataset including tightly aggregated nuclei of multiple tissues for the training of machine learning-based nuclear segmentation algorithms. The proposed dataset covers sample preparation methods frequently used in quantitative immunofluorescence microscopy. We demonstrate the heterogeneity of the dataset with respect to multiple parameters such as magnification, modality, signal-to-noise ratio and diagnosis. Based on a suggested split into training and test sets and additional single-nuclei expert annotations, machine learning-based image segmentation methods can be trained and evaluated.

show abstract

“…HaCaT and NB samples were acquired with a 63x magnifying objective whereas ganglioneuroma samples were acquired using a 10x magnifying objective. Trained undergraduate students created nuclear label masks using a recently developed machine learning framework [27]. They annotated all images except two.…”

Section: Dataset Descriptionmentioning

confidence: 99%

Evaluation of Deep Learning Architectures for Complex Immunofluorescence Nuclear Image Segmentation

Kromp

Fischer

Bozsaky

et al. 2021

IEEE Trans. Med. Imaging

Self Cite

View full text Add to dashboard Cite

Separating and labeling each nuclear instance (instance-aware segmentation) is the key challenge in nuclear image segmentation. Deep Convolutional Neural Networks have been demonstrated to solve nuclear image segmentation tasks across different imaging modalities, but a systematic comparison on complex immunofluorescence images has not been performed. Deep learning based segmentation requires annotated datasets for training, but annotated fluorescence nuclear image datasets are rare and of limited size and complexity. In this work, we evaluate and compare the segmentation effectiveness of multiple deep learning architectures (U-Net, U-Net ResNet, Cellpose, Mask R-CNN, KG instance segmentation) and two conventional algorithms (Iterative h-min based watershed, Attributed relational graphs) on complex fluorescence nuclear images of various types.We propose and evaluate a novel strategy to create artificial images to extend the training set. Results show that instanceaware segmentation architectures and Cellpose outperform the U-Net architectures and conventional methods on complex images in terms of F1 scores, while the U-Net architectures achieve overall higher mean Dice scores. Training with additional artificially generated images improves recall and F1 scores for complex images, thereby leading to top F1 scores for three out of five sample preparation types. Mask R-CNN trained on artificial images achieves the overall highest F1 score on complex images of similar conditions to the training set images while Cellpose achieves the overall highest F1 score on complex images of new imaging conditions. We provide quantitative results demonstrat-Manuscript resubmitted on February 13, 2021.

show abstract

Machine learning framework incorporating expert knowledge in tissue image annotation

Cited by 7 publications

References 14 publications

Proteomics and transcriptomics of peripheral nerve tissue and cells unravel new aspects of the human Schwann cell repair phenotype

Proteomics and transcriptomics of peripheral nerve tissue and cells unravel new aspects of the human Schwann cell repair phenotype

An annotated fluorescence image dataset for training nuclear segmentation methods

Evaluation of Deep Learning Architectures for Complex Immunofluorescence Nuclear Image Segmentation

Contact Info

Product

Resources

About