Home / Blog / Image Analysis Quality Control Part 2: Tissue Section QC

Image Analysis Quality Control Part 2: Tissue Section QC
4 June 2020
By Hannah Thomson

Getting the image quality right before embarking on advanced image analysis may be an obvious statement to make.  However, there are multiple ways to assess image quality and this speaks to the need for robust quality control (QC) processes throughout an image analysis workflow. In the second part of our 3-part Image Analysis Quality Control Blog series, we look at a number of potential issues identified during Tissue Section QC and the impact they can have on image analysis data if not remedied.

Condition of the Section:

It is important to review the section for any damage, for example any folds, tears, knife marks, creasing or chattering. Damaged or obscured areas of tissue should be manually excluded using annotations and if they cover a large area of tissue, this can result in an image failing QC and being removed from the data set.  Damage to the peripheral tissue of sections is particularly evident in needle biopsies, where a high degree of trauma to the tissue can occur during the harvesting process. If not annotated out of the region of interest (ROI), this can skew the analysis of the image.

Right: Damaged tissue section with chattering present

Orientation & Alignment:

“Serial section” is the term used to describe sequential slices of tissue taken from a histology block.  Often serial sections are used to allow staining using H&E on one section and biomarker staining on an adjacent section. Annotations of slides (defining ROI such as tumour or stroma) are often formed using the H&E staining of one section and then the ROI electronically overlaid or co-registered within the image analysis software on the section containing the stained biomarker.

With a standard histology section being generally 4 μm thick, one can see the value in using sequential, or serial sections for the H&E and biomarker staining but the orientation and alignment of the sections must be checked to ensure that successful co-registration is feasible.  Generally, this technique is sufficient to define biomarker location at the tissue level, but co-registration of serial samples to robustly define co-staining at the cellular level is prone to errors due to the distance between the sections.

Features of the Sample:

Images should be reviewed to check if there are any tissue features that would not be relevant to the analysis, for example, blood vessels or areas of necrosis. Areas of necrosis are common in sections from PDX models where the highly proliferative tissue out-paces the nutritive process of angiogenesis. In addition, the presence of any deposits, for example, carbon deposits in lung tissue or melanin in melanoma samples, should be recorded at QC stage. Any irrelevant tissue areas or deposits can then be removed during annotation procedures.

Left: Carbon deposits present in a lung tissue section

Sufficient amount of sample:

In order to generate robust data, there must be a sufficient amount of viable tissue area present to analyse.  As a guideline, the area of analysed tissue on large sections should not be less than 1.0 mm2. Similarly, if analysis is required in a specific region of interest (ROI), e.g. tumour, there must be sufficient area of tissue ROI for quantification. Without sufficient biomarker quantification within a section then the confidence in making judgements on the difference between samples would be hampered.

Right: Exemplar images with insufficient sample present on the slide

There are many factors to consider when deciding if an image is suitable for image analysis and these are just some to include when conducting Tissue Section QC. To find out what else we include in our QC check, click here to read Image Analysis Quality Control Part 1: Image Scan QC and Image Analysis Quality Control Part 3: Staining QC.

Encountering some of the issues above does not necessarily mean that an image cannot be analysed as some problems can be overcome, for example by negative annotation or additional algorithm development. Unfortunately, at OracleBio we do have to fail a significant number of images every year because the image analysis data generated will not be robust. If you are unsure if your images are suitable for quantitative analysis or need advice on how to avoid encountering some of these issues, please contact us. 

About the Author: Nicole Stillie

Senior Image Analysis Scientist

Nicole joined OracleBio in 2016 after graduating with a B.Sc. in Biomedical Science. She is now a Senior Image Analysis Scientist and QCs hundreds of images every month. Connect with Nicole on LinkedIn. 

Discover how we maintain high levels of quality

Explore OracleBio's Image Analysis Blog

Stay up to date with OracleBio

Subscribe to our mailing list to stay informed about news, webinars, events and more