Fourier-ring descriptor to characterize rare circulating cells from images generated using immunofluorescence microscopy

Emerson, Tegan; Kirby, Michael; Bethel, Kelly; Kolatkar, Anand; Luttgen, Madelyn; O’Hara, Stephen; Newton, Paul K.; Kühn, Peter

doi:10.1016/j.compmedimag.2014.10.003

Cited by 6 publications

(7 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Having cells freely released into solution and collected on optimized optical substrates without interfering bound beads allows for high quality fluorescence and brightfield imaging that reveals morphologies unique to CTCs. Based on standard CK and CD45 immunostaining and morphological features that are diagnostic in cytopathology, a set of criteria was developed to classify cells (Figure 2A ), based on both existing methods [ 35 , 38 , 39 ] and observations of cell populations captured from healthy donor samples using Vortex HT. Classifications were comprised of 3 categories: debris, WBCs, or CTCs.…”

Section: Resultsmentioning

confidence: 99%

“…Since many cells may transition to a mesenchymal state [ 32 ], traditional epithelial cell staining techniques may overlook a significant number of candidate cells [ 47 ], resulting in underreported performances especially in size-based isolation platforms. While most devices are characterized using probes for CK, CD45, and DAPI, the introduced CTC identification criteria makes use of a sequential checklist that includes well-defined morphological criteria associated with malignancy—which take advantage of accumulated cytopathology knowledge [ 10 , 38 , 39 , 48 ]—and may help minimize user-errors in manual enumeration. Morphological characterization may also help classify large cells that stain negative for common CTC markers, which may arise from size-based isolation methods, and cytometric analyses may sufficiently distinguish CTCs from other cell types present in blood, like monocytes, granulocytes, and cancer-associated non-CTCs such as disseminated tumor-activated macrophages [ 49 ].…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Classification of large circulating tumor cells isolated with ultra-high throughput microfluidic Vortex technology

et al. 2016

View full text Add to dashboard Cite

Circulating tumor cells (CTCs) are emerging as rare but clinically significant non-invasive cellular biomarkers for cancer patient prognosis, treatment selection, and treatment monitoring. Current CTC isolation approaches, such as immunoaffinity, filtration, or size-based techniques, are often limited by throughput, purity, large output volumes, or inability to obtain viable cells for downstream analysis. For all technologies, traditional immunofluorescent staining alone has been employed to distinguish and confirm the presence of isolated CTCs among contaminating blood cells, although cells isolated by size may express vastly different phenotypes. Consequently, CTC definitions have been non-trivial, researcher-dependent, and evolving. Here we describe a complete set of objective criteria, leveraging well-established cytomorphological features of malignancy, by which we identify large CTCs. We apply the criteria to CTCs enriched from stage IV lung and breast cancer patient blood samples using the High Throughput Vortex Chip (Vortex HT), an improved microfluidic technology for the label-free, size-based enrichment and concentration of rare cells. We achieve improved capture efficiency (up to 83%), high speed of processing (8 mL/min of 10x diluted blood, or 800 μL/min of whole blood), and high purity (avg. background of 28.8±23.6 white blood cells per mL of whole blood). We show markedly improved performance of CTC capture (84% positive test rate) in comparison to previous Vortex designs and the current FDA-approved gold standard CellSearch assay. The results demonstrate the ability to quickly collect viable and pure populations of abnormal large circulating cells unbiased by molecular characteristics, which helps uncover further heterogeneity in these cells.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Classification of large circulating tumor cells isolated with ultra-high throughput microfluidic Vortex technology

et al. 2016

View full text Add to dashboard Cite

show abstract

“…In order to do this, new sets of features may have to be identified that complement or even replace the color content of the cutouts. Examples of possible features would be further morphological quantities and Fourier-ring descriptors [ 38 ]. To apply machine learning to the subgrouping task would require more data than used here and a more rigorous manual classification performed by experts.…”

Section: Resultsmentioning

confidence: 99%

Automated Classification of Circulating Tumor Cells and the Impact of Interobsever Variability on Classifier Training and Performance

Svensson

Hübler

Figge

2015

Journal of Immunology Research

View full text Add to dashboard Cite

Application of personalized medicine requires integration of different data to determine each patient's unique clinical constitution. The automated analysis of medical data is a growing field where different machine learning techniques are used to minimize the time-consuming task of manual analysis. The evaluation, and often training, of automated classifiers requires manually labelled data as ground truth. In many cases such labelling is not perfect, either because of the data being ambiguous even for a trained expert or because of mistakes. Here we investigated the interobserver variability of image data comprising fluorescently stained circulating tumor cells and its effect on the performance of two automated classifiers, a random forest and a support vector machine. We found that uncertainty in annotation between observers limited the performance of the automated classifiers, especially when it was included in the test set on which classifier performance was measured. The random forest classifier turned out to be resilient to uncertainty in the training data while the support vector machine's performance is highly dependent on the amount of uncertainty in the training data. We finally introduced the consensus data set as a possible solution for evaluation of automated classifiers that minimizes the penalty of interobserver variability.

show abstract

“…Visual identification of CTCs is tedious, requires a higher level of training and can be subjective. Therefore, there is a critical need for technologies to automate and standardize optical detection of CTC in order for these techniques to become more robust and more amenable to routine use [ 35 ]. Nevertheless, we demonstrate that CTC analysis using RosetteSep™ and ScreenCell ® is sensitive, compared to EasySep™ and Dynabeads ® , and that the technique can be used for detection of CTC in metastatic breast cancer patients.…”

Section: Discussionmentioning

confidence: 99%

Comparative performance of different methods for circulating tumor cell enrichment in metastatic breast cancer patients

et al. 2020

View full text Add to dashboard Cite

The isolation and analysis of circulating tumor cells (CTC) has the potential to provide minimally invasive diagnostic, prognostic and predictive information. Widespread clinical implementation of CTC analysis has been hampered by a lack of comparative investigation between different analytic methodologies in clinically relevant settings. The objective of this study was to evaluate four different CTC isolation techniques–those that rely on surface antigen expression (EpCAM or CD45 using DynaBeads ® or EasySep™ systems) or the biophysical properties (RosetteSep™ or ScreenCell ® ) of CTCs. These were evaluated using cultured cells in order to calculate isolation efficiency at various levels including; inter-assay and inter-operator variability, protocol complexity and turn-around time. All four techniques were adequate at levels above 100 cells/mL which is commonly used for the evaluation of new isolation techniques. Only the RosetteSep™ and ScreenCell ® techniques were found to provide adequate sensitivity at a level of 10 cells/mL. These techniques were then applied to the isolation and analysis of circulating tumor cells blood drawn from metastatic breast cancer patients where CTCs were detected in 54% (15/28) of MBC patients using the RosetteSep™ and 75% (6/8) with ScreenCell ® . Overall, the ScreenCell ® method had better sensitivity.

show abstract

Fourier-ring descriptor to characterize rare circulating cells from images generated using immunofluorescence microscopy

Cited by 6 publications

References 25 publications

Classification of large circulating tumor cells isolated with ultra-high throughput microfluidic Vortex technology

Classification of large circulating tumor cells isolated with ultra-high throughput microfluidic Vortex technology

Automated Classification of Circulating Tumor Cells and the Impact of Interobsever Variability on Classifier Training and Performance

Comparative performance of different methods for circulating tumor cell enrichment in metastatic breast cancer patients

Contact Info

Product

Resources

About