![[WDAF 2026] Institutions turn to tokenization as buyers, issuers: Ondo Finance](https://cryptonews100.com/wp-content/uploads/2026/11/news-p.v1.20261109.92fd66e07214413db4355da131b5277f_P1-120x86.jpg)
Summary
Background and Goal
The heterogeneous response of most cancers cells to radiation necessitates complete investigations into cell survival, which are sometimes performed utilizing clonogenic assays. These assays contain laborious and subjective handbook enumeration of cell colonies. This paper presents an progressive digital picture processing approach for automated cell colony counting.
Strategies
This analysis centered on HT-29 cells (colon most cancers). The methodology makes use of Hough strategies to attain exact detection, mathematical morphology, and institution of the optimum minimal colony measurement utilizing ROC curve. This approach seamlessly built-in into ImageJ software program as a ColCounter plugin.
Outcomes
Pictures with decision of 1200*1200 pixels achieved sensitivity and specificity of 92.88% and 92.62%, respectively. To find out the software program´s effectiveness of picture evaluation of completely different resolutions in contrast with handbook counting, an analysis of the automated technique’s reliability was performed. The findings revealed an general intraclass correlation coefficient (ICC) of 0.89 (95% CI: 0.81–0.93). The bounds of settlement had been decided utilizing the Bland-Altman technique.
Conclusions
The proposed methodology demonstrated interchangeability with the traditional handbook enumeration approach. The outcomes affirm the software program’s elevated efficiency, even in altered contexts. ColCounter gives handbook modifying operate, enhancing its utility as a streamlined, fast, and exact device for quantifying tumor cell colonies in clonogenic assays.
Quotation: Forero MG, Medina LA, Patiño-Aldana AF, Hernandez-Rodríguez ADP, López-Molina G, Mena HH, et al. (2026) A novel approach for automated counting of tumor cell colonies. PLoS One 21(7):
e0354467.
https://doi.org/10.1371/journal.pone.0354467
Editor: Ling Ding, College of Nebraska Medical Heart, UNITED STATES OF AMERICA
Acquired: January 14, 2026; Accepted: July 8, 2026; Printed: July 23, 2026
Copyright: © 2026 Forero et al. That is an open entry article distributed below the phrases of the Creative Commons Attribution License, which allows unrestricted use, distribution, and copy in any medium, offered the unique creator and supply are credited.
Knowledge Availability: The minimal knowledge set and accompanying code is offered on the following hyperlinks: Pictures of most cancers cell colonies: https://doi.org/10.34848/DSSJMY Guide and plugin counts: https://doi.org/10.34848/A87IDN Concordance evaluation and ROC curve: https://doi.org/10.34848/YB5H2L ColCounter Code: https://www.researchgate.net/publication/367603821_Automatic_colony_count.
Funding: This analysis was funded by the Ministry of Science, Know-how and Innovation of Colombia, MinCiencias (https://minciencias.gov.co/), grant numbers 874 (A.O-M) and 887 (A.D.P.H-R); by Universidad del Rosario (https://urosario.edu.co/) and Universidad de Ibagué (https://www.unibague.edu.co/), Colombia. The funders had no function in examine design, knowledge assortment and evaluation, resolution to publish, or preparation of the manuscript.
Competing pursuits: Andrés Felipe Patiño-Aldana, Andrea Del Pilar Hernández-Rodríguez, Mateo Díaz-Quiroz, Margarita Garcia, Alejandro Ondo-Méndez stories monetary assist was offered by Colombia Ministry of Science Know-how and Innovation. The opposite authors declare that they haven’t any recognized competing monetary pursuits or private relationships that would have appeared to affect the work reported on this paper. This doesn’t alter our adherence to PLOS ONE insurance policies on sharing knowledge and supplies.
1. Introduction
Radiotherapy performs an necessary function within the therapy of most cancers, being utilized in roughly 50% of sufferers [1]. Its antitumor impact is because of DNA injury by direct and oblique mechanisms, which activate cell demise pathways [2]. Nevertheless, DNA restore mechanisms could also be activated in surviving most cancers cells to keep up their proliferation capability. On this context, clonogenic assays characterize the in vitro gold normal for assessing radioresistance, as they consider the flexibility of cells to outlive, proliferate, and kind colonies after publicity to a cytotoxic agent [3–5]. Though extensively used to check radioresistance, clonogenic assays are additionally employed in different oncologic fields like drug screening and quantifying “stemness” and tumorigenicity of most cancers stem cells [6,7].
Cell survival curves describe the connection between the radiation dosage and the survival cell fraction [8]. The development of these graphics requires colony counting, which tends to be handbook, making it a labor-intensive and error-susceptible process [9,10]. Automated and semi-automated colony depend strategies based mostly on picture evaluation have been proposed to reinforce accuracy, reproducibility, and expedite the counting course of. Examples of these strategies embrace ColonyCountJ, an ImageJ plugin for quantifying colony parameters in clonogenic assays [10]; CoCoNut, a standalone cell colony counter [11]; and a high-throughput technique based mostly on edge detection for automated colony and cell counting [12].
As automated methodologies are developed to sort out advanced microscopy picture evaluation throughout various domains, standard cell counting software program usually falls quick when confronted with colony counting challenges, which embrace routinely figuring out wells, mitigating illumination inconsistencies inherent in scanner-based scans, and facilitating colony choice based mostly on cell counts. Thus, an pressing want arises for strategies and software program explicitly tailor-made to satisfy these necessities.
Analysis efforts on this specific area carry on increasing, completely different proposals have been developed exploring distinct approaches just like the use of the Watershed approach for picture segmentation, estimation of effectively floor space lined by colonies, and the identification of areas of curiosity and native segmentation to establish and label cancerous colonies [11,13,14]. Some novel approaches depend on the realm lined by the colony (ACC) somewhat than the colony depend itself, utilizing completely different strategies like spatial Fuzzy C-Means clustering, Circle Hough Remodel, and native adaptive thresholding for picture processing [15,16]. Extra just lately, synthetic intelligence-based fashions have been developed, for instance, Klett et al. utilized a time-lapse microscopy for real-time monitoring of colony formation, and Lavitt et al. developed a mannequin based mostly on convolutional neural networks for cell counting that may very well be extrapolated to clonogenic assays [17,18].
Though options for colony counting have been designed, some limitations embrace excessive relative error values, restrictions when figuring out merged or overlapping colonies, the necessity for excessive decision photos, calculation of the survival fraction based mostly on the ACC, whose parameter, regardless of correlating with the depend, doesn’t exchange it; and the requirement of specialised gear for picture acquisition that’s tough to entry. It’s nonetheless essential to develop fashions that reveal good efficiency and that don’t characterize acquisition boundaries.
Exploring completely different strategies for creating instruments that facilitate colony counting would result in the technology of high-performance, user-friendly, and cost-effective strategies that overcome the constraints of handbook counting. The current examine aimed to explain the event of an progressive mannequin based mostly on digital picture processing, particularly utilizing the Hough approach, for automated cell colony counting “ColCounter” and to judge its efficiency in contrast with handbook counting.
2. Supplies and strategies
2.1. Cell tradition and irradiation protocols
HT-29 cells (colorectal most cancers; ATCC code HTB-38TM) [19] had been grown in Dulbecco´s Modified Eagle´s medium (DMEM, Biowest®) supplemented with 10% Fetal Bovine Serum (FBS, Hyclone®) and seeded in 60 mm diameter cell tradition dishes till they reached confluence. Cells had been irradiated with 6 MV X-ray photons at doses of 0, 1, 2, 4, 6, 8, 10, 12, 16, and 20 Gy. To make sure digital equilibrium and dose homogeneity within the goal quantity, the dishes had been irradiated on a paraffin block. A clonogenic assay was carried out following the plating after therapy protocol [3]. Cells had been seeded in 6 effectively plates in triplicate in a spread of 200–2000 cells per effectively, relying on the radiation dose. Cells had been cultured in DMEM supplemented with 10% FBS and incubated at 37°C in an environment containing 5% CO2 for 14 days to permit adhesion. Lastly, colonies had been mounted and stained with 4% formaldehyde and 1% crystal violet, respectively. The ensuing photos, encompassing six plates, had been acquired utilizing a Hewlett-Packard scanner (HP Scanjet Professional) at resolutions of 1024 x 5320 and 456 x 456 pixels, as illustrated in Fig 1.
In complete, 17 six-well plates had been used, leading to a collective depend of 102 wells, that had been analyzed to find out the optimum colour channel for subsequent processing. Amongst these, the colonies from 14 plates (equal to 84 wells) had been recognized manually by two specialists who collaborated to achieve a consensus. This meticulous course of yielded ground-truth photos, exemplified in Fig 2, which had been subsequently employed to outline the minimal colony measurement, which is a vital parameter for creating the plugin. Subsequent, 15 completely different plates had been used to validate the strategy utilizing handbook counting as a comparative technique. Nevertheless, 4 wells out of these 15 plates (90 wells) had been discarded because of poor scan high quality (heterogenous mild and dark-colored background). Lastly, 86 wells had been used for the concordance evaluation.
Fig 2. Floor-truth photos for technique validation.
Pink signifies colonies comprising 50 or extra cells, and blue represents clusters of fewer than 50 cells that don’t meet the standards for classification as colonies.
2.2. Statistical evaluation
To develop the plugin, a ROC curve was used to evaluate the sensitivity and false-positive charge (1-specificity). The target of this approach was to find out the optimum minimal colony space that would obtain the best true optimistic charge whereas concurrently minimizing the false optimistic charge.
As soon as the plugin was configured, the concordance between handbook depend and plugin outcomes was assessed. This analysis aimed to gauge the plugin’s efficacy in eventualities the place researchers are confronted with the need to work with photos of resolutions aside from these specified within the unique protocol. To check the adaptability of the device throughout varied resolutions, images of completely different resolutions (200, 300, and 600 dpi) had been used.
To quantify the extent of settlement, the intra-class correlation coefficient was calculated. Moreover, the Bland-Altman technique was employed to analyze the common bias and delineate the boundaries of settlement between the 2 strategies [20]. Moreover, a Passing-Bablok regression was fitted, enabling a non-parametric evaluation of the null speculation regarding the equivalence of efficiency between handbook counting and the ColCounter plugin [21]. To conduct statistical evaluation, R software program was employed at the side of packages together with tidyverse, irr, blandr, and mcr [22–26].
2.3. Strategies
Fig 3 reveals a complete flowchart of technique growth. Upon receiving the enter picture, the next step entails choosing the colour channel that gives essentially the most informative knowledge. It’s necessary to notice that not all colour channels supply related data. As depicted in Fig 1, colonies exhibit a faint purplish hue arising from the fusion of blue and crimson parts, whereas the background maintains a grayish look. Consequently, the distinction between colonies and the frivolously coloured background in channels like crimson and blue was much less pronounced in comparison with the inexperienced channel, as a result of absence of inexperienced colour inside the colonies. The picture representations for every colour channel are proven in Fig 4. These photos are accompanied by the resultant grayscale picture generated by averaging the values from all three colour channels. This computation is mathematically summarized in Equation 1.
To find out which colour channel contained extra data and mathematically confirm the choice to pick out inexperienced for colony detection over the opposite two channels, the entropy—a measure of data—was calculated for every channel utilizing Equation 2. The place h signifies the normalized histogram of every colour channel, q represents the colour degree, and L pertains to the utmost depth inside every colour channel. Table 1 lists the entropy values corresponding to every colour channel and monochromatic picture, obtained by averaging the colour values from every channel throughout all pixels. The entropy values had been measured throughout 17 plates. The outcomes reveal that the inexperienced channel persistently displayed the best entropy values, which is per the expectation. This end result bolsters the rationale for choosing the inexperienced channel for subsequent processing of your complete set of photos.
Upon choosing the suitable colour channel, the photographs had been segmented to differentiate colonies from background. Nevertheless, as a result of of the substantial variability in colony depth inside the photos, a thresholding technique was discovered unsuitable. As a substitute, Black Prime Hat transformation was employed to differentiate between the background and darker areas, which corresponded to the colonies, as illustrated in Fig 5. This approach enhances the visibility of smaller, darker buildings in opposition to a comparatively uniform background. This was achieved by subtracting the morphological closing end result from the unique picture. Mathematically, if Q represents the unique picture and B is the structuring ingredient, the Black Prime Hat transformation T is calculated as follows (Equation 3):
On this examine, an octagon with a radius bigger than the measured colony measurement was chosen because the structuring ingredient B. This transformation enhanced the distinction between the colonies and the background, facilitating the next processing steps.
One other benefit of using the Black Prime Hat technique as a substitute of thresholding is its potential to handle reflection artifacts which will come up throughout scanning. The shadow, with a radius bigger than that of the colonies, was successfully eliminated, as demonstrated in Fig 6.
Fig 6. The Black Prime Hat Operator eliminates artifacts attributable to edge reflections on plates throughout scanning.
A. Shut-up view of a effectively extracted from the unique picture. The artifact ensuing from the plate edge reflection on the effectively’s edge is highlighted inside the crimson field. B. Outcome obtained by making use of the Prime Hat operator with an octagonal structuring ingredient. The black field signifies the digital disappearance of the artifact.
To differentiate the colonies from the background, the Otsu thresholding technique was employed, which yielded superior outcomes in contrast with the opposite strategies, regardless of the histogram not exhibiting two clearly identifiable Gaussian modes, as illustrated in Fig 7. Conversely, the histogram was predominantly unimodal, making it difficult to find the mode akin to the colonies. Fig 8 reveals the end result of the thresholding course of utilizing the Otsu technique.
To find out the quantity of colonies current, it’s essential to initially establish every effectively individually. Given the round form of the wells, two distinct strategies had been assessed for detecting the perimeters of every plate. These strategies embrace the Hough remodel, which is often used to establish geometric shapes, and the Random Pattern Consensus (RANSAC) approach. Each strategies had been optimized to particularly search for circles that carefully matched the scale of the wells. Moreover, to mitigate the time complexity related to assessing a big quantity of pixels, the enter binarized picture underwent skeletonization, ensuing within the discount of the perimeters to a single-pixel width, as depicted in Fig 9. The Hough remodel demonstrated superior efficiency for precisely figuring out circles [27], thus establishing it as the popular approach for this goal.
Following effectively localization, the ensuing picture was used to generate a masks utilizing morphological filling, successfully separating the wells from the background, as proven in Fig 10. To isolate the colonies inside all wells, the minimal worth at every pixel between the binarized picture, proven in Fig 8, and the masks was computed, as illustrated in Fig 11.
Earlier than colony counting, it’s important to get rid of the define of every plate. To attain this, after the colonies have been efficiently separated from the background, the round picture obtained by the Hough remodel is dilated and subtracted from the binarized picture utilizing Otsu’s technique. The ensuing output is proven in Fig 12.
2.4. Colony counting
After segmenting the colonies and eradicating the background noise, every effectively was labeled utilizing 8-connectivity. This assigns a definite label to every effectively and creates a masks to establish every one of them. The objects detected inside every effectively had been then labeled, once more utilizing 8-connectivity, and subsequently, the colonies had been counted, as demonstrated in Fig 13. The selection of 8-connectivity for labeling colonies somewhat than 4-connectivity is justified as a result of the previous is extra appropriate for high-quality object labeling, stopping conditions wherein colonies related solely by a diagonal pixel are marked as separate entities.
Fig 13. Automated identification and quantification of colonies inside a effectively utilizing the proposed technique.
A. Masks utilized to establish particular person colonies within the plate. B. Colonies inside the effectively. C. Labeled and counted colonies.
To make sure correct colony counting and keep away from the inclusion of cell particles or smaller artifacts that could be inaccurately counted as colonies, it was crucial to determine the minimal colony measurement. In pursuit of this objective, the minimal acceptable measurement of every colony was fine-tuned by contemplating 9 attainable space values starting from 2 to 10 pixels. An ROC curve was constructed to judge the sensitivity and specificity of every worth, as depicted in Fig 14. This complete evaluation revealed that the optimum minimal measurement, which struck the very best steadiness between sensitivity and specificity, was eight pixels. Consequently, particles with areas under eight pixels had been systematically excluded, as demonstrated in Fig 15.
3. Outcomes
To evaluate the efficiency of the developed technique, the outcomes had been in contrast with these obtained manually by specialists. The analysis concerned calculating the sensitivity, specificity, and diagnostic accuracy measures. When evaluating the strategies, the outcomes had been categorized as true positives (TP), false positives (FP), false negatives (FN), and true negatives (TN). Table 2 presents the outcomes derived from the comparability between the handbook depend, utilized as the bottom fact, and the outcomes obtained utilizing the developed plugin, known as ColCounter, within the six wells of the 17 photos.
To find out specificity and sensitivity, the colonies recognized within the floor fact photos had been in contrast with these detected by the plugin. True positives (TP) had been outlined as colonies marked within the floor fact that had been appropriately recognized within the plugin-generated photos. True negatives (TN) corresponded to background areas or non-colony particles appropriately recognized as such by the plugin. False positives (FP) had been areas incorrectly recognized as colonies by the plugin however not current within the floor fact. False negatives (FN) had been colonies current within the floor fact that the plugin did not detect. Sensitivity was calculated as TP/(TP + FN), reflecting the strategy’s potential to appropriately establish precise colonies, whereas specificity was calculated as TN/(TN + FP), reflecting its potential to appropriately reject non-colony objects. It is very important observe that the sensitivity and specificity of the automated detection are decided by the algorithm’s potential to establish colony buildings within the picture somewhat than by the preliminary quantity of seeded cells. Though the quantity of cells could affect colony density in clonogenic assays, the detection standards utilized by the ColCounter plugin are impartial of the preliminary cell depend, offered that colonies stay sufficiently separated to permit particular person identification.
ColCounter achieved a sensitivity of 92.88% and a specificity of 92.62%, exceeding the values obtained with the identical samples and reported by Roldán et al. in 2019, which had been 82.3% for sensitivity and 76% for specificity [14]. Roldán et al. additionally in contrast their technique with the CellCounter and OpenCFU [28] packages, which ColCounter outperformed. The diagnostic accuracy was calculated to be 92.85%, indicating a superb proportion of appropriately recognized colonies and non-colony components utilizing the proposed technique. To additional improve the counting course of, outcomes could be simply and shortly edited utilizing ImageJ instruments, that are built-in into the ColCounter menu upon startup.
Moreover, Table 3 presents the outcomes of the common time required for every effectively and for all six wells in each the handbook counting and ColCounter strategies, demonstrating that the proposed technique is, on common, greater than 10 occasions quicker than handbook counting.
In some circumstances, acquiring photos on the decision specified within the protocol could not at all times be attainable. Due to this fact, to evaluate how the developed technique performs with photos at resolutions completely different from the perfect, the outcomes obtained from 23 wells scanned at 200 dpi, 60 wells at 300 dpi, and 25 wells at 600 dpi had been examined. The general intraclass correlation coefficient (ICC) was 0.89 (95% CI 0.81–0.93). Nevertheless, there have been variations within the ICC values based mostly on the picture decision. For photos scanned at 200 dpi, the ICC was 0.55 (95% CI 0.063–0.808). In distinction, for photos acquired at a decision of 300 dpi, the ICC was 0.96 (95% CI 0.931–0.976). In the meantime, photos scanned on the protocol-specified decision of 600 dpi exhibit an ICC of 0.99 (95% CI 0.989–0.998).
As anticipated, the Bland-Altman evaluation revealed variations within the limits of settlement relying on the acquisition decision. Table 4 presents the bias and limits of settlement, together with their 95% confidence intervals, between handbook and plugin counts for photos acquired at 200, 300 and 600 dpi. For photos with a decision of 300 dpi, the imply distinction between strategies was 2.96 colonies (0.26 to five.6), whereas for photos at 600 dpi, the imply distinction was 4.72 colonies (−5.89 to fifteen.33).
Moreover, as anticipated, the boundaries of settlement between the strategies different for completely different resolutions. Pictures at 300 dpi exhibited limits of settlement starting from 23.46 colonies (18.82 to twenty-eight.10) to −17.53 colonies (−22.17 to −12.88). In distinction, photos at 600 dpi confirmed limits of settlement spanning from 55.15 colonies (36.73 to 73.51) to −45.68 colonies (−64.07 to −27.29). It’s important to notice that these limits had been decided based mostly on the idea that the distinction between strategies follows a traditional distribution. Nevertheless, the info for photos at 300 dpi didn’t fulfill this assumption. In distinction, variations in photos at 600 dpi gave the impression to be roughly usually distributed (S1 Fig).
To delve deeper into this skewness, additional exploration revealed a big linear relationship for photos acquired at 300 dpi, indicating that the distinction between strategies barely elevated because the quantity of colonies elevated (Fig 16 A). In distinction, no important development was noticed for photos acquired at 600 dpi decision (Fig 16 B).
Fig 16. Comparability of outcomes obtained with handbook counting and utilizing the ColCounter plugin.
A. Bland-Altman plot for the comparability of each strategies utilizing 300 dpi photos. B. Bland-Altman plot for the comparability of each strategies utilizing 600 dpi photos. C. Passing-Bablok regression at 300 dpi. D. Passing-Bablok regression at 600 dpi.
In any other case, a comparability of the strategies utilizing the Passing-Bablok regression is offered in Table 5. For photos taken at a decision of 300 dpi, the intercept and regression crossed the null worth (0 and 1, respectively). Consequently, there have been no fixed or proportional variations between handbook depend and the one carried out utilizing the ColCounter plugin (Fig 16C).
In distinction, for photos at a decision of 600 dpi, the numerous intercept revealed a constant distinction between handbook counting and that performed by ColCounter, amounting to −6.66 (95% CI −12.88; −2.41). Nevertheless, because the intercept crossed the null worth of 1, this distinction seems to be non-proportional, because it didn’t escalate with the quantity of colonies (Fig 16D).
4. Dialogue
On this examine, an computerized technique for counting tumor cell colonies was developed and applied as an ImageJ plugin. The proposed technique reveals excessive reliability, which surpasses earlier functions in accuracy. As a result of acquiring photos with the specified decision isn’t at all times possible, an investigation into the efficiency of the proposed technique for low-resolution photos was performed. The outcomes revealed that the strategy maintains a excessive degree of reliability, though it’s considerably depending on picture decision. For photos with a decision of 300 dpi, the intraclass correlation coefficient (ICC) was 0.96 (95% CI 0.931–0.976), which additional improved to 0.99 (95% CI 0.989–0.998) for photos with a decision of 600 dpi.
The evaluation of imply bias and limits of settlement between handbook and automated counting strategies is offered in Table 4. The findings indicated a slight overestimation, with a imply distinction of 2.96 colonies (0.26 to five.6) in automated counting in comparison with handbook counting for photos at 300 dpi decision. For photos at 600 dpi decision, the imply distinction was 4.72 (−5.89 to fifteen.33). Though it exceeded the null worth of 0, the arrogance interval for this distinction was comparatively broad.
Moreover, the evaluation of the boundaries of settlement between strategies revealed that for photos at 300 dpi, 95% of the observations might have variations starting from 23.46 colonies (18.82 to twenty-eight.10) to −17.53 colonies (−22.17 to −12.88). These variations tended to exhibit a optimistic skew because the quantity of colonies elevated, as indicated in Fig 16. For photos acquired at 600 dpi, the boundaries of settlement had been wider. Nevertheless, on this case, the distinction gave the impression to be impartial of the quantity of colonies.
Moreover, Passing-Bablok regression evaluation demonstrated that there have been neither fixed variations (Intercept = 0) nor proportional variations (Beta = 1) between the strategies when utilizing photos at 300 dpi. Conversely, for photos at 600 dpi, a continuing distinction of −6.66 colonies (−12.88, −2.41) was noticed between handbook and automated counting. Due to this fact, on this state of affairs, the plugin persistently underestimated the quantity of colonies.
No acceptable bias vary was pre-established on this examine. Nevertheless, a suitable technique mustn’t systematically over or underestimate the colony counts. The Bland Altman plot statistics examined the choice speculation of a real bias completely different from zero, thus the rejection of the null speculation, urged a scientific bias with photos at 200 dpi and 300 dpi, however no bias with photos took a 600 dpi (p = 0.3679). Buryska et al. developed a semiautomated algorithm approach and in contrast it to microscopic handbook counting carried out by impartial observers [9]. The authors reported a robust correlation between each strategies (Pearson correlation coefficient = 0.985 and 0.965 for two observers), with limits of settlement (−50 to +60), and no important magnitude-dependent bias when evaluating the algorithm with microscopic counts, indicating good settlement between the semiautomated and handbook microscopic quantification strategies [9]. The bounds of settlement noticed in our examine had been narrower (±40) than these reported by Buryska et al. Total, these findings recommend {that a} semiautomated approach for colony counting can present dependable outcomes compared to conventional strategies.
However, Siragusa et al. [11] recognized a number of essential points in strategies counting on picture evaluation for clonogenic assays. These considerations encompassed challenges like the necessity for uniform lighting throughout imaging, efficient background identification, separating fused colonies, creating a fast calibration process, and creating user-friendly and cost-effective strategies. Numerous automated colony counting strategies have been proposed prior to now 20 years to mitigate the subjective and time-consuming points of handbook counting. One of the primary strategies “Clono-counter”, developed in Java [29], based mostly on densitometric and particular picture parameters can not right for variations in illumination generally encountered in clonogenic assays. Moreover, it could not supply consumer tunability, which limits its adaptability to particular experimental circumstances. One other instance is the strategy developed by Guzman et al. [13] referred to as “ColonyArea”, an ImageJ plugin. This approach employs native thresholding by analyzing the primary and second derivatives of the proportion space operate. Then, it estimates the proportion of floor space lined by viable colonies. Though the device gives an easy-to-use interface and is freely out there, its validation was carried out in opposition to another technique. As well as, the requirement for high-quality photos (better than 800 dpi) could restrict its broader applicability
Various options emerged, such because the one proposed by Choudhry [12], who developed one other ImageJ macro referred to as Cell Colony Edge (IMJ Edge) that demonstrated correct cell and colony detection below varied testing circumstances, together with difficult eventualities like low-contrast photos, uneven illumination, and irregularly formed cells. The IMJ Edge was in comparison with two various digital strategies, together with the one described on this paper, which confirmed excessive correlation with handbook measurements. Nevertheless, IMJ Edge does have some limitations. The strategy depends on a considerable quantity of calibration parameters to automate picture counting, making it doubtlessly advanced to arrange and fine-tune. Moreover, it’s prone to counting points on the edges of plates and should miscount fused colonies.
Later, the “CoCoNUT” colony counter was developed by the Nutech division at DTU and examined utilizing the V79 and HeLa cell traces [11]. This automated approach can be based mostly on ImageJ and structured as a macro. It depends totally on a single parameter, particularly, the radius of the smallest counted colony, which contributes to its robust edge detection potential. Nevertheless, it’s value noting that the requirement for a photographic mild field, whereas cost-effective, might introduce challenges and variations through the picture acquisition course of.
Curiously, the event of softwares for the evaluation of clonogenic assays has developed, focusing not solely on colony counts but additionally on different parameters. For instance, Militello et al. (2017) and Militello et al. (2020) developed packages that took under consideration the realm lined by colonies to calculate the survival fraction, because the evaluated therapy could not impact the quantity of colonies, however on their measurement [15,16]. Equally, the softwares countPHICS and Schiefer Counter systematically analyze colony measurement, nevertheless, not bearing in mind its proportionality with the quantity of colonies to calculate the survival fraction, however as a complement to the statistical evaluation of the assay [30,31].
Current options, akin to those developed by Gomes et al. and Qian et al., are based mostly on nuclei counting utilizing fluorescence strategies somewhat than colony measurement, contemplating that completely different remedies could affect colony morphology [32,33]. Nevertheless, the strategies are proposed for high-throughput screening. However, a promising technique based mostly on the mixture of timelapse microscopy, deep studying, and microfluidics has been developed to investigate colony formation [17], nonetheless, the mannequin was utilized to cells cultured in 3D methods, which can limit its software.
In abstract, present colony counting strategies are sometimes tailor-made to particular circumstances associated to picture acquisition, lighting, cell morphology, distinction, and uniformity. Whereas progress has been made in leveraging laptop expertise to handle these particular challenges, additional developments are wanted to create extra versatile options that may accommodate a broader vary of cell traces.
The analysis of the proposed ColCounter plugin throughout completely different picture acquisition resolutions highlights a number of key benefits. On this examine, the plugin was examined utilizing photos scanned at 200, 300, and 600 dpi to evaluate its efficiency below various acquisition circumstances. These benefits embrace affordability, user-friendliness, and the flexibility to generate high-quality outcomes. This examine considerably contributes to the event of standardized cell and colony counting strategies for clonogenic assays. It’s value noting that the use of a scanner for picture acquisition, which is a extensively out there device, enhances the strategy’s practicality and accessibility.
The event of the ColCounter plugin confirmed low price, ease of use, and high quality outcomes relating to the 2 cell traces, which helped within the growth of a standardized technique of cell and colony counting for clonogenic assays. A scanner was used as a standardized picture acquisition technique that may very well be extensively out there. The proposed technique confirmed glorious and interchangeable settlement with handbook counting utilizing photos at 300 dpi. Nevertheless, the strategy additionally confirmed some limitations; its efficiency relied on the decision of the picture acquired, and the colonies on the edges of the plate remained tough to establish by automated processing. However, the plugin included easy-to-use macros that enable customers to pick out and delete colonies for repeat counting, which will increase confidence within the technique as it may be simply edited by consultants.
5. Conclusions
Assessing the clonogenic capability of cells retained after publicity to cytotoxic brokers is a helpful device for evaluating in vitro responses to numerous therapies. This assay is crucial for exploring how cells reply to completely different cytotoxic brokers. The event of automated picture processing workflows is of paramount significance because it considerably reduces interobserver variability and enhances the reliability of clonogenic assay research. Moreover, automation streamlines the method, thereby lowering the educational curve. In our examine, we launched a novel approach for computerized colony counting that we efficiently built-in as an add-on to ImageJ software program, attaining a sensitivity of 92.88% and specificity of 92.62%.
Clonogenic assays had been used to evaluate the influence of radiation on a colorectal most cancers cell line, evaluating handbook and computerized counting strategies. Future work will give attention to optimizing the device efficiency for varied cell traces and colony morphologies. The examine of this particular cell line is especially important because of its scientific significance, rating among the many prime three worldwide in phrases of each incidence and mortality. The newly developed plugin serves as an automated counting technique and consists of handbook modifying capabilities, making it a considerable technical contribution to radiobiology analysis and the work of radiation oncology employees. This method gives an environment friendly, quick, and correct approach to quantify tumor cell colonies throughout clonogenic assays, finally advancing most cancers radiotherapy.
Acknowledgments
The authors thank the technicians of Centro de Management de Cáncer Ltda, for they assist on the irradiation course of of cell cultures, Sonia Perilla for her laboratory help and Dr. Thierry Pourcher from the TIRO Laboratory (Université Côte d’Azur) for his form donation of the HT-29 cell line. The authors additionally wish to thank all of the members of the Lún and SiBio Analysis Seedbeds for their assist and collaboration through the growth of the challenge.
References
- 1.
Evans E, Staffurth J. Rules of most cancers therapy by radiotherapy. Surgical procedure (Oxford). 2018;36(3):111–6. - 2.
Sia J, Szmyd R, Hau E, Gee HE. Molecular mechanisms of radiation-induced most cancers cell demise: A primer. Entrance Cell Dev Biol. 2020;8:41. pmid:32117972 - 3.
Franken NAP, Rodermond HM, Stap J, Haveman J, van Bree C. Clonogenic assay of cells in vitro. Nat Protoc. 2006;1(5):2315–9. pmid:17406473 - 4.
Munshi A, Hobbs M, Meyn RE. Clonogenic cell survival assay. Strategies Mol Med. 2005;110:21–8. pmid:15901923 - 5.
Brix N, Samaga D, Hennel R, Gehr Ok, Zitzelsberger H, Lauber Ok. The clonogenic assay: Robustness of plating efficiency-based evaluation is strongly compromised by mobile cooperation. Radiat Oncol. 2020;15(1):248. pmid:33121517 - 6.
Esquer H, Zhou Q, Abraham AD, LaBarbera DV. Superior Excessive-Content material-Screening Purposes of Clonogenicity in Most cancers. SLAS Discovery. 2020;25(7):734–43. - 7.
Ye R, Qiao Y, Singh PK, Wang Y, He J, Li N, et al. Excessive-Content material Clonogenic Survival Display screen to Determine Chemoradiation Sensitizers. Worldwide Journal of Radiation Oncology*Biology*Physics. 2021;111(5):e27–e37. - 8.
Giridhar P, Rath GK. Medical significance of cell survival curves. Pract Radiat Oncol. 2020;:171–5. - 9.
Buryska S, Arji S, Wuertz B, Ondrey F. Utilizing bland-altman evaluation to establish applicable clonogenic assay colony counting strategies. Technol Most cancers Res Deal with. 2023;22:15330338231214250. pmid:37997353 - 10.
Kumar Maurya D. ColonyCountJ: A user-friendly picture j add-on program for quantification of completely different colony parameters in clonogenic assay. J Clin Toxicol. 2017;07. - 11.
Siragusa M, Dall’Olio S, Fredericia PM, Jensen M, Groesser T. Cell colony counter referred to as CoCoNut. PLoS One. 2018;13(11):e0205823. pmid:30403680 - 12.
Choudhry P. Excessive-throughput technique for automated colony and cell counting by digital picture evaluation based mostly on edge detection. PLoS One. 2016;11(2):e0148469. pmid:26848849 - 13.
Guzmán C, Bagga M, Kaur A, Westermarck J, Abankwa D. ColonyArea: An ImageJ plugin to routinely quantify colony formation in clonogenic assays. PLoS One. 2014;9(3):e92444. pmid:24647355 - 14.
Roldán N, Rodriguez L, Hernandez A, Cepeda Ok, Ondo-Méndez A, Cancino Suárez SL, et al. A New Automated Most cancers Colony Forming Items Counting Technique. In: Lecture Notes in Pc Science (Together with Subseries Lecture Notes in Synthetic Intelligence and Lecture Notes in Bioinformatics), 2019. 465–72. - 15.
Militello C, Rundo L, Minafra L, Cammarata FP, Calvaruso M, Conti V, et al. MF2C3: Multi-feature fuzzy clustering to reinforce cell colony detection in automated clonogenic assay analysis. Symmetry. 2020;12(5):773. - 16.
Militello C, Rundo L, Conti V, Minafra L, Cammarata FP, Mauri G, et al. Space-based cell colony surviving fraction analysis: A novel absolutely computerized approach utilizing general-purpose acquisition {hardware}. Comput Biol Med. 2017;89:454–65. pmid:28886482 - 17.
Klett A, Raith D, Silvestrini P, Stingl M, Bermeitinger J, Sapre A, et al. Leveraging automated time-lapse microscopy coupled with deep studying to automate colony forming assay. Entrance Oncol. 2025;15:1520972. pmid:40046624 - 18.
Lavitt F, Rijlaarsdam DJ, van der Linden D, Weglarz-Tomczak E, Tomczak JM. Deep studying and switch studying for computerized cell counting in microscope photos of human most cancers cell traces. Utilized Sciences. 2021;11(11):4912. - 19.
Fogh J, Trempe G. New human tumor cell traces. Human Tumor Cells in Vitro. Springer US. 1975. 115–59. - 20.
Gerke O. Reporting requirements for a bland-altman settlement evaluation: A evaluation of methodological opinions. Diagnostics (Basel). 2020;10(5):334. pmid:32456091 - 21.
Bilić-Zulle L. Comparability of strategies: Passing and Bablok regression. Biochem Med (Zagreb). 2011;21(1):49–52. pmid:22141206 - 22.
R Core Staff. R: A language and setting for statistical computing. Vienna, Austria: R Basis for Statistical Computing. 2021. - 23.
Wickham H, Averick M, Bryan J, Chang W, McGowan L, François R, et al. Welcome to the Tidyverse. JOSS. 2019;4(43):1686. - 24.
Datta D. Blandr: A Bland-Altman Technique Comparability package deal for R. Zenodo. 2017. https://github.com/deepankardatta/blandr - 25.
Gamer M, Lemon J, Fellows I, Singh P. irr: Numerous Coefficients of Interrater Reliability and Settlement. 2019. - 26.
Manuilova E, Schuetzenmeister A, Mannequin F. mcr: Technique Comparability Regression. 2021. - 27.
Forero MG, Medina LA, Hernández NC, Morera CM. Analysis of the Hough and RANSAC strategies for the detection of circles in organic exams. Proceedings of SPIE. 2020;11510:387–95. - 28.
Geissmann Q. OpenCFU, a brand new free and open-source software program to depend cell colonies and different round objects. PLoS One. 2013;8(2):e54072. pmid:23457446 - 29.
Niyazi M, Niyazi I, Belka C. Counting colonies of clonogenic assays through the use of densitometric software program. Radiat Oncol. 2007;2:4. pmid:17212832 - 30.
Brzozowska B, Gałecki M, Tartas A, Ginter J, Kaźmierczak U, Lundholm L. Freeware device for analysing numbers and sizes of cell colonies. Radiat Environ Biophys. 2019;58(1):109–17. pmid:30673853 - 31.
Santos AF, Schiefer EM, Atherino MC, Atherino JC, Negri LH, Weffort-Santos AM, et al. Schiefer counter: An alternate technique for clonogenic assay analysis. J Pharmacol Toxicol Strategies. 2020;106:106911. pmid:32805386 - 32.
Gomes NP, Frederick B, Jacobsen JR, Chapnick D, Su TT. A Excessive Throughput Display screen with a Clonogenic Endpoint to Determine Radiation Modulators of Most cancers. Radiation Analysis. 2022;199(2):null. - 33.
Qian H, Baglamis S, Redeker F, Raaijman J, Hoebe RA, Sheraton VM, et al. Excessive-Content material and Excessive-Throughput Clonogenic Survival Assay Utilizing Fluorescence Barcoding. Cancers. 2023;15(19):4772.
![[WDAF 2026] Institutions turn to tokenization as buyers, issuers: Ondo Finance](https://cryptonews100.com/wp-content/uploads/2026/11/news-p.v1.20261109.92fd66e07214413db4355da131b5277f_P1-350x250.jpg)
