Why intelligent EHRs are the next big step in radiology
Better integration of images and analyses of their implications can provide actionable insights that can drive care improvement.
In November 1895, Wilhelm Röntgen, a German mechanical engineer, professor and physicist, produced an X-ray image of his wife’s hand. It was the first time in history that anyone had detected electromagnetic radiation in a wavelength range.
His pioneering work earned him the first Nobel Prize for Physics in 1901 and opened up the world of medical imaging modalities, including CT and MRI machines. Over time, imaging has become clearer, radiation exposure decreased and our ability to collect CT and MRI data skyrocketed.
Although we have made incredible strides in imaging capabilities in the past 123 years, our capability to generate actionable information from the report has changed very little since Röntgen first conducted his experiment.
While radiology and image-driven workflows generate tremendous amounts of useful information, it is stored as unstructured data in the form of PDFs or paragraph text. On its own, this data doesn’t have much worth to radiologists and other physicians who rely on electronic health records (EHRs) to convey and disseminate information.
While many might see this as a failure within our industry, it’s an opportunity to place the radiologist at the center of the care team, transform the next generation of reporting and connect the mass amounts of imaging information with the physicians and caregivers who need it.
To understand how radiologists can improve their processes through health IT, we must start by defining diagnostics and explaining the role of the EHR for radiology and imaging experts. Put simply, the EHR serves two primary purposes for clinicians—guiding patients to the right care pathway and acting as an information system that both inputs and retrieves information.
Clinicians will be the first to tell you that they have no use for raw data without context. It is only after we have entered this data into the EHR and provided a meaningful workflow that we are able to transform data into higher levels of information that can help us guide our patients to the appropriate treatments.
Actionable information provides value to consumers and clinicians. The digitization of insights, delivery of practical information and automatic follow-up appointment scheduling are critical to improving outcomes and enhancing the healthcare experience.
In a perfect world, radiologists and imaging experts could quantify the human anatomy and digitize the thousands of reputable medical journal articles that are published each year. From there, we could integrate algorithms, such as the St. John Sepsis Surveillance Agent, into EHRs and develop intelligent decision support tools. We’re at a pivotal point in medicine: once we start to properly use the immense amount of data that we have at our disposal, care pathways can be triggered automatically, and there will be a dramatic decrease in human error.
Imagine that a physician is evaluating a patient who has suffered from lung cancer for three years, measuring a 40 mm nodule on the upper left lobe of the lung. With current processes, a physician would have to look at the printed PDFs of separate reports side by side, then highlight those changes year-over-year to detect the growth. Once our industry unleashes the coded value that is hidden within these reports, algorithms built into our EHRs can do the heavy lifting of the diagnoses. This will result in saved time and money for health care organizations and more efficient patient care.
There is a reason that medical error and lost communication are a leading cause of death. Digitizing radiology in a more meaningful way will solve many of the inconsistencies in healthcare.
The role of partnerships cannot be understated when looking at the day-to-day workflows of radiologists and imaging experts.
Siemens Healthineers, for instance, has more imaging AI patents than any other company, and voice recognition and reporting vendors, such as Nuance, are able to take raw data and transform it into actionable insights. Through strong partnerships, vendors can quantify the information received from imaging AI and place it into the EHR workflow, which will ultimately spawn the next steps in the care process. Partnerships like this have never been done before and are critical to the improvement of the field of radiology.
When we look at intelligence—whether it’s human-created or computer-generated—it’s all equally valuable. We’re experiencing a great deal of hype for the promise of artificial intelligence, but there has not been tangible evidence that the AI that radiologists have at their fingertips now can undoubtedly improve outcomes. This will change soon.
As an industry, we’re primarily using AI for the automation of arduous tasks. For example, if a patient is in a serious car accident, the clinician might spend the first 10 minutes post-imaging labeling the ribs and spine of the patient to depict which are fractured. But AI can do that for us, saving physicians a great deal of time on the critical front end when every second counts. What used to take minutes now takes seconds.
AI has also made a huge impact on our imaging capabilities, especially regarding the recognition of abnormalities. Researchers have proven that computers can do just as well, if not better, than humans at detecting cancerous moles. This technology is making an impact in mammography by double checking the work of radiologists. In the past couple years, use cases like these have increased, techniques and processes have improved, and algorithms are of a higher quality.
Take, for instance, a case study from Intermountain Healthcare in Salt Lake City. A team developed a care process model that summarizes and updates evaluation and treatment recommendations for patients experiencing pneumonia. If a radiologist identifies that a patient likely has pneumonia, the EHR then automatically spawns the next step of the care pathway, triggers a prescription for the appropriate antibiotics and calls for the necessary respiratory care protocols.
Use cases like this plainly illustrate the role of the EHR for radiologists. With digitized information identifying all the symptoms of pneumonia, physicians wouldn’t have to think about the appropriate next steps in a patient’s treatment. Instead, the automated workflow would provide seamless care guidance throughout the patient’s visit.
When a patient gets their blood tested, their physician receives a snapshot of their overall health. Someday radiologists will depend on full-body scans and AI to get an in-depth look at a patient’s health.
There are still challenges to overcome before we can reach this ideal future state. Radiology, as we know it, lives within a fee-for-service world. As we shift toward value-based care, we must make changes that will favor the quality of care over the quantity.
Through improved technology, collaborative partnerships and more efficient communication, radiology has a considerable opportunity to change primary care. This is a natural next step that radiologists have desired for years and one that innovators like Wilhelm Röntgen would be proud of.
This blog originally appeared on Cerner's web site.
His pioneering work earned him the first Nobel Prize for Physics in 1901 and opened up the world of medical imaging modalities, including CT and MRI machines. Over time, imaging has become clearer, radiation exposure decreased and our ability to collect CT and MRI data skyrocketed.
Although we have made incredible strides in imaging capabilities in the past 123 years, our capability to generate actionable information from the report has changed very little since Röntgen first conducted his experiment.
While radiology and image-driven workflows generate tremendous amounts of useful information, it is stored as unstructured data in the form of PDFs or paragraph text. On its own, this data doesn’t have much worth to radiologists and other physicians who rely on electronic health records (EHRs) to convey and disseminate information.
While many might see this as a failure within our industry, it’s an opportunity to place the radiologist at the center of the care team, transform the next generation of reporting and connect the mass amounts of imaging information with the physicians and caregivers who need it. To understand how radiologists can improve their processes through health IT, we must start by defining diagnostics and explaining the role of the EHR for radiology and imaging experts. Put simply, the EHR serves two primary purposes for clinicians—guiding patients to the right care pathway and acting as an information system that both inputs and retrieves information.
Clinicians will be the first to tell you that they have no use for raw data without context. It is only after we have entered this data into the EHR and provided a meaningful workflow that we are able to transform data into higher levels of information that can help us guide our patients to the appropriate treatments.
Actionable information provides value to consumers and clinicians. The digitization of insights, delivery of practical information and automatic follow-up appointment scheduling are critical to improving outcomes and enhancing the healthcare experience.
In a perfect world, radiologists and imaging experts could quantify the human anatomy and digitize the thousands of reputable medical journal articles that are published each year. From there, we could integrate algorithms, such as the St. John Sepsis Surveillance Agent, into EHRs and develop intelligent decision support tools. We’re at a pivotal point in medicine: once we start to properly use the immense amount of data that we have at our disposal, care pathways can be triggered automatically, and there will be a dramatic decrease in human error.
Imagine that a physician is evaluating a patient who has suffered from lung cancer for three years, measuring a 40 mm nodule on the upper left lobe of the lung. With current processes, a physician would have to look at the printed PDFs of separate reports side by side, then highlight those changes year-over-year to detect the growth. Once our industry unleashes the coded value that is hidden within these reports, algorithms built into our EHRs can do the heavy lifting of the diagnoses. This will result in saved time and money for health care organizations and more efficient patient care.
There is a reason that medical error and lost communication are a leading cause of death. Digitizing radiology in a more meaningful way will solve many of the inconsistencies in healthcare.
The role of partnerships cannot be understated when looking at the day-to-day workflows of radiologists and imaging experts.
Siemens Healthineers, for instance, has more imaging AI patents than any other company, and voice recognition and reporting vendors, such as Nuance, are able to take raw data and transform it into actionable insights. Through strong partnerships, vendors can quantify the information received from imaging AI and place it into the EHR workflow, which will ultimately spawn the next steps in the care process. Partnerships like this have never been done before and are critical to the improvement of the field of radiology.
When we look at intelligence—whether it’s human-created or computer-generated—it’s all equally valuable. We’re experiencing a great deal of hype for the promise of artificial intelligence, but there has not been tangible evidence that the AI that radiologists have at their fingertips now can undoubtedly improve outcomes. This will change soon.
As an industry, we’re primarily using AI for the automation of arduous tasks. For example, if a patient is in a serious car accident, the clinician might spend the first 10 minutes post-imaging labeling the ribs and spine of the patient to depict which are fractured. But AI can do that for us, saving physicians a great deal of time on the critical front end when every second counts. What used to take minutes now takes seconds.
AI has also made a huge impact on our imaging capabilities, especially regarding the recognition of abnormalities. Researchers have proven that computers can do just as well, if not better, than humans at detecting cancerous moles. This technology is making an impact in mammography by double checking the work of radiologists. In the past couple years, use cases like these have increased, techniques and processes have improved, and algorithms are of a higher quality.
Take, for instance, a case study from Intermountain Healthcare in Salt Lake City. A team developed a care process model that summarizes and updates evaluation and treatment recommendations for patients experiencing pneumonia. If a radiologist identifies that a patient likely has pneumonia, the EHR then automatically spawns the next step of the care pathway, triggers a prescription for the appropriate antibiotics and calls for the necessary respiratory care protocols.
Use cases like this plainly illustrate the role of the EHR for radiologists. With digitized information identifying all the symptoms of pneumonia, physicians wouldn’t have to think about the appropriate next steps in a patient’s treatment. Instead, the automated workflow would provide seamless care guidance throughout the patient’s visit.
When a patient gets their blood tested, their physician receives a snapshot of their overall health. Someday radiologists will depend on full-body scans and AI to get an in-depth look at a patient’s health.
There are still challenges to overcome before we can reach this ideal future state. Radiology, as we know it, lives within a fee-for-service world. As we shift toward value-based care, we must make changes that will favor the quality of care over the quantity.
Through improved technology, collaborative partnerships and more efficient communication, radiology has a considerable opportunity to change primary care. This is a natural next step that radiologists have desired for years and one that innovators like Wilhelm Röntgen would be proud of.
This blog originally appeared on Cerner's web site.
More for you
Loading data for hdm_tax_topic #better-outcomes...