Deep Learning Reveals Cancer Metastasis and Therapeutic Antibody Targeting in the Entire Body

Pan, Chenchen; Schoppe, Oliver; Parra-Damas, Arnaldo; Cai, Ruiyao; Todorov, Mihail Ivilinov; Gondi, Gabor; Neubeck, Bettina von; Böğürcü-Seidel, Nuray; Seidel, Sascha; Sleiman, Katia; Veltkamp, Christian; Förstera, Benjamín; Mai, Hongcheng; Rong, Zhouyi; Trompak, Omelyan; Ghasemigharagoz, Alireza; Reimer, Madita Alice; Cuesta, Ángel M.; Coronel, Javier; Jeremias, Irmela; Saur, Dieter; Acker‐Palmer, Amparo; Acker, Till; Garvalov, Boyan K.; Menze, Bjoern; Zeidler, Reinhard; Ertürk, Ali

doi:10.1016/j.cell.2019.11.013

Cited by 164 publications

(151 citation statements)

References 58 publications

Supporting

Mentioning

148

Contrasting

Order By: Relevance

“…AIMOS as an enabling resource will be helpful in many areas of biomedical research, including tumor research [15][16][17]54 , organ lesion studies 18 , drug delivery 20 , and nanoparticle uptake [22][23][24][25][26] . It provides high-quality organ segmentations within a second.…”

Section: Discussionmentioning

confidence: 99%

“…The light-sheet microscopy dataset was partially acquired for a prior study 17 but was extended for this study. This imaging modality enables whole-body scans at microscopic resolution (6 μm) for entire mice without the need for destructive cryoslicing.…”

Section: Methodsmentioning

confidence: 99%

“…Due to its fast technological advancement, whole-body imaging with diverse modalities plays an increasingly important role in murine research [4][5][6][7][8][9][10][11][12][13][14] . For a broad range of research areas, from cancer [15][16][17] to organ lesion studies 18 , from radiation studies 19 to drug delivery 20,21 and nanoparticle uptake [22][23][24][25][26][27] , quantitative and comparative analyses of the acquired imaging data require segmentation of the mouse anatomy. Delineation of the major organs and other structures of interest allows the extraction of quantitative information such as organ shape and size, drug uptake, signals from biomarkers, or metastatic distribution from the imaging data.…”

mentioning

confidence: 99%

See 2 more Smart Citations

Deep learning-enabled multi-organ segmentation in whole-body mouse scans

et al. 2020

Self Cite

View full text Add to dashboard Cite

Whole-body imaging of mice is a key source of information for research. Organ segmentation is a prerequisite for quantitative analysis but is a tedious and error-prone task if done manually. Here, we present a deep learning solution called AIMOS that automatically segments major organs (brain, lungs, heart, liver, kidneys, spleen, bladder, stomach, intestine) and the skeleton in less than a second, orders of magnitude faster than prior algorithms. AIMOS matches or exceeds the segmentation quality of state-of-the-art approaches and of human experts. We exemplify direct applicability for biomedical research for localizing cancer metastases. Furthermore, we show that expert annotations are subject to human error and bias. As a consequence, we show that at least two independently created annotations are needed to assess model performance. Importantly, AIMOS addresses the issue of human bias by identifying the regions where humans are most likely to disagree, and thereby localizes and quantifies this uncertainty for improved downstream analysis. In summary, AIMOS is a powerful open-source tool to increase scalability, reduce bias, and foster reproducibility in many areas of biomedical research.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

mentioning

confidence: 99%

See 1 more Smart Citation

Deep learning-enabled multi-organ segmentation in whole-body mouse scans

et al. 2020

Self Cite

View full text Add to dashboard Cite

show abstract

“…Recent advances in medical imaging explicitly identified local image sub-regions that determine the training of classifier deep neural networks (Courtiol et al, 2019;Fu et al, 2019;Pan et al, 2019;Shamai et al, 2019). Localization of sub-regions that were particularly important for the classifier result permitted a visual assessment and pathological interpretation of distinctive image properties.…”

Section: Interpretation Of Latent Features Discriminating High and Lomentioning

confidence: 99%

Interpretable deep learning of label-free live cell images uncovers functional hallmarks of highly-metastatic melanoma

Zaritsky

Jamieson

Welf

et al. 2020

Preprint

View full text Add to dashboard Cite

Deep convolutional neural networks have emerged as a powerful technique to identify hidden patterns in complex cell imaging data. However, these machine learning techniques are often criticized as uninterpretable "black-boxes" -lacking the ability to provide meaningful explanations for the cell properties that drive the machine's prediction. Here, we demonstrate that the latent features extracted from label-free live cell images by an adversarial auto-encoding deep convolutional neural network capture subtle details of cell appearance that allow classification of melanoma cell states, including the metastatic efficiency of seven patientderived xenograft models that reflect clinical outcome. Although trained exclusively on patientderived xenograft models, the same classifier also predicted the metastatic efficiency of immortalized melanoma cell lines suggesting that the latent features capture properties that are specifically associated with the metastatic potential of a melanoma cell regardless of its origin.We used the autoencoder to generate "in-silico" cell images that amplified the cellular features driving the classifier of metastatic efficiency. These images unveiled pseudopodial extensions and increased light scattering as functional hallmarks of metastatic cells. We validated this interpretation by analyzing experimental image time-lapse sequences in which melanoma cells spontaneously transitioned between states indicative of low and high metastatic efficiency.Together, this data is an example of how the application of Artificial Intelligence supports the identification of processes that are essential for the execution of complex integrated cell functions but are too subtle to be identified by a human expert. -Gurion University of the Negev, Israel (to AZ). We thank Sean Morrison for PDX-derived cell models. We thank Andrew R. Cohen for LEVER. Author ContributionAZ, ESW and GD conceived the study. ESW designed the experimental assay, optimized culture conditions for PDX cells, and trained AN and JC to perform the experiments. AN performed all live imaging and mouse experiments. JC performed additional experiments. UE generated the PDX samples. AZ and ARJ developed analytic tools, analyzed, and interpreted the data. AZ drafted the manuscript. All authors wrote and edited the manuscript and approved its content.GD mentored all authors.

show abstract

“…For example, DeepMACT technology was recently established, and it enables detection of cancer metastasis and drug targeting at the cellular level in the whole mouse body. Here, an AI-based method could speed up the discovery of effective cancer treatments (Pan et al, 2019). In the future, further development of machine learning pipelines could pave the way for faster, unbiased and more accurate analysis of multidimensional samples.…”

Section: Conclusion and Future Perspectivementioning

confidence: 99%