Not Available in Your Country
Olympus AI for Quantitative Analysis of Fluorescently Labeled Cells with Ultra-Low Light Exposure

Rationale

Fluorescent labels are an invaluable tool in modern microscopy-based cell studies. The high exposure to excitation light, however, influences the cells directly and indirectly through photochemical processes. Adverse experimental conditions can lead to photodamage or phototoxicity with an observable impact on cell viability. Even if no direct effect is observed, strong light exposure can influence the cells’ natural behavior, leading to undesired effects. In long-term live cell experiments, fluorescence observation with minimal light exposure is particularly desirable. However, lower light exposure results in lower fluorescent signal, which typically decreases the signal-to-noise ratio (SNR), making it difficult to carry out quantitative analyses on the images (Figure 1).

Figure 1 Overview image for illustration of the different light levels involved in this study.

Figure 1
Overview image for illustration of the different light levels involved in this study.
 

igure 2 From left to right: DAPI-stained nuclei of HeLa cells with optimal illumination (100%), low light exposure (2%), very low light exposure (0.2%) and extremely low light exposure (0.05%). The SNR decreases signifi cantly, because signal levels decrease to the camera’s noise level, eventually reaching the detection threshold of the camera. Contrast optimized per SNR for visualization only.

Figure 2
From left to right: DAPI-stained nuclei of HeLa cells with optimal illumination (100%), low light exposure (2%), very low light exposure (0.2%) and extremely low light exposure (0.05%). The SNR decreases significantly, because signal levels decrease to the camera’s noise level, eventually reaching the detection threshold of the camera. Contrast optimized per SNR for visualization only.
 

Deep Learning Technology

From a technological point of view, the analysis of cells with ultra-low light exposure can simply be seen as a problem of analyzing images with very low SNR (Figure 2). To address this issue and analyze low-signal images with unprecedented robustness and precision, Olympus has integrated a new image analysis approach, based on deep neural networks, into its scanR HCS analysis software. This kind of neural network architecture has recently been described as the most powerful object segmentation technology available [Long et al. 2014: Fully Convolutional Networks for Semantic Segmentation].

Neural networks of this kind feature an unrivaled adaptability to various challenging image analysis tasks, making it an optimal choice for the non-trivial quantitative analysis of cells with ultralow light exposure. In a training phase, the neural networks automatically learn how to predict the desired parameters, for example positions and contours of cells or cell compartments, a process called segmentation of objects of interest.

During the training phase, the network is fed with pairs of example images and “ground truth” data (i.e. object masks where the objects of interest are annotated). Once the network is trained, it can be applied to new images and predict the object masks with very high precision.

Typically in machine learning, the annotations (e.g. boundaries of cells) are provided by human experts. This can be a very tedious and time-consuming step because neural networks require large amounts of training data in order to fully exploit their large potential.

To overcome these difficulties, Olympus uses a concept coined self-learning microscopy. In self-learning microscopy, the microscope automatically generates the ground truth required for training the neural network by acquiring reference images during the training phase.

For example, in order to teach the neural network the robust detection and segmentation of nuclei in ultra-low exposure images, the microscope automatically acquires large numbers of image pairs where one image is taken under optimal lighting conditions and the other is underexposed. These pairs are used to train the neural network to correctly analyze the noisy images created with ultra-short exposure levels.

Since this approach to ground truth generation requires little human interaction, huge amounts of training image pairs can be acquired within a short time. This makes it possible for the neural network to adapt to variations including different SNR levels and illumination inhomogeneities during the training phase, which results in a learned neural network model that is unaffected by all these issues.

Low Signal Segmentation Training

To train the neural network to detect nuclei robustly in varying SNR conditions, the Olympus scanR system is set up to acquire a whole 96 well plate with 3×3 positions in each well. The following images are acquired for each position:

  • DAPI image with 100% light exposure for optimal SNR (200 ms exposure time and 100% LED excitation light intensity)
  • DAPI image with 2% light exposure (4 ms exposure time and 100% LED excitation light intensity)
  • DAPI image with 0.2% light exposure (4 ms exposure time and 10% LED excitation light intensity)
  • DAPI image with 0.05% light exposure (4 ms exposure time and 2.5% LED excitation light intensity)

 90 wells are used for training, while 6 wells are excluded and used for validation later. Nuclei positions and contours are detected in the DAPI image with optimal SNR with standard image analysis protocols. The segmentation masks are paired with the images with reduced SNR and used for training as shown in Figure 3. Training took about 2:40 hours on a PC with NVIDIA GeForce GTX 1070 GPU. After training, the neural network is capable of detecting nuclei robustly at all exposure times.

Applying Low Signal Segmentation

The trained neural network can now be applied to new images with different SNRs following the workflow depicted in Figure 4. The resulting contours at different SNRs are shown in Figure 5.

Figure 3 Training the neural network. Pairs of images with high and suboptimal SNR are used to teach the neural network object detection in all SNR conditions.

Figure 3
Training the neural network. Pairs of images with high and suboptimal SNR are used to teach the neural network object detection in all SNR conditions.
 

Figure 4 Applying the trained neural network (inference). The network has been trained to predict object positions and contours in varying conditions, including very low SNR images.

Figure 4
Applying the trained neural network (inference). The network has been trained to predict object positions and contours in varying conditions, including very low SNR images.
 

Figure 5 Examples of detected objects in images with different SNRs acquired with light exposure levels of (from left to right) 100%, 2%, 0.2% and 0.05%. Contrast optimized per SNR for visualization only.

Figure 5
Examples of detected objects in images with different SNRs acquired with light exposure levels of (from left to right) 100%, 2%, 0.2% and 0.05%. Contrast optimized per SNR for visualization only.
 

Direct comparison of the segmentation results at different SNR (Figure 6) and an overlay of contours at different SNRs clearly demonstrates the network’s detection capabilities at reduced exposure levels (Figure 7). The contours deduced from the images overlap almost perfectly (red, yellow, teal), except for the lowest light exposure (blue contours). This indicates that the limit for robust detection with the neural network lies between 0.2% and 0.05% light exposure.

Validation of the Results

In order to validate the results of the neural network and confirm the limit for quantitative analysis, images of two wells per illumination condition are analyzed. Nuclei positions and contours are determined, the nuclei are counted and the area of the nucleus and the mean intensity of the DAPI signal are measured. The results under optimal conditions (Figure 8, top left) show two distinct populations, correlating with cells in the G1 (single DNA content) and G2 stage (double DNA content) of the cell cycle.

Figure 6 Nuclei segmentation results for different SNR images acquired with light exposure levels of (from top left to bottom right) 100%, 2%, 0.2% and 0.05%. The contours derived from the lowest SNR (bottom right) deviate significantly from the correct contours and indicate the limit of the technique for quantitative analysis at ultra-low exposure levels is between 0.2% and 0.05% of the usual light exposure. Contrast optimized per SNR for visualization only.

Figure 6
Nuclei segmentation results for different SNR images acquired with light exposure levels of (from top left to bottom right) 100%, 2%, 0.2% and 0.05%. The contours derived from the lowest SNR (bottom right) deviate significantly from the correct contours and indicate the limit of the technique for quantitative analysis at ultra-low exposure levels is between 0.2% and 0.05% of the usual light exposure. Contrast optimized per SNR for visualization only.
 

Figure 7 Nuclei contours detected by the neural network at 4 different SNRs shown on top of the image acquired with the lowest light exposure level (0.05%). Contrast optimized for visualization only.

Figure 7
Nuclei contours detected by the neural network at 4 different SNRs shown on top of the image acquired with the lowest light exposure level (0.05%). Contrast optimized for visualization only.
 

When the images are analyzed at only 2% exposure, the dynamic range is reduced dramatically (Figure 8, top right). However, after rescaling the plot, the same distinct populations are visible, and a similar distribution is found as summarized in Table 1 for all light exposure levels. Even when the same method was applied to images exposed at 0.2%, the values for cell count and cells in G1 and G2 state are almost identical.

At the lowest exposure level (0.05%), the dynamic range is reduced even further. At this exposure, the distinct populations are no longer clearly visible (Figure 8, bottom right) and as a result, the cell count is underestimated by about 4% while the percentage of cells, for example, in G2 is overestimated by about 1%. This indicates that when high precision measurements are required, the SNR limit for successful analysis has been reached. However, the results can at least be used for a rough estimate, even with this ultra-low light exposure.

Figure 8 Cell cycle diagrams derived from images with optimal SNR (100% light exposure) and reduced SNR (2%, 0.2%, 0.05% light exposure, respectively). Note the different scaling of the y-axis.

Figure 8
Cell cycle diagrams derived from images with optimal SNR (100% light exposure) and reduced SNR (2%, 0.2%, 0.05% light exposure, respectively). Note the different scaling of the y-axis.
 

Measurement Results Derived at Different Light Exposure Levels

Light exposure 100% 2% 0.2% 0.05%
Detected nuclei count 11072 11015 11007 10595
G1 state 63.9% 63.9% 63.8% 63.2%
G2 state 36.1% 36.1% 36.2% 36.8%

Conclusions

Quantitative analysis of fluorescence images of nuclei benefits many areas of life science research. Automated detection using an AI-based approach can save time and improve reproducibility, however, the method must be robust and should produce reliable results even when the fluorescence signal is low.

Olympus’ scanR HCS software, with its integrated AI-based self-learning microscopy and image analysis capabilities, enables reliable detection of nuclei at a range of SNR levels. The results presented here show that the convolutional neural networks of Olympus’ AI technology can carry out the analysis with as little as 0.2% of the light usually required. This makes it highly suitable for fast, robust quantitative analysis of large numbers of images at challenging lighting conditions. The dramactic decrease in light exposure ensures negligible influence on cell viability and enables fast data acquisition and long term observation of living cells.

Author

Dr. Mike Woerdemann
Product Manager
Olympus Soft Imaging Solutions
GmbH
Münster, Germany

对不起,此内容在您的国家不适用。

相关产品
生命科学用高内涵筛选工作站

scanR

  • 快速精确地图像采集和分析
  • 超精确的实时实验控制
  • 分析模块

This site uses cookies to enhance performance, analyze traffic, and for ads measurement purposes. If you do not change your web settings, cookies will continue to be used on this website. To learn more about how we use cookies on this website, and how you can restrict our use of cookies, please review our Cookie Policy.

OK