Using Digital Chest Images to Monitor the Health of Coal Miners and Other Workers

Posted on by Michael Attfield, PhD, and David Weissman, MD

doctors examine a chest x-ray

Under the Federal Coal Mine Health and Safety Act of 1969, as amended by the Federal Mine Safety and Health Act of 1977, NIOSH administers a long-standing chest radiographic (x-ray) monitoring program for underground coal miners (The Coal Workers X-ray Surveillance Program [CWXSP]). Conventional screen-film chest radiographic imaging has been an indispensable tool for monitoring the lung health of miners and other dust-exposed workers. In these surveillance programs, trained readers assess a worker’s chest radiograph for the presence and severity of occupational lung disease (parenchymal a, pleural b, and other abnormalities) using a classification system developed by the International Labour Office (ILO). The ILO classification system was designed to systematically record the abnormalities seen on standard screen-film chest radiographs that occur as a result of dust inhalation. It requires the reader to compare any abnormalities observed on the worker’s radiograph to a set of standard radiographs exemplifying various types and severity of dust diseases, provided as hard copy films by the ILO.

NIOSH is also involved in the training and certification of physicians to use the ILO system through the B Reader Program. The B Reader Program aims to ensure competency by testing the ability of applicant readers to classify a test-set of radiographs. NIOSH has long experience with the B Reader Program and with employing the ILO system for monitoring, research, and other activities using radiographic assessment. NIOSH continues to have an extensive impact on national and international practices in using chest imaging for workforce medical screening, research, hazard evaluations, clinical practice, compensation, and standard setting in occupational respiratory disease.

Increasingly, medical imaging and patient information systems are utilizing digital approaches for data acquisition, transmission, storage, display, and interpretation. Over the next decade, conventional film-based radiography will be completely replaced by digital radiography systems in the United States and elsewhere. NIOSH is in the process of adopting and enabling the application of digital chest imaging to surveillance for occupational lung disorders. This will benefit programs assessing occupational lung disease both nationally and internationally. NIOSH is working together with partner organizations such as the American College of Radiology and the International Labour Office.

There are many advantages to using modern digital imaging systems for monitoring miners’ lung health, including:

  • More consistent image quality
  • Faster results
  • Increased ability to share images with multiple readers
  • Simplified storage of images
  • Reduced risk for technicians and the environment due to the elimination of chemicals for developing film

Digital diagnostic imaging is rapidly being adopted world-wide, including by many developing countries. Thus, there is international demand for adaptation of the ILO classification system to a digital format. Peer-reviewed research findings indicate that digital images can be used in the classification of work-related parenchymal disorders, providing results equivalent to conventional film-based radiographs.1 Further information on the science and application of digital imaging to assessing occupational respiratory disease is available from a NIOSH workshop. Until the ILO endorses the use of digital standards, however, readers must continue to use the current ILO reference films and conventional chest radiographs for classifying using the ILO system.

As we move forward with this conversion, NIOSH would like to hear from you on aspects of the transition to digital imaging that concern you, or about which you have questions or comments. Particular issues might include:

  1. How digital and film-based ILO readings compare
  2. Effect of image manipulation on readings and need for standardization
  3. Digital implications for black lung clinics and other medical facilities
  4. Implications for the B Reader Program
  5. Implications for the ILO system
  6. Conversion of x-ray film images to digital media
  7. Ethical practices for classification of digital images
  8. Infrastructure (computer, monitor, etc.) needs for digital classification

Dr. Attfield is a Senior Scientist in the Surveillance Branch of the NIOSH Division of Respiratory Disease Studies.

Dr. Weissman is Director of the NIOSH Division of Respiratory Disease Studies.


a. The lung parenchyma comprises the main tissue of the lung where gas exchange takes place.

b. The pleura is the membrane that lines the lung.


  1. Franzblau A, Kazerooni EA, Sen A, Goodsitt MM, Lee SY, Rosenman KD, Lockey JE, Meyer CA, Gillespie BW, Petsonk EL, Wang ML. Comparison of Digital Radiographs with Film Radiographs for the Classification of Pneumoconiosis. Acad Radiol. 2009 Apr 1. [Epub ahead of print]
Posted on by Michael Attfield, PhD, and David Weissman, MDTags

15 comments on “Using Digital Chest Images to Monitor the Health of Coal Miners and Other Workers”

Comments listed below are posted by individuals not associated with CDC, unless otherwise stated. These comments do not represent the official views of CDC, and CDC does not guarantee that any information posted by individuals on this site is correct, and disclaims any liability for any loss or damage resulting from reliance on any such information. Read more about our comment policy ».

    The above questions outlined in your article would be of great importance to me as a senior claims examiner with the Department of Labor/Federal Black Lung Program. In addition, when a chest x-ray is taken on a digital machine is the image transferable immediately to a disc or from the machine via e-mail, etc. to (for instance) the Black Lung District Office?

    If so, would it be a viable, reliable, method of re-reading a digital x-ray to forward via e-mail the digital x-ray to a B-read for re-reading?

    I think digitizing chest x-rays is meritorious within the context that that our colleagues have described, but is akin to using a 21st century cleaning solution to get the dust off the Victrola’s vaccuum tube.

    The entire “B” reading system is of greatest use to the lawyers and to help document the damage which has occurred to populations in the previous decades.

    If one wishes to detect asbestos-related disease early, then one is served best by employing the 64 slice or greater high-resolution CT.

    If one wishes to detect disease with greater certitude, then enhanced programs for interpretation for the CT unit can also be employed.

    My own feeling is, obviously, that fewer resources should be spent on this particular brand of “sitting the ICU patient up in bed and giving him a shave” and more resources spent on developing technologically current approaches to detecting early disease which will do more than document asbestosis which began 15 to 40 years prior to the radiograph, digital or not.

    We thank Dr. Harbut for his thoughtful comments. First, we would like to clear up a possible misunderstanding in that our principal interest is in digital images and not directly in digitized film x-rays. While digitizing film x-rays may be necessary on occasion, when side-by-side or other assessment of old chest films is called for, this is not our primary focus. Rather, our interest is in transitioning to acquiring and using digital chest images to identify and evaluate occupational lung disease, often by classifying them according to the International Labour Office (ILO) classification system.

    We agree with Dr. Harbut that CT and other advanced imaging techniques, when used for the purpose of clinical diagnosis, offer the potential for more sensitive identification of the pneumoconioses in individuals. However, for the purposes of screening and surveillance of working populations, the standard chest image remains the modality of choice. CT scans cause far more exposure to radiation than standard chest images. It is unclear that sufficient benefit would result from exposing workers, many of whom would not have occupational disease, to the levels of radiation required by CT imaging, to justify the use of that approach. Also, unlike chest radiographs, there are no widely accepted, validated international standards for accurately and reliably quantifying CT changes of pneumoconiosis. Finally, using advanced imaging techniques to screen populations would be prohibitively expensive for many organizations. This is not to say at all that we at NIOSH rule out use of advanced imaging techniques. However, further work is necessary in order to develop methods and to identify situations where such approaches can be justified. We will be pursuing such lines of thinking once we have dealt with our immediate concerns with digital radiography.

    In the Federal Black Lung program, there is a widespread perception that there is bias on the part of B-readers.

    The defendant/employers usually hire the same B-readers, who seem to consistently read x-rays as negative. There are three B-readers in Baltimore who are routinely hired by employers for reading in contested claims.

    In several recent cases with which I an aware, one of these Baltimore B-readers has opined that coal workers’ pneumoconiosis has become a very rare disease, “virtually non-existent disease after World War II.” This B-reader further states that coal workers’ pneumoconiosis is very rare because the coal mines that were operated after World War II are relatively dust free compared to the coal mines operated in the era prior to WW II. It is unclear that this B-reader’s opinion relies upon coal mine dust surveys and coal workers’ pneumoconiosis epidemiologic surveys.

    I am unaware of any such surveys that would have measurements going back some six or seven decades. This B-reader’s opinion on the epidemiology of coal workers’ pneumoconiosis seems to be inconsistent with more recent studies regarding the prevalence of coal workers’ pneumoconiosis over time. I am unaware of any such studies that describe the prevalence of coal workers’ pneumoconiosis as “virtually nonexistent.”

    This employer’s B-reader opinion gives rise to the perception in the claimant community that the employers’ B-readers are biased.

    No doubt the employer/defendants have their views about claimant B-reader bias. In general the employers seem to have more financial resources to garner medical reports, such as B-readings, than the claimants have.

    In contested claims the B-readers are not selected randomly from the largest pool of B-readers available, as is recommended by NIOSH in Classification of Chest Radiographs: Practices in Contested Proceedings.

    This B-reader bias, be it real or merely perceived, underscores the need for “blinding” in contested claims.

    Once digital x-rays are validated and approved for the process of determining the extent of coal workers’ pneumoconiosis, it may be possible to amend the evidentiary procedures to have the B-readers, who are selected by a party contesting a claim, read the contested digital submission in “line up” fashion with other non contested digital results. The practicality of this idea is largely dependant on the use, storage and secure electronic transmission of digital x-rays.

    This idea also assumes that a large pool of digital x-rays can be stored, and sorted into three groups: the first group negative (i.e. legally negative) for coal workers’ pneumoconiosis (perfusion up to 1/0), the second group positive for coal workers’ pneumoconiosis (perfusion 1/1 to 3/4 without a large opacity), and the third group of with large opacities over one centimeter.

    The B-reader would required to read the contested digital film in a “blind” line up with other digital films randomly selected from the three pool groups of stored films, perhaps as few as three (one from each category) or as many as six (two from each category). All of these films would be identified with unique codes (known only to NIOSH), but not identified by personally identifiable information such as the SSN. The B-reader would read the four or seven digital films, and send the results back to NIOSH. The B-readers interpretations of the non-contested films could be kept on file under the B-reader’s name to monitor the proficiency of the B-reader and the contested film interpretation could be forwarded to the parties to the claim.

    I realize that this procedure would require a regulatory change in the processing of coal workers’ pneumoconiosis claims, but if and when digital films are feasible, I believe it would be worth considering.

    While this suggested procedural change would increase the cost of x-ray readings for a contested claim, if the “line up” process could eliminate B-reader bias it may result in earlier settlement of claims, and therefore be a cost savings in the overall litigation process.

    We thank Mr. Spence for his thoughtful post.

    Although it is not our policy to comment on individual readers, Mr. Spence raises some important issues.

    First, our recent publications certainly do not support the view that pneumoconiosis is non-existent. One of our latest surveillance reports shows that although the prevalence of coal workers’ pneumoconiosis (CWP) dropped substantially from 1970 through 1999, there is evidence that the prevalence is rising. In addition, in our coal workers’ x-ray program we unfortunately continue to encounter relatively young (<50 years old) coal miners with severe occupational lung disease. We also note that the International Labour Office (ILO) instructions for the system for reading radiographs for the pneumoconioses explicitly state that only for clinical readings may the reader interpret the findings medically in terms of believing or suspecting their origin. Our own ethical guidelines state, "B Readers shall recognize the limitations of chest radiograph classifications, and shall not make clinical diagnoses about the pneumoconioses based on chest radiograph classification alone."

    As Mr. Spence notes, we have published guidelines for appropriate methodologies for different scenarios, including epidemiology, surveillance, worker monitoring, contested proceedings, government programs, and medical diagnosis. Critical components contributing to valid reading of chest x-rays for contested proceedings include random selection of readers, multiple readings, and blinding to the individual's status. Other important considerations are remuneration that is not contingent on the individual's medical condition and the application of quality assurance procedures. Mr. Spence offers some interesting details on how these conditions might be achieved, although, as he notes, in certain areas regulatory changes would be needed. As he also notes, the transition to digital images facilitates the achievement of valid, unbiased readings. Digital images provide the theoretically greater potential for easy circulation to members of a reader pool as well as the ability to anonymize images and to include quality assurance images in the reading process. NIOSH welcomes proposals from interested parties for developing approaches to achieving unbiased and reliable readings of the pneumoconioses in all situations.

    So, basically you are saying that the Government of the United States of America as manifested in NIOSH has determined that it is more protective of the public-health to detect pneumoconioses 15-40 years after the first exposure, which is the essential capability of the chest x-ray, and describe the penetration of disease, rather than employ current technology for early detection and a resultant prevention of disease.

    Anybody who has not been living in a cave on Mars for the last 20 years knows that the “B” reading system has been abused, although digitizing may ultimately allow electronic data analysis of the film instead of having the interpretation be subject to the whim or funding of the reader.

    Although I agree with levels of radiation being worrisome, perhaps NIOSH should consider funding some pilot studies to determine how many lives can be saved using early imaging techniques and thusly limiting and/or ceasing exposure to agents which cause pneumoconiosis and cancer.

    NIOSH is open to consideration of medical techniques that can lead to occupational disease prevention without imposing adverse risks or consequences on individuals. NIOSH provides support for investigator-initiated research. Should someone want to conduct research in this area, they may apply for funding by mechanisms available and described under “Funding Opportunities” at Proposals are funded annually based on availability of funds, relevance of the proposed work to the NIOSH mission, and the quality of the science proposed.

    I need to ask when or are you accepting digital images. We are going to be totally digital this Fall yet still maintain a processor for plain films. I have called many times and been told it is not in the near future yet it looks to me that it is. Can you please let me know what is actually happening. It is very hard to maintain these systems for plain films.

    We infer from your question that you are referring to the Coal Workers X-ray Surveillance Program (CWXSP), a federal program that NIOSH administers by which underground coal miners can receive periodic chest x-rays for the purpose of preventing disease exacerbation.

    You are certainly not alone in being under pressure to deal with this problem. The scientific and technical issues are of course very complex, especially with the medical-legal implications of pneumoconiosis, while the need to transition to one of the available digital radiographic technologies for medical surveillance is extremely acute. NIOSH shares your concern and is moving as rapidly as possible towards a digital format for the Coal Workers’ X-ray Surveillance Program (CWSXP), including both services for coal miners and training and testing for physicians under the B Reader Program. We have assembled an excellent team that is working hard to address needs of both the CWSXP and of physicians performing International Labour Organization (ILO) classifications for other purposes. The American College or Radiology and the ILO are critical partners in this effort. We expect to begin accepting digital images in 2010.

    One of the potential barriers to implementation of digital images for surveillance of the pneumoconioses has been the question of whether the use of digital images would lead to the same findings as does film. In this respect, an important milestone has been the recent publication of NIOSH-funded work indicating that, for the purposes of the CWSXP, digital chest imaging will provide ILO classification results equivalent to those of traditional film-based chest radiographs (Franzblau A, Kazerooni EA, Sen A, Goodsitt MM, Lee SY, Rosenman KD, Lockey JE, Meyer CA, Gillespie BW, Petsonk EL, Wang ML. Comparison of digital radiographs with film radiographs for the classification of pneumoconiosis. Acad Radiol. 2009 Jun;16(6):669-77. Epub 2009 Apr 2). This research was a necessary foundation for implementing the transition to digital format; these findings had to be established before we could move forward.

    Currently, NIOSH is making progress in transitioning the CWXSP to accept digital images in addition to films. This involves two activities. The first is to put into place the necessary hardware and infrastructure so that digital images can be received, processed, and stored at NIOSH. We will be selecting contractors for this in the next few weeks. After that, we expect the process to take about a year for the equipment to be installed and the infrastructure to be set up and tested. The second activity involves working with the International Labour Organization (ILO) to create digitized versions of the reference x-rays used in their pneumoconiosis classification system. Currently, the ILO mandates that classifications can only be made using the film-based references. Working with ILO, it is our goal to have digital-format ILO standard images available sometime in 2010.

    We have also contracted for development of software that physicians can use to view ILO standard images and examinee images on side by side diagnostic-quality monitors and record ILO classification results. We also aim to make this software, along with specifications for how it should be used to perform ILO classifications, available to the public sometime in 2010.

    For now, everyone should please attempt to hold onto the film screen systems to maintain the capacity for medical surveillance for dust exposed workers. We appreciate your continuing commitment to occupational health!

    During our hygiene survey a high concerntration of coal dust and silica were detected in the boiler house; are there any other tests that I can do or specific CXR that can be done to detect problems?

    While your question is important it does not directly deal with the topic of this blog—digital radiographs. We will have someone contact you directly to discuss your question.

    I am interesting in learning more about this technology and the plans that are in place to implement it into other fields. Was this project created mainly for use within the mining industry? What other industries do they plan on introducing it to?

    Thank you for your comment. Both the technology of digital radiology and the standardized methodology for evaluating the medical signs of certain dust-related diseases of the lung have applications beyond the mining industry. The standard methodology for assessing occupational lung disease is appropriate for the pneumoconiosis – fibrotic diseases of the lung typically caused by inhalation of asbestos, silica, and coal dust, and, more rarely, other types of dust. Such dusts are certainly encountered in mining, particularly coal dust, but many other workers are at risk (e.g., in construction and manufacturing). Although asbestos exposure has been severely limited in recent years, many workers involved in building demolition and remediation are still at risk.

    More information on the transition to digital radiology can be found in the NIOSH publication Application of the ILO International Classification of Radiographs of Pneumoconioses to Digital Chest Radiographic Images.

    I never knew that there was a safety act for miners. It’s nice to know that there is. I wonder what the results would look like since these people would encounter different challenges in the mine everyday. They might have inhaled different substances from the mine as well. Digital Chest monitors are really handy this way. But also, constant testing must be done.

Post a Comment

Your email address will not be published. Required fields are marked *

All comments posted become a part of the public domain, and users are responsible for their comments. This is a moderated site and your comments will be reviewed before they are posted. Read more about our comment policy »

Page last reviewed: April 27, 2018
Page last updated: April 27, 2018