Automatic instance segmentation of mitochondria in electron microscopy data

Nightingale, Luke; Folter, Joost de; Spiers, Helen; Strange, Amy; Lm, Collinson; Jones, Martin L.

doi:10.1101/2021.05.24.444785

Cited by 11 publications

(5 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This method has been further developed since submission to the MitoEM challenge into an open-source model called MitoNet [34] . FCI (London) 11 [35] : Four separate convolutional neural networks were trained to predict mitochondria binary masks in MitoEM-H, mitochondria boundaries in MitoEM-H, mitochondria binary masks in MitoEM-R, and mitochondria boundaries in MitoEM-R, respectively. All networks share a common architecture based on a 5-level 3D U-Net [44] with Inception-like blocks [55] , 32 initial filters and a dropout rate of 0.3.…”

Section: Summary Of Segmentation Methodsmentioning

confidence: 99%

Current Progress and Challenges in Large-Scale 3D Mitochondria Instance Segmentation

Franco-Barranco,

Lin,

Jang

et al. 2023

IEEE Trans. Med. Imaging

Self Cite

View full text Add to dashboard Cite

In this paper, we present the results of the MitoEM challenge on mitochondria 3D instance segmentation from electron microscopy images, organized in conjunction with the IEEE-ISBI 2021 conference. Our benchmark dataset consists of two large-scale 3D volumes, one from human and one from rat cortex tissue, which are 1,986 times larger than previously used datasets. At the time of paper submission, 257 participants had registered for the challenge, 14 teams had submitted their results, and six teams participated in the challenge workshop. Here, we present eight top-performing approaches from the challenge participants, along with our own baseline strategies. Posterior to the challenge, annotation errors in the ground truth were corrected without altering the final ranking. Additionally, we present a retrospective evaluation of the scoring system which revealed that: 1) challenge metric was permissive with the false positive predictions; and 2) size-based grouping of instances did not correctly categorize mitochondria of interest. Thus, we propose a new scoring system that better reflects the correctness of the segmentation results. Although several of the top methods are compared favorably to our own baselines, substantial errors remain unsolved for mitochondria with challenging morphologies. Thus, the challenge remains open for submission and automatic evaluation, with all volumes available for download.

show abstract

Section: Summary Of Segmentation Methodsmentioning

confidence: 99%

Current Progress and Challenges in Large-Scale 3D Mitochondria Instance Segmentation

Franco-Barranco,

Lin,

Jang

et al. 2023

IEEE Trans. Med. Imaging

Self Cite

View full text Add to dashboard Cite

show abstract

“…You can find more detailed information about the method in the later work presented by the authors on [56]. • ABCS (FNL) 12 : Two simple 3D versions of the original U-Net architecture [31] were trained to simulate different fields of views with input sizes of 64 × 128 × 128 voxels and 64 × 256 × 256, respectively.…”

Section: A Participants' Methodsmentioning

confidence: 99%

Current Progress and Challenges in Large-scale 3D Mitochondria Instance Segmentation

Franco-Barranco¹,

Lin²,

Jang³

et al. 2023

Preprint

View full text Add to dashboard Cite

<p>In this paper, we present the results of the MitoEM challenge on mitochondria 3D instance segmentation from electron microscopy images, organized in conjunction with the IEEE-ISBI 2021 conference. Our benchmark dataset consists of two large-scale 3D volumes, one from human and one from rat cortex tissue, which are 3,600 times larger than previously used datasets. At the time of paper submission, 257 participants had registered for the challenge, 14 teams had submitted their results, and six teams participated in the challenge workshop. Here, we present eight top-performing approaches from the challenge participants, along with our own baseline strategies. Posterior to the challenge, annotation errors in the ground truth were corrected without altering the final ranking. Although several of the top methods are compared favorably to our own baselines, substantial errors remain unsolved for mitochondria with challenging morphologies. Thus, the challenge remains open for submission and automatic evaluation, with all volumes available for download. Additionally, we present a retrospective evaluation of the scoring system performed using TIMISE, our novel open-source evaluation toolbox, which revealed that (1) the challenge metric was permissive with the false positive predictions and (2) the size-based grouping of instances did not correctly categorize mitochondria of interest. Thus, we propose a new scoring system that better reflects the correctness of the segmentation results. </p>

show abstract

“…MitoEM-R MitoEM-H Wei [30] 0.521 0.605 Nightingale [20] 0.715 0.625 Li [14] 0.890 0.787 Chen [13] 0.917 0.82 STT-UNET (Ours) 0.958 0.849…”

Section: Methodsmentioning

confidence: 99%

Video Instance Segmentation via Multi-Scale Spatio-Temporal Split Attention Transformer

Thawakar

Narayan

Cao

et al. 2022

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Accurate 3D mitochondria instance segmentation in electron microscopy (EM) is a challenging problem and serves as a prerequisite to empirically analyze their distributions and morphology. Most existing approaches employ 3D convolutions to obtain representative features. However, these convolution-based approaches struggle to effectively capture long-range dependencies in the volume mitochondria data, due to their limited local receptive field. To address this, we propose a hybrid encoder-decoder framework based on a split spatio-temporal attention module that efficiently computes spatial and temporal selfattentions in parallel, which are later fused through a deformable convolution. Further, we introduce a semantic foreground-background adversarial loss during training that aids in delineating the region of mitochondria instances from the background clutter. Our extensive experiments on three benchmarks, Lucchi, MitoEM-R and MitoEM-H, reveal the benefits of the proposed contributions achieving state-of-theart results on all three datasets. Our code and models are available at https://github.com/OmkarThawakar/STT-UNET.

show abstract

Automatic instance segmentation of mitochondria in electron microscopy data

Cited by 11 publications

References 6 publications

Current Progress and Challenges in Large-Scale 3D Mitochondria Instance Segmentation

Current Progress and Challenges in Large-Scale 3D Mitochondria Instance Segmentation

Current Progress and Challenges in Large-scale 3D Mitochondria Instance Segmentation

Video Instance Segmentation via Multi-Scale Spatio-Temporal Split Attention Transformer

Contact Info

Product

Resources

About