Supervised learning to quantify amyloidosis in whole brains of an Alzheimer’s disease mouse model acquired with optical projection tomography

Nguyen, David; Uhlmann, Virginie; Planchette, Arielle; Marchand, Paul J.; Ville, Dimitri Van De; Lasser, Theo; Radenović, Aleksandra

doi:10.1364/boe.10.003041

Cited by 12 publications

(12 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Similar results are found when the considered processing operations are performed after the 3D reconstruction (Figure 4g). In this case, the averaged standard scores of the gaussian smoothing operators applied over all voxels of the reconstructed 3D images are around 0 because smoothing is already performed in the applied FBP algorithm 26 .…”

Section: D Cell Segmentationmentioning

confidence: 99%

“…For OPT measurements, we processed datasets of full intact mouse brains previously studied in 26 using an SL approach to quantify amyloidosis of an Alzheimer's disease mouse model. The epiimage projections were acquired with a custom mesoscopic OPT setup consisting of a calibrated CMOS camera (ORCA-Flash 4.0 V2, Hamamatsu) coupled to a 300-mm achromat objective lens providing 0.5X magnification of the sample 33 .…”

Section: Optmentioning

confidence: 99%

“…The final plaque segmentation and quantification was then realized by applying a threshold value of 0.5 to the 3D plaque probability map. This pipeline was the same as that provided by the authors of a previous study 26 .…”

Section: Supplementary Filesmentioning

confidence: 99%

See 2 more Smart Citations

Statistical Distortion of Supervised Learning Predictions in Optical Microscopy Induced by Image Compression

Pomarico

Schmidt

Chays

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

The growth of data throughput in optical microscopy has triggered the extensive use of supervised learning (SL) models on compressed datasets for automated analysis. Investigating the effects of image compression on SL predictions is therefore pivotal to assess their reliability, especially for clinical use.We quantify the statistical distortions induced by compression through the comparison of predictions on compressed data to the raw predictive uncertainty, numerically estimated from the raw noise statistics measured via sensor calibration. Predictions on cell segmentation parameters are altered by up to 15% and more than 10 standard deviations after 16-to-8 bits pixel depth reduction and 10:1 JPEG compression. JPEG formats with higher compression ratios show significantly larger distortions. Interestingly, a recent metrologically accurate algorithm, offering up to 10:1 compression ratio, provides a prediction spread equivalent to that stemming from raw noise. The method described here allows to set a lower bound to the predictive uncertainty of a SL task and can be generalized to determine the statistical distortions originated from a variety of processing pipelines in AI-assisted fields.

show abstract

Section: D Cell Segmentationmentioning

confidence: 99%

Section: Optmentioning

confidence: 99%

See 1 more Smart Citation

Statistical Distortion of Supervised Learning Predictions in Optical Microscopy Induced by Image Compression

Pomarico

Schmidt

Chays

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…, 2016 ; Bulten et al. , 2019 ; Nguyen et al. , 2019 ; Oktay and Gurses, 2019 ; van der Heyden et al.…”

Section: Introductionunclassified

Deep learning approach for quantification of organelles and misfolded polypeptide delivery within degradative compartments

Morone

Marazza

Bergmann

et al. 2020

MBoC

View full text Add to dashboard Cite

show abstract

“…OPT is also capable of utilizing many colored and fluorescent dyes that were developed for tissue-specific or gene-specific staining, which is important to three-dimensional observations of specific tissues because it allows a computer to automatically determine the three-dimensional structure of the target tissue. Many universities [3][4][5][6][7][8][9][10][11] have adopted OPT systems to help them do biomedical research.…”

Section: Introductionmentioning

confidence: 99%

Optical Projection Tomography Using a Commercial Microfluidic System

Fei

Liu

et al. 2020

Micromachines

View full text Add to dashboard Cite

Optical projection tomography (OPT) is the direct optical equivalent of X-ray computed tomography (CT). To obtain a larger depth of field, traditional OPT usually decreases the numerical aperture (NA) of the objective lens to decrease the resolution of the image. So, there is a trade-off between sample size and resolution. Commercial microfluidic systems can observe a sample in flow mode. In this paper, an OPT instrument is constructed to observe samples. The OPT instrument is combined with commercial microfluidic systems to obtain a three-dimensional and time (3D + T)/four-dimensional (4D) video of the sample. “Focal plane scanning” is also used to increase the images’ depth of field. A series of two-dimensional (2D) images in different focal planes was observed and compared with images simulated using our program. Our work dynamically monitors 3D OPT images. Commercial microfluidic systems simulate blood flow, which has potential application in blood monitoring and intelligent drug delivery platforms. We design an OPT adaptor to perform OPT on a commercial wide-field inverted microscope (Olympusix81). Images in different focal planes are observed and analyzed. Using a commercial microfluidic system, a video is also acquired to record motion pictures of samples at different flow rates. To our knowledge, this is the first time an OPT setup has been combined with a microfluidic system.

show abstract

Supervised learning to quantify amyloidosis in whole brains of an Alzheimer’s disease mouse model acquired with optical projection tomography

Cited by 12 publications

References 42 publications

Statistical Distortion of Supervised Learning Predictions in Optical Microscopy Induced by Image Compression

Statistical Distortion of Supervised Learning Predictions in Optical Microscopy Induced by Image Compression

Deep learning approach for quantification of organelles and misfolded polypeptide delivery within degradative compartments

Optical Projection Tomography Using a Commercial Microfluidic System

Contact Info

Product

Resources

About