Multi-Scale Deep Neural Network for Mitosis Detection in Histological Images

Kausar, Tasleem; Wang, Mingjiang; Wu, Boqian; Idrees, Muhammad; Kanwal, Benish

doi:10.1109/iciibms.2018.8549967

Cited by 8 publications

(4 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Furthermore, contextual aggregation is limited by the input dimensions, meaning that a receptive field size can only exceed the input dimensions if padded artificial input pixels are used (a technique usually referred to as the use of same padding), which do not contain contextual information. While reducing the original input dimensions can be used to focus on scale information (Kausar et al (2018), Li et al (2018)), the potential contextual information remains unchanged.…”

Section: The Receptive Field and The Field Of Viewmentioning

confidence: 99%

HookNet: Multi-resolution convolutional neural networks for semantic segmentation in histopathology whole-slide images

Rijthoven

Balkenhol

Siliņa

et al. 2021

Medical Image Analysis

121

View full text Add to dashboard Cite

Section: The Receptive Field and The Field Of Viewmentioning

confidence: 99%

HookNet: Multi-resolution convolutional neural networks for semantic segmentation in histopathology whole-slide images

Rijthoven

Balkenhol

Siliņa

et al. 2021

Medical Image Analysis

121

View full text Add to dashboard Cite

“…9 show the experimental results of our method and some other previously proposed methods on the 2014 MITOSIS testing set. With an F-score of 0.575, our method outperforms all other approaches, including: STRASBOURG [3], YILDIZ [3], MINES-CURIE-INSERM [3], CUHK [3], which were the 4 winners of the 2014 ICPR MITOS-ATYPIA challenge, MFF-CNN [18], DeepMitosis [20], MSSN [19], CasNN [17] with its two proposed versions, namely the 'single' version that employs only one classification network and the 'average' version that employs three different classification networks and fuses their results, as well as the different configurations of the SegMitos model [22].…”

Section: ) Quantitative Evaluationmentioning

confidence: 99%

“…The drawback of this method is that its two networks are trained separately, which may be an impediment to the integration of the system. MFF-CNN [18] design a multi-scale fused CNN for mitosis detection. The model comprises two multi-scale branches that fuse features across different layers.…”

Section: Related Workmentioning

confidence: 99%

PartMitosis: A Partially Supervised Deep Learning Framework for Mitosis Detection in Breast Cancer Histopathology Images

2020

View full text Add to dashboard Cite

Detection of mitotic tumor cells per tissue area is one of the critical markers of breast cancer prognosis. The aim of this paper is to develop a method for the automatic detection of mitotic figures from breast cancer histological slides using a partially supervised deep learning framework. Unlike the previous literature, which has focused on solving the problem of mitosis detection in the weakly annotated datasets using centroid pixel labels (weak labels) only without taking advantage of the available pixel-level labels (strong labels) of other datasets, in this paper, we design a novel partially supervised framework based on two parallel deep fully convolutional networks. One of them is trained using weak labels and the other is trained using strong labels, together with a weight transfer function. In the detection phase, we fuse the segmentation maps produced by the two networks to obtain the final mitosis detections. Our system exploits the available large sets of mitosis detection samples with mitosis centroid annotation, such as the 2014 ICPR dataset and the AMIDA13 dataset, and only a small set of samples with the annotation of all mitosis pixels, such as the 2012 ICPR dataset, to perform a more accurate mitosis detection on weakly labeled data. This enables us to outperform all previous mitosis detection systems by achieving F-scores of 0.575 and 0.698 on the 2014 ICPR dataset and the AMIDA13 dataset respectively.

show abstract

“…The designed FF-CNN model fused multi-level features from various layers and correspondingly a deep cascaded approach is applied to build an effective training dataset. Further, in [34] we have designed a multi-scale fully convolution (MFF-CNN) model where multi-level and multi-scale features are fused to detect mitosis from HPFs images.…”

Section: Related Workmentioning

confidence: 99%

SmallMitosis: Small Size Mitotic Cells Detection in Breast Histopathology Images

et al. 2021

Self Cite

View full text Add to dashboard Cite

Multi-Scale Deep Neural Network for Mitosis Detection in Histological Images

Cited by 8 publications

References 14 publications

HookNet: Multi-resolution convolutional neural networks for semantic segmentation in histopathology whole-slide images

HookNet: Multi-resolution convolutional neural networks for semantic segmentation in histopathology whole-slide images

PartMitosis: A Partially Supervised Deep Learning Framework for Mitosis Detection in Breast Cancer Histopathology Images

SmallMitosis: Small Size Mitotic Cells Detection in Breast Histopathology Images

Contact Info

Product

Resources

About