All posts in Our Blog

The elephants in the room

Categories: Our Blog
Comments: No

Last week the Digital Pathology Association (“DPA”) announced that the ‘de novo’ application process may be acceptable by the FDA for clearance of whole slide imaging devices (“WSI” devices) for primary diagnostic uses (see this article from the Digital Pathology Blog for more information).  Why is this significant?  The de novo process lowers the primary hurdle for marketing approval from that of a class III device to essentially a class II device. But WSI device manufacturers aren’t out of the woods yet.  There are other proverbial “elephants in the room” with which they’ll have to contend. Batter up With medical devices, the manufacturer who is first-to-market generally bears the cost burden to conduct the clinical trials.  But somebody has to submit first.  Will WSI device manufacturers actually cooperate to share the high cost of the first approval, speeding approval of all WSI device applications? Where have all the pathologists gone? Read more

5 Keys to Better Images

Categories: Our Blog
Comments: No

It never ceases to amaze me how easy it is to take digital images with microscopes, yet the resulting images can be so poor. Just take a stroll through the poster area at a scientific conference, or flip through a scientific journal (peer-reviewed or not), and the evidence is obvious – poor color, artifacts, and inconsistent appearance in image panels. Here are five basic keys that we should all be using when we take color images on a microscope (and some apply to fluorescence, too). Köhler Alignment Basically, Köhler alignment (or Köhler illumination, or simply “Köhler”) optimizes the illumination of your specimen on the microscope. This simple process aligns the illumination of your microscope, and ensures the optimal balance of resolution and contrast. Focus on a specimen. Close the field diaphragm until it is visible in the field of view. Use the condenser-focusing knob located behind the condenser on the Read more

You’ve captured images across multiple sessions, under various user-defined settings, and perhaps on different equipment or in different labs.  Maybe you even received images from multiple sources.  The major challenge now is getting uniformity in the images for evaluation, comparison or analysis, or for presentation, publication or reports.  Sure, you can use Adobe® Photoshop® and other image editing software, but that takes time and effort. With Datacolor CHROMACAL™ image software, we offer you the flexibility to select the tool you prefer: EZ Image Prep™ for white balancing, brightness matching and flatfield-correction, or full color calibration (which also includes white balancing and brightness matching).  Both solutions are provided within in a simple-to-use, user-friendly, one-step batch processing environment.  With either solution, CHROMACAL delivers remarkable results (Figure 1). Primarily interested in consistent white-balance and brightness matching?  Choose the EZ Image Prep feature built into CHROMACAL. With this solution, you don’t need a CHROMACAL Read more

When we consider whole slide imaging (“WSI”) devices, we expect high performance in both speed and image quality, together with consistency over time and among devices from the same manufacturer.  Quality control measures employed during manufacturing certainly contribute to performance and consistency, but what is being done to ensure accuracy of the output?  After all, it’s the image that matters for diagnosis and records, right? The FDA will soon be issuing guidelines for WSI device technical assessments.  These assessments will be required for any device that will be promoted for use in digital pathology, and may be considered a benchmark for traditional imaging systems.  In fact, the guidelines may be adopted more widely for any imaging instrumentation used for research, including slide scanners. Upon examination of the assessment categories in the draft FDA guidance (see the document HERE), the first tier of assessments would be performed at the individual component Read more

Obstacles Facing Adoption of Digital Pathology

Categories: Our Blog
Comments Off on Obstacles Facing Adoption of Digital Pathology

A recent post to LinkedIn (http://bit.ly/1MKpg8E), highlights that transitioning from traditional microscope-based pathology to digital pathology in the clinical setting (or even for research) is no simple matter.  This observation was echoed by Dr. Keith Kaplan on tissuepathology.com (http://bit.ly/1RXzfFF).  As a microscope user and enthusiast for over 3 decades, it’s going to take more than just the allure of cost savings and convenience to convince me to give up my trusty, dependable scope. In the healthcare arena, cost, speed, efficiency and quality of care matter far beyond our imaginations.  But these are only hypothetical considering that the “product” must first deliver on expectations (“fully baked” as some may say).  In this case, the first expectation is for a quality image.  Only once a quality image is established can we begin to recognize benefits in workflow, efficiency, analysis, records, and ultimately patient care.  But at this time there is no standardization Read more

Color Infidelity (reposted)

Categories: Our Blog
Comments: No

(reposted with permission from Mike Tjepkema, W. Nuhsbaum, Inc. View the original here) Color infidelity: Why using a light source incorrectly is cheating on your data Knowledge is power when minimizing error in imaging Maybe you are one of the people who engaged in arguments about whether the dress is white and gold or blue and black. Or maybe you are just a regular person trying to take a picture of a dress to share it with your friends. Either way, color is important! With microscopes, most users of digital cameras are familiar with the simple process of going through a white balance procedure to get a very nice white background in images. However, many users often notice that although the background is white, the color of their image is incorrect. This leads to arguments, although less viral than “the dress,” about the true color of a sample! Through the Read more

It’s pretty simple – you just want a good image, and in this day and age, we expect that our technology is smart enough to give us what we want.  After all, our cell phones take pretty good pictures, so why is color imaging on a microscope any different?  It comes down to the intended purpose of the images.  While cell phones are great for capturing the moment with family or friends, scientific-grade cameras are about precision and accuracy, not selfies. One of the first steps in color imaging is to white balance the camera.  Contrary to its name. white balancing is intended to make neutral objects, such as the area around a piece of tissue, appear neutral grey (NOT white).  The figure to the right (courtesy of BitesizeBio.com) shows the histogram of an “empty” area on a slide before (left) and after (right) white balancing.  In this example, the Read more

Ever since being dubbed the resident microscopy guru at a long-since merged and absorbed pharma company, I’ve been advocating for good microscopy practices.  It makes sense afterall – manage what you can control, and you can greatly improve the quality and consistency of your images.  Theoretically, the remaining variation results from changes in experimental outcomes.  Without solid processes and controls, there is truth in the old adage, “Garbage in, garbage out.” Yesterday I attended “Getting Correct Color from Your Camera,” a webinar hosted by BitesizeBio (bitesizebio.com) and sponsored by Lumenera Corporation (www.lumenera.com).  I am surprised that researchers continue to struggle with microscopy and images, and I was delighted to see that there is still interest in developing and following good imaging practices.  During the brief 30-minute presentation, the speaker reviews everything from microscope set-up (i.e. Köhler alignment) to camera acquisition settings, and image calibration to monitor calibration.  View the webinar Read more

You probably already have a good idea of what “color” means.  By definition, “integrity” equates to honesty; being whole; and a sound, unimpaired, perfect condition.  Sounds utopian – well, at least it’s something to strive for.  The concept of color integrity relates to the accurate representation of color in an image, and “image” must include both the digital image file (what I call bits and bytes) AND the display of the image file AND visualization of that image on a computer monitor.  So, you have an expensive microscope camera and a new(er) monitor.  No problem, right?  Not so fast.

Metamerism and the Microscope

Categories: Our Blog
Comments: No

There are some great articles available that discuss the importance of color reproduction in digital microscope images (here’s a good, succinct article: http://ow.ly/Rc221).  However, I have yet to see a discussion on a phenomenon called METAMERISM, and this is a challenge that we all experience every day.

Quality control in microscope images (whether from whole slide scanners or traditional microscopes) is paramount to accurate analysis, especially across studies, imaging devices and repositories.  Color correction (including methods such as color normalization, color standardization and color calibration) has been shown to accommodate for color batch effects, and is especially beneficial for automated analysis and computer-assisted diagnosis. Most approaches to correcting color variation as a batch artifact are found only in the literature, but commercial solutions are now available, and can be adapted to qualify or even calibrate acquisition devices, as is the proposal by the FDA for digital pathology.   How do you ensure that you’re getting the best information from your images?  Share a Comment below.   Click Here for a list of references.

In IBM’s announcement of its acquisition of Merge Healthcare (http://www-03.ibm.com/press/us/en/pressrelease/47435.wss), there are several notable points that highlight the importance, along with the challenges, of advancing medical imaging: “Medical images are by far the largest and fastest-growing data source in the healthcare industry and perhaps the world – IBM researchers estimate that they account for at least 90% of all medical data today – but they also present challenges that need to be addressed.” “Medical images are some of the most complicated data sets imaginable, and there is perhaps no more important area in which researchers can apply machine learning and cognitive computing. That’s the real promise of cognitive computing and its artificial intelligence components – helping to make us healthier and to improve the quality of our lives.” For the promise of machine learning and artificial intelligence to be realized in medical imaging, it remains critical that imaging standards are Read more

Color in medical images will soon be standardized for display, just as radiological images are today.  The display of images, however, is only part of the equation.  The images themselves – the data behind what we see – should also accurately represent the specimen so that there is complete confidence in our observations. Yes medicine seems to be the central focus for color standards in imaging, and options exist today for researchers to calibrate computer monitors, and equally important to calibrate microscope images.  It is through the collaboration of organizations like the ICC Medical Imaging Working Group, the FDA and the ISO Technical Committees that healthcare and research alike benefit from standardization. http://www.auntminnie.com//index.aspx?sec=sup&sub=pac&pag=dis&ItemID=111567&wf=1236 Join the conversation.  Leave a Reply below.

Inconsistency in Imaging Systems

Categories: Our Blog
Comments: 2

Microscope imaging systems are not alike and give different results, even if configured identically.  There is little, if any, industry standardization relating to the equipment, software or performance, so how could we ever expect to compare images among imaging sessions let alone among imaging systems? Attendees at the Summit on Color in Medical Imaging (May 8-9, 2013, held at the FDA) agreed that challenges relating to color – and consistency – exist across research and medical disciplines, and impact evaluation, interpretation and, in some cases, diagnostic decisions.  Representing academia, industry and medical practice, the authors found that a serious need exists to manage and control the color in both our imaging systems and displays. Collaboration, however, is occurring throughout the industry.  The International Color Consortium (“ICC”) formed the ICC Medical Imaging Working group to explore solutions to these issues.  In advance of releasing its guidance on digital pathology, the FDA Read more

How important are your images?  We take time and care to capture the best images we can from our microscope systems, but did we really get it right?  If we’re not using a calibrated monitor, we just don’t know.  The image below is courtesy of Dr. Stefan Hamann (Senior Scientist, Translation Pathology Laboratory, Biogen Inc.) and clearly illustrates the disparity between three new(er) monitors displaying the same image (same computer, same graphics card, and two monitors are even the same model purchased at the same time!).  Which displayed image is right? Objective monitor calibration (using a colorimeter or spectrophotometer) is the best way to ensure that our monitors are using their full color and tonal rendering capability.  Yes, there are “calibration” options in our computer’s operating system, and this is akin to adjusting color on our televisions – moving sliders until the color “looks” right – but this approach is Read more

A recognized expert in microscopy and image analysis, who also happens to give workshops on using Photoshop for scientific images, recently shared his experience with a new color image standardization method. “[CHROMACAL] is the first tool in digital microscopy for brightfield that fulfills an industry-wide need for image standardization, and it should be in every lab that is serious about good imaging and science.” It is common practice — even acceptable — for researchers to “tweak” their scientific images using Photoshop, but this method is cumbersome, subjective, adjustments generally go undocumented and hence not repeatable, and the whole process is not very scientific.  While evaluating CHROMACAL’s benefits for visualizing image detail, this expert discovered that CHROMACAL had advantages over using Photoshop for scientific images. Read the case study for yourself to see why he feels CHROMACAL outperforms Photoshop for objectivity, consistency and usability: http://scientific.datacolor.com/wp-content/uploads/2015/07/Case-Study-Outperforms-Photoshop-Sedgewick.pdf

Now follow CHROMACAL on Facebook

Categories: News, Our Blog
Comments Off on Now follow CHROMACAL on Facebook

Follow the CHROMACAL community on Facebook for the latest information on digital imaging, image quality, color consistency and, of course, CHROMACAL!  

In Parts 1 and 2 of this series, a few brief examples were presented illustrating the importance of image consistency and color standardization for confidence in both image comparison and image analysis. Let’s now consider image visualization.  But before we get started, we have to assume that the data (the bits and bytes of the image) are an accurate description of the specimen.  To visualize an image, we must use viewer software and a display device (a.k.a. computer monitor).  What is often overlooked is that different software interpret the data differently and, in some cases, the difference in appearance is remarkable [1-3]. Assuming that the data is accurate (e.g. perfectly captures the essence of our specimen) and we have total confidence in the viewing software, how do we know we’re seeing the image accurately?  Just because we’re using a new monitor doesn’t mean that the color and tonal display capabilities Read more

In Part 1 of this series, I presented some compelling examples illustrating that our current digital images still fall short of the gold standard – the optical microscope.  This second installment of the series highlights the impact of image quality on image analysis. Improve Automated Analysis with Image Consistency Regardless of the study, the evidence demonstrates that you can’t rely on your instrumentation to deliver quality and consistent images (Figure 1, from Badano et al) [1].  Images are impacted not only by the acquisition device, but also by the acquisition and viewing software, and the device with which you visualize the images.  In fact, color monitors suffer from instabilities that result in color shift, contrast reduction and a reduction of the total viewable colors [2]. It stands to reason that image consistency would enable more precise and accurate analysis and comparison of images.  In fact, several examples in the literature Read more

It is absolutely critical that forensic analysis of a crime scene and trace evidence be conducted with utmost diligence.  We’ve all heard of cases in which sloppy evidence collection or presentation resulted in a criminal going free, as is often the subject of popular forensic science dramas on television.  In scientific research, it has become acceptable to “adjust” images for consistency using consumer-grade image editing software (such as that offered by Adobe Systems Inc.) – the stipulation, however, is that any manipulation of the images must be disclosed and described to your audience so that they can reproduce the experiments.  But such manipulation is not generally acceptable in forensics.  Why?  Because image processing generally involves human bias and subjectivity, adjusting image data based on impression and perception.  In forensics where admissibility of images is of concern, objectivity is paramount. A forensic scientist at Indiana University Purdue University Indianapolis (IUPUI) recently Read more

The other week I gave an invited talk at a biotechnology company at which I spoke about the digital imaging environment as it exists today.  As I prepared for the presentation, I began a review of the literature and found numerous examples where image quality and image consistency directly contributed to sometimes spurious results.  There is considerable concern regarding image quality for digital pathology, however these challenges also apply to traditional and non-clinical research applications.  In most if not all cases, Datacolor CHROMACAL would have provided greater image consistency, leading to improvements in visualization, evaluation and image analysis.  I would like to take a moment and share some of those observations. The Current Digital Environment is Inferior to the Optical Standard Despite the advances in imaging technology, digital images still fail to meet the expectations of researchers, scientists, clinicians and physicians who rely on their perceptive skills to identify features Read more

Reproducibility in Research Matters

Categories: Our Blog
Comments: No

Earlier this week, PLoS Biology published an article that explores the economic impact of irreproducible research results [1].  Based on their analysis, the authors claim that approximately $28 billion dollars annually are spent to deliver results that cannot be replicated.  As Stefano Bertuzzi, ASCB Executive Director, pointed out in his rebuttal [2], such a claim would also implicate nearly half of the published papers as being irreproducible.  Importantly, Bertuzzi correctly identifies that science is an evolving process, continuously improving, and that the PLoS Biology paper failed to accurately capture this concept. Although the economic figures can certainly be debated, a key takeaway is that scientific reproducibility is important and better approaches must be utilized to control and document the processes.  In the area of image reproducibility in scientific research, new methods like ChromaCal should be considered for delivering consistency, reproducibility, and quality in images.   [1] Leonard P. Freedman, Iain Read more

A recent article in the Wall Street Journal [1] really lit up the social media scene.  It was tweeted and retweeted, and appeared in numerous blogs with plenty of reader commentary, and was front and center in my LinkedIn Pulse.  Contributors to the chatter ranged from pathologists to politicians, and product marketers to patients. In its simplest form, digital pathology is the use of digital images of patient specimens (i.e. biopsy, smear, etc.) for diagnosis or, at least in the US, for consultation, education and training.  It is certainly much easier, faster, and more cost effective to send a digital image than it is to send a glass microscope slide via snail mail to another city or even state to permit a second pair of eyes to render their opinion.  Technology is improving to allow these images to more accurately represent the original slide in all aspects (color, contrast, even Read more

This week, FDA released a draft guidance document entitled “Technical Performance Assessment of Digital Pathology Whole Slide Imaging Devices“.  As they have previously implemented in the digital radiology area, FDA is moving to establish standards for color integrity and color consistency in WSI systems.  We are all aware of the variability in color rendering within WSI equipment models as well as between the different manufacturers’ equipment. The volume of images that can be produced by these systems, and the efficiency that can be realized with their implementation are both tremendous.  And the potential impact to the workflow for evaluation, analysis and assessment is equally significant.  So more rigor in the calibration and color standardization of these systems just makes sense.  And we also need tools to test that the systems remain in compliance, tools that should be built into the WSI systems and also available to WSI users so that Read more

Social media is abuzz about whether #thedress is white/gold or black/blue dress (Science Can Tell You the Color of the Dress).  It just goes to show that our eyes are not reliable. So, if you find yourself editing color images in software like Photoshop or ImageJ, you are not only at the mercy of the quality of your monitor, but you are also vulnerable to the color interpretation by your eyes. That’s why it is critical to use an objective color calibration solution to standardize your color.  For science, medicine and research, your choices are limited.  Yes, several vendors offer monitor calibration solutions, but that single aspect falls short of solving the problem.  You need to ensure color consistency and objectivity by addressing your monitor and your digital images. It’s already been proven that scientific cameras are inconsistent, and further prone to color anomalies caused by user-controlled settings.  Even whole Read more

Recently published online with Bioscience Technology, “Imaging and Analysis with Flying Colors: Part One” explores variables in the imaging process that contribute to image inconsistency, and reviews current methods in use to improve color reproduction in brightfield images.  The author’s narrative includes the following on the importance of ensuring the scientific integrity of digital images: “Researchers are facing increasing demands from colleagues, peers and publishers for process documentation including adequate controls, and for extensive documentation of experimental parameters. Without such consideration, there would be little chance to repeat, or even validate, findings. [Microscope] images are visual representations of experimental outcomes and, therefore, should be considered and treated as data. In today’s research climate, scientists should anticipate and be prepared to accommodate heightened demands for scientific integrity by providing enhanced rigor in the capture, processing and communication of images.”

On Nov 20, 2014 at 1pm ET, join Dr. Keith Kaplan for a review of issues with color in histology and histopathology, sources of color variation, implications on medical and research practice, and possible directions for resolution.  Click here to register About the speaker: Keith J. Kaplan, MD, is a practicing pathologist and laboratory medical director in North Carolina. He is board certified in anatomic and clinical pathology. Dr. Kaplan currently serves as a member of the College of American Pathologists, American Society of Clinical Pathology and the American Society of Cytopathology as well as the American Pathology Foundation. He is an executive board member of the American Pathology Foundation. Dr. Kaplan is the publisher of the Digital Pathology Blog at tissuepathology.com, the industry’s leading pathology blog.

In a recent blog posting on ParticleImaging.com, Lew Brown (Technical Director at Fluid Imaging Technologies, Inc.) writes: “Color to us is subjective in general, but in digital imaging we have the benefit of numbers to describe color in quantitative terms.  So if those numbers used to describe color are calibrated to a known standard, those numbers now describe color in a completely quantitative way, and can be used for legitimate comparisons of images.” Read more of Mr. Brown’s insights here

In an article published in the October 2014 issue of BioPhotonics, authors Mark Clymer and Dr. Eduardo Rosa-Molinar discuss how establishing a color standard could allow microscopists to verify and reproduce results.  In summary, they comment: “With little consistency among imaging systems today, the development of color-management standards for research imaging is highly desirable.  When scientists eliminate unpredictable variation that occurs during image capture and display, they reduce an unwanted variable in their experiments and analysis.  Color calibration in bright-field microscopy is becoming a vital step in this process.” And for those interested in learning more about the science behind color science, the article includes a sidebar entitled “Color science in black and white” by Dr. Hong Wei.

In this webinar by Jerry Sedgewick (Photoshop expert, image analysis consultant and previous core facility manager), Mr. Sedgewick delivers some bottom-line advice on how to standardize the color in your microscopy images, while maintaining linearity, for greater confidence in quantitative analysis.  He also highlights a new diagnostic tool to check whether your imaging system is capturing scientific images that are linear.  An introduction is provided by Michael Linden, MD, PhD (University of Minnesota) who shares his insights regarding Datacolor ChromaCal.

In this webinar presented by Jerry Sedgewick (Photoshop expert, image analysis consultant and previous core facility manager), he offers an alternative to Photoshop for color brightfield images.  Mr. Sedgewick compares post-processing using a solution (calibration slide and software) provided by a company known for color calibration (Datacolor, Inc., NJ) against common methods for post-processing in Photoshop. Mr. Sedgewick discusses how this solution (ChromaCal) also prevents saving over original images, provides metadata to track what has been done, and tags the image (if desired) to show it has been post-processed in ChromaCal.  In addition, he reviews ChromaCal’s ability to automatically white balance, adjust brightness to a target value, and standardize color to a known set of colors along incremental wavelengths in the visible range sans user interaction.

Why be linear?

Categories: Our Blog
Comments: No

As a photographer, one of the most appreciated change from film to digital became one of the most annoying. It’s all related to that little screen that comes with consumer and professional DSLR cameras. While it was over-the-top gratifying to see the picture I just took rather than nervously waiting for results after the film was processed and printed, the screen image always came through too bright. No matter how I turned down the brightness of the screen, the picture would not provide me the deeper tones I wanted to see. From a camera manufacturer perspective, this makes sense. It’s likely that you would not want any user of your cameras to complain about not being able to see a darker feature in the image. But the absence of the deeper tones and colors removes contrast, and it’s contrast that gives depth and liveliness to a picture. For the users Read more

Incorrect Color as Artifact

Categories: Our Blog
Comments: No

At a recent immunohistochemistry and microscopy course in Woods Hole, MA, a speaker began his presentation with a compelling perspective. The speaker was an experienced microscopist from a leading microscopy company who started his career in Electron Microscopy (EM). He noted that in the EM world the focus was always on image artifacts: defects that could be seen in the image, often caused by poor specimen preparation and by issues with the microscope. In that world, microscopists labored over ways to prevent artifacts, sometimes spending days and weeks to eliminate artifacts. Then he changed fields and entered into the light microscopy world. In that shift to light microscopy he encountered a turn around: a complete lack of attention to artifacts. To this day he remains puzzled by the lack of attention to artifacts among those who use light microscopes. From the point of view of an expert in microscopy, defects Read more

Problems with Color Perception

Categories: Our Blog
Comments: No

Even when all the color controls are in place—we’ve white balanced the camera, calibrated the image color to a standard, and are viewing on calibrated monitors—human perception still gets in the way. We may think we are evaluating an image’s colors correctly, but we’re not. That’s because we rarely focus on color to see it accurately. Seldom do we view a color by itself: it’s always against some kind of background tone, color or number of colors. We view a color that is lit by a light source that can affect the hue, shifting the color to warmer colors (yellows, reds) when the light source is tungsten, or to cooler colors (blues, greens) when you stand under the shade of a tree (see color temperature blog, White Balancing and Accurate Color: http://color-integrity.com/?p=128). And we view a color that may be lit by transmitted light, which adds to its contrast and Read more

White Balancing and Accurate Color

Categories: Our Blog
Comments: No

White balancing can be misinterpreted. Without understanding why an image should be white balanced, it could be misinterpreted as making the background of color brightfield images white, versus the real reason for white balancing: to correct for overall hue shifts, and, in the process, create white and gray areas that are “neutral.” In other words, whites look white and grays look gray versus whites and grays that look a little yellow, blue, red or green. First, it’s useful to understand that human vision does not operate like a camera: even when there are colors in white objects, we don’t see them. Even at dusk when the sky is red we can still see white objects. We perceive the object as white, when the white object is really reddish. This is multiplied by numerous settings in which we automatically perceive white objects when whites are colored: a room illuminated by tungsten Read more

Most scientists and medical practitioners take the appearance of images for granted when these are seen on computer screens. We have an intrinsic trust that the way an image appears on the screen is an accurate representation of the image itself. How easily we are fooled. Even when the computer display is calibrated by a hardware device, the application software that you are using can make an image appear in ways that can be deceiving. And yet, to the eye, images look “just fine.” But sometimes “just fine” isn’t fine at all. Software can alter the appearance of an image in 3 substantial ways: A) by showing images that are non-pixelated at high zoom, B) by brightening images through auto-scaling, and C) by brightening images when displaying a only a portion of the full tonal range represented by the file format. Non-pixelated. Even when zooming in, software can introduce a Read more

The Impact of Color Memory

Categories: Our Blog
Comments: No

Years ago at the beginnings of the digital imaging era, I tried to convince an Orthopedist to digitally photograph his x-ray films. Without any hesitation, he told me the digital resolution was not good enough, so I set out to prove him wrong. I digitally imaged an x-ray film of a ribcage and then sent the image to him. I received an email back: inadequate resolution. So I looked carefully at the film, then looked at the image on my monitor and I compared the two. It dawned on me that the film was bluish, and so I introduced the same bluish tone into the image and sent it to him with only that change. I received an email back. He said, “I don’t know what you did to the resolution, but this image is well resolved and it exceeds what I ever thought possible.” It was then that the Read more

Consistent Color for Image Analysis

Categories: Our Blog
Comments: 1

Humans are really good at workarounds.  We can find new ways to avoid highly trafficked roads, and workarounds for a software program that doesn’t quite do what we want. But some things shouldn’t require workarounds.  That is especially true with image analysis of color specimens from microscopes.  But in microscopy, the rationale used to justify the workaround is this:  because colors are so often inconsistent, something in the image analysis software has to compensate for the inconsistency, or a human must intervene to make image by image decisions.  The former can be riddled with hours spent trying to get the software to work; the latter reeks of human bias when it comes to scientific research. For those new to image analysis–also known as image quantitation, quantization and machine analysis—this is the way in which a computer automatically detects important areas of the image and then spits out data, or compares Read more

A Room with a View… of Your Monitor

Categories: Our Blog
Comments: No

When getting a monitor calibration device, it’s easy to feel like a kid in a candy shop. You can’t just stop at calibrating the first monitor, but every last monitor in the room needs to be calibrated. It’s one of those things you never knew you missed until it’s in your hands. For those who are in labs and clinics this is especially true. After one monitor is calibrated, everyone’s computer is included. Then, suddenly, everyone is seeing the same images, and it becomes a wonder that a calibration device wasn’t something that was purchased with the first monitor. It is a wonder. When images convey scientific and clinical proof of various conditions, it would seem important that everyone in the lab or clinic see the same view of the images. Yet, in both the scientific and medical worlds, the idea of being able to see the same colors on Read more

Accurate Color: Nothing to Blink At

Categories: Our Blog
Comments: No

In the book “Blink,” the author describes how years of experience results in the ability to make an intuitive judgment in the blink of an eye. But the author doesn’t discuss what it takes to make several hundred-to-thousands of these judgments every day, one “blink” for every judgment. He doesn’t address a lot of blinking. In the world of science and medicine, that level of blinking can be commonplace. For those who read from digital images, a dermatologist once reported that 50 judgments can be made in a single minute, and these judgments can continue for an 8-hour day. But that’s on those days that only last 8 hours. Many days for scientists and practitioners can last 12- to 15-hours. Think of the excessive amounts of blinks–with associated neuronal sparks–in a longer day, and how easy it can be to blink yourself into fatigue and exhaustion. After one or many Read more

In the world of light microscopy, chromogenic-stained samples stand apart from others when it comes to using a microscope and camera. No other kind of sample, it seems, requires so much hand/eye interaction, and so much back and forth between the microscope and computer to get a decent image. For that reason alone, it is fertile ground for mistakes, and especially for those new to the field. At each microscopy session, color brightfield images typically suffer from incorrect overall coloring, over-exposure, uneven illumination, and incorrect appearance of the sample itself. At a big picture level, color brightfield images suffer inconsistency from session-to-session, and inconsistency from camera-to-camera. Here are the common mistakes―and solutions, although only five out of six mistakes can be solved today. 1. Overexposure. Never over-expose (a.k.a., saturate, or over-saturate) images. Why not? Three logical reasons, and a fourth that helps you get along with others: Because image information Read more

For the tens-of-thousands of dollars invested in state-of-the-art color imaging systems found in scientific labs and medical centers, it would seem absurd that little to no attention has been paid to color integrity.  After all, we rely on both medicine and science to give us honest and impartial interpretations of conditions.  If scientists and medical doctors are taking color pictures of these conditions, wouldn’t they follow suit and provide not just images that are focused and have adequate contrast, but colors that are consistent and closer to that of the specimen? Currently, some obstacles stand in the way.  For example, it is difficult to obtain colors close to that of the specimen, along with consistent colors, without a calibration standard.  For microscopists in particular, this is a problem because microscopic calibration standards are not currently available.  These are desperately needed because cameras from the various manufacturers each contain proprietary color Read more