07 Feb 2019

Artificial Intelligence all set to Virtualize Radiology Brains

Mr. Prashant Akhawat’s interview with Live Mint

  • AI and automation will not replace radiologists, but will act as an assistant to radiologist in every step of the imaging detection, diagnosis and prognosis
  • There is an increase in interest and enthusiasm for AI among the radiology community as the discussion is moving upwards on from considering AI as a threat

Accelerating with an exponential growth, artificial intelligence (AI) is all set to move from experimental stages to live industry implementations and all is set to mark its presence across all industry verticals. AI is all about virtualizing human cognitive functions in the form of software brains. For organizations, harnessing AI is not optional, albeit it is critical to stay competitive. Gartner in its recent study (2018), predicts the business value derived from AI to reach $3.9 trillion by 2022. With the disruptive potential, the investments in AI are ever-increasing. It is redefining industries with automation processes, and personalization. The healthcare industry has been one of the foremost adopters of the AI amongst all others. The advancement, the healthcare industry is reaping with the use of AI to maintain medical records, do mundane tasks more accurately and faster, to design care pathways, digital consultations, medication management, drug creation, image analytics, bigdata analytics, medical robotics, health monitoring and host of other things.

Powering Radiology with AI

Today, in India, there is a growing demand for improved medical care. We see an increase in insurance penetration, the rise in chronic disease and an aging population — all this demand for better imaging diagnoses and treatment. Radiology industry is facing its own set of ongoing challenges – the shortage of radiologists topping the list. The radiology industry is a functional domain for automation by virtualizing domain intelligence into an intelligent software. The quantum leap in medical imaging technology has led to exponential growth of medical imaging data stored digitally. Deep Learning algorithms and Image Analytics can help in improving medical diagnosis and aid radiologists with better reporting efficiencies.

Nowadays, with the growing incidence of lifestyle diseases, the requirement of frequent imaging and getting multiple scans at a higher resolution has increased. There are scenarios where the hospitals have X-Rays, MG, CT and MRI equipment’s but sans radiologist to read a report. To bridge the gap where AI can play a pivotal role by performing specific tasks such as image recognition – nodule detection, hemorrhagic or ischemic stroke detection, fracture detection, breast cancer analysis with prior case analysis and other narrow tasks requisite identify potential findings in medical images, which is one set of tasks performed by radiologists. This gives a window to the radiologists to focus more on image-guided medical interventions, defining clinical parameters of imaging examinations, relating findings from images with medical records and test reports, consult physicians for treatment based on the diagnosis, discussing procedures and results with patients. Artificial Intelligence is taking over image reading and interpretation so that radiologists can read more images in a short span with better accuracy as the number of images has increased more in the last decade than the number of radiologists.

Radiologists need to adapt themselves to new skill sets and technologies for attaining better productivity by integrating AI with radiology practice. AI and automation will not replace radiologists, but will act as an assistant to radiologist in every step of the imaging detection, diagnosis and prognosis and will also help in prior analysis and comparison. AI Algorithms will have a stronghold in medical imaging and will become an integral part of RIS-PACS, as frequently we hear some or the other algorithm is developed to detect tumors, lesions, fractures and host of other things.

With the help of RIS-PACS workflow enabled with AI, a radiologist sitting at any location can do the reads for any healthcare center located remotely, which addresses the accessibility challenges, subspecialty reads, point in time care with reduced costs. Further helping them to optimize results without compromising on the accuracy of diagnosis. For emergency reporting (trauma and stroke cases) AI can assist the radiologist with a preliminary report with which they can conclude with their final interpretation of the case within the stipulated time. AI Enabled RIS-PACS can prioritize stroke or stat cases and reduces turnaround time in turn will improve patient care. The algorithms are trained to assist the radiologists with various detections, diagnosis, staging, sub classification of different medical conditions. Hence, deep learning is a shoulder to the increasing workload in radiology.

Global Market for AI

Artificial intelligence’s complexity is not a deterrent to its adoption instead it is as disruptive as the internet was. The future of investment is AI, and it has not just been observed, but also proved that AI helps in improving the radiologist’s productivity, efficiency and quality diagnosis, hence this would help them reduce costs and increase return on investments. In the healthcare industry, AI is rapidly rising in the medical imaging domain. Globally, the AI market in medical imaging is forecasted topping US$ 2 billion by 2023. Notably, there is an increase in interest and enthusiasm for AI among the radiology community as the discussion is moving upwards on from considering AI as a threat. Also, clinical applications have shown improved clinical results with the use of AI.

All the research trends underline how AI is revolutionizing radiology in the long run. AI-based companies have learned to warm-up radiologists. Realizing the technological potential of AI, radiology practitioners are partnering with AI ventures to have a seamless RIS-PACS workflow. AI is going to augment the way care is provided by healthcare practitioners.

Source: https://www.livemint.com/ai/artificial-intelligence/artificial-intelligence-all-set-to-virtualize-radiology-brains-1549459969839.html

04 Feb 2019

Picture Archiving and Communication System (PACS) and Its Benefits

Technology and innovation play a major role in today’s healthcare system as it is crucial for sustaining health. It has enhanced the quality of medical care offered to patients, and PACS system in hospitals & health facilities is one such example of technology improving medical care.

PACS (Picture Archiving and Communication System) has changed the way radiology works and is now considered one of the most essential requirements in healthcare facilities. Sharing of instant medical images electronically and reporting them remotely is now very easy and quick, thanks to this invaluable software. With AI (Artificial Intelligence) enabled technologies now becoming available in PACS, their functionality is growing by the day

What is PACS?

PACS (Picture Archiving and Communication System) is a medical imaging technology that provides economical storage, presentation, retrieval, distribution, and management of medical images. Transmission of electronic images and reports takes place digitally via PACS. Thus, manual filing, retrieving and distribution of film jackets is no longer required. It allows storage and viewing of all types of medical imaging by healthcare organizations both internally and externally.

A radiology PACS is often deployed with RIS. An RIS is used to record patient history and schedule appointments, whereas PACS focuses on image storage and retrieval.

The four major components of PACS are:

  • The imaging modalities
  • Transmission of patient information through a secured network
  • Interpreting and reviewing of images through a workstation
  • Storage archives for retrieval of images and patient reports

Benefits of PACS

PACS improves efficiency in electronic data handling workflow. It offers a cost and space advantage due to decreasing price of digital storage. Benefits provided by PACS are many, but here we highlight few of the most important ones:

Improved Viewing and Analysis

An effective viewing and analysis is possible as PACS’ digital images enable you to zoom in and operate the images for a more elaborate analysis.

Where the conventional film can only exist in one place at one time, PACS enables simultaneous multi location viewing of images. It enables collaboration among radiologists as they can seek each other’s opinions by viewing the cases simultaneously and discussing them under peer review module. The interpretive skills of the professionals prove beneficial for the patients as well. The high-quality images make it possible to give a more accurate diagnosis.

Easy Accessibility to Images and Reports

PACS enables instant and easy access to images and reports. No matter where the tests are performed reporting can be done remotely and results can be shared anywhere, even if it is an isolated facility. PACS enables submitting reports, archiving images and transferring them through a portable media anywhere in the world. Practitioners at different physical locations can access the same information concurrently for teleradiology. In addition, quick access to prior images is also possible at the same institution. Radiology history of patients is available, which allows comparison with previous studies.

User-friendly Software

User-friendly since there are several customizations available for easy use of the software for staff and beneficial for patients. PACS is a great integration platform for other automation systems such as Radiology Information System (RIS), Electronic Medical Record (EMR), and Hospital Information System (HIS). The PACS database automatically groups all images chronologically, correctly labelled according to their examination. They can also be retrieved easily with the help of criteria’s such as name, hospital, referring clinician, etc.

Efficient Data Management

The system provides an efficient, seamless review of radiology cases within a physician’s daily workflow as it makes it easier to store and organize imaging data, with a centralized and accessible system. Data management becomes more efficient as the number of duplicate images can be reduced as previous data is available with the system.

Steps to Consider While Purchasing or Upgrading PACS

While purchasing or upgrading to a new PACS, one should look for features such as scalability and user-friendliness. The features should be easily configurable and should have uploading features for prior studies. Other than that, it should have a voice recording feature, integration with the Hospital Information System (HIS) and the Cloud feature. It is also beneficial switching to a new system when it has in-built AI algorithms.

An integrated RIS-PACS is a great advantage to radiology department as it can give evidence-based insights such as enhanced productivity of radiologist, turn-around time, modalities being used for improved workflow, and referring physicians sending maximum exams. RADSpa, one of the best integrated RIS-PACS available, provides a customizable work-flow suiting the requirements of radiologists.

A RIS-PACS’ rich intuitive AI integrated workflow, supporting all DICOM modalities, an orchestrated work-flow, multilingual software, advanced application supporting MIP/MPR and 2D/3D viewer, a scalable architecture and ability to integrate with any existing PACS and 3rd party AI algorithm, is sure to add value and increase productivity for any radiology department.

28 Dec 2018
Dr Anjali

Artificial Intelligence in Emergency Radiology

A Report, based on the Keynote Speech delivered
– By –
Dr. Anjali Agrawal, Head, Teleradiology Solutions, Delhi Operations
@ Artificial Intelligence in Radiology 2018 Symposium, November 10, 2018
Organized by Telerad Tech and Image Core Lab

Dr. Anjali Agrawal addressed the audience about how AI was going to be relevant in emergency care and emergency radiology.

Deep learning and the human brain

Speaking broadly on artificial intelligence (AI), Dr Anjali said that AI is a more gen­­­eral term and includes machine learning (ML) and deep learning (DL). Machine learning, a specific type of AI, gives computers the ability to learn without being explicitly programmed. Deep learning, a subset of machine learning, mimics the human brain configuration, where the multiple neuronal layers or neurons can crunch vast amounts of data and draw conclusions. In particular, DL has immense relevance for radiology and healthcare. The availability of large amounts of annotated image datasets and increased computational power had made AI a reality and it wasn’t an illusion anymore. It is moving from experimental to the implementation phase now.

Current state of Emergency and Trauma Care

Dr. Anjali drew attention to the current state of emergency and trauma care in India. From trauma registry, it was a documented fact that trauma related deaths in India occurred every 1.9 minutes. The mortality in serious injuries was 6 times worse in a developing country such as India, as compared to a developed country. A WHO survey revealed that there were more deaths due to lack of timely care than due to other diseases like AIDS, Malaria and TB, all put together.

Dr. Anjali said that, more than 80% of Indians didn’t get care within the golden hour and she highlighted the challenges in emergency care and radiology. There was tremendous pressure on the limited resources that were available at one’s disposal. The process from the scene of accident to the emergency room is disorganized. Dr. Anjali drew attention to the education system and training, which was quite heterogeneous.

AI will transform the ER services

As per Dr. Anjali, AI could be very useful for triage in the emergency room. Studies have shown that Emergency Severity Index assignment by doctors and nurses is correct only 60% of the time. They ended up under-triaging almost 27% of the patients. And a vast majority of those, nearly half, went into the mid-acuity group, level 3 – a typical human tendency to play safe.

AI and deep learning could help by analyzing the complex data from various sources – the age and sex of the patient, presenting history, complaints, vital signs, what was the mode of transport – did the patient walk-in or was s/he brought by an ambulance, past medical history etc. This could help in transforming the emergency department operations. By using these algorithms one could accurately triage patients, so that the critically ill patients got the attention of the emergency medicine physician and were managed appropriately. These algorithms could also help in allocating resources appropriately, minimizing mismatch between staff and patient case load, with improved patient outcomes. AI could help expedite interpretation of emergent imaging studies. These algorithms would be able to make predictions of adverse events and help make an individual-specific follow-up plan.

AI will revolutionize the radiology workflow and create smart enterprises

Dr. Anjali maintained that disruptive new generation healthcare technologies such as AI, robotics, machine learning and deep learning will revolutionize radiology workflow in many ways and that it was going to lead to massive improvements in quality, value and the depth of contribution of radiology towards patient care. She stated that one of the most well researched applications is emergency radiology.

Dr. Anjali cited that, AI could help make an informed decision regarding the need for imaging and the choice of modality based on analysis of the patient records. AI enabled algorithms could play a huge role in reducing radiation doses of CT examinations or reducing scan time for MRI, by using various enhancement and post-processing techniques. However, the overall decision making and communication with the referring physician or the patient would require the intervention of a human radiologist for quite some time, despite being aided by AI.

AI can help reduce scan timings and dosages

Dr. Anjali said that, the algorithms could enhance very noisy, grainy and undersampled data, such as from MRI, which were being produced in shortened timeframes and produce high-resolution MRI images– with huge implications in the emergency room, where one tends to shy away from doing an MRI because of time constraints. For e.g., if an MRI could be done in two- thirds or one-third the time required, one would be more comfortable in sending a sick patient for a suspected hip fracture for an MRI.

Similarly, these algorithms could also be applied to Computed Tomography (CT) scanners to help reduce the CT radiation dose – a huge advantage, as the reduction in radiation dose from CT would be comparable to a standard chest x-ray. One would be able to do ultra low-dose CTs, and get more information from diagnostic quality images, compared to a radiograph.

AI can ease the workflow of a radiologist in many ways

Showcasing a typical workflow of a radiologist, Dr. Anjali said that the radiologist logs into the system, reviews his/her work list and selects a study to review. The radiologist reviews other information such as patient history, prescriptions, etc., that are related to the case. Once the images are presented, the radiologist, in most cases, adjusts the hanging protocols to enable him/her to perform the interpretation and generate a report. The initial process of arranging studies is time consuming. Citing the recent developments in reading protocols, Dr. Anjali said that it could help hang the images in an interactive manner, learning each time, catering to individual preferences, and saving time.

AI can help in detection of findings, segmentation, quantification and reporting, in a manner that is easier to understand by both the referring physicians and the patients.

Dr. Anjali maintained that AI could help the radiologist by triaging cases, such that only the positive ones could be seen by the radiologists for further interpretation. She was of the thought that, as opposed to a few articles quoting that AI would replace a radiologist for particular targeted applications,  AI would assist a radiologist, where a radiologist would act as a second reader or vice versa.

She gave one particular example where an article looked into automated detection of critical findings on non-contrast CT examinations of the head –hemorrhage, mass effect, and hydrocephalus using an AI algorithm, which would be helpful in triage. If the algorithm found the non-contrast CT to be negative, it went through another stroke algorithm. In the case of it being positive, it was labeled as a critical imaging finding, and if they were both negative, it was labelled as “no critical imaging finding”. The algorithm had a good sensitivity of 62% and a specificity of 96%, comparable to a radiologist in the detection of acute ischemia. The sensitivity and specificity were higher for detection of hemorrhage, hydrocephalus and mass effect, matching the performance of the radiologists. Therefore, the study concluded that there was a huge potential for AI algorithm in terms of screening and detection of critical findings in the emergency setting.

Dr. Anjali said that she had seen similar data in her group and had detected intracranial hemorrhage with a very high sensitivity and specificity using a hybrid approach of convoluted neural networks and factorial image analysis.  The data and the results were comparable to the existing literature and that more data pertaining to quantification and localization of intracranial hemorrhage is underway.

Juxtaposing her earlier statement on triage being the low hanging fruit for AI applications in radiology, Dr. Anjali, mentioned that apart from acute neurologic conditions, these triage tools had been used in the detection of chest radiographic findings by classifying them into normal or abnormal with a high accuracy of almost 95%.

She quoted another example of AI application – wrist fractures. These algorithms were trained by senior orthopedic surgeons. When the emergency room physicians, not trained orthopedicians or radiologists, used them, their sensitivity improved from 81% to almost 92% and specificity from 88% to 94%; with a relative reduction in misinterpretation rate of almost 47%. Dr. Anjali simplified it further by saying that the algorithm was able to emulate the diagnostic acumen of the experts by providing the labels on where the fracture was and also put a heat map, assigning a confidence level to the detected fracture.

Adding on, Dr. Anjali said that AI applications would also be useful in detection of non-acute findings the emergency radiology setting. These may be overlooked because the focus is on the critical life-threatening illnesses. AI could help with measurements of bone density, detection of fatty liver, coronary calcifications, and the presence of emphysema, which may not be relevant in the acute setting, but would have implications in the future.

AI can make expertise widely available and scalable

Dr. Anjali then stated that according to her, AI would become a huge leveler in terms of the expertise and experience of radiologists. Good radiology consult would become easily accessible, affordable, as well as scalable. In the scenario of a mass casualty incident, the algorithms could be put to use to quickly distinguish between critical and non-critical cases. Handling massive imaging volumes would also become easier.

Dr. Anjali mentioned about one particular study on automated bone age estimation where the algorithms were extremely accurate as well as reproducible, with an interpretation time of fewer than 2 seconds. This is a huge achievement because every radiologist knows how tedious and time-consuming the task of bone age determination is.

Man with machine synergy

She went to add that many similarities had been drawn between the fields of medicine and aviation. The pilot, as well as the doctor, needs to be highly skilled, as they are responsible for human lives. Both the professions have benefitted tremendously from automation. There is no flight without a human pilot, and similarly, there would be no healthcare without human doctors, because the legal responsibility would always be with the doctors. Dr. Anjali urged the doctors to not forget that medicine was an art and that the physicians needed to practice it like an art to stay relevant. AI would not replace radiologists. According to Dr. Anjali, it would be the synergy between man and machine that will help the profession as well as benefit the patients.

26 Dec 2018
New Pacs

Replacing existing RIS-PACS? Or a first time RIS-PACS buyer?

Struggling with what to do about an outdated RIS-PACS? Or, planning for RIS-PACS for the first time? In either case, it is not an easy and simple decision to take. But prolonging the decision will not help either. There will always be a dilemma, for an existing RIS-PACS user, regarding whether to pay the upgradation cost and stick to the same old vendor who ask for big bucks for each single support or whether to replace existing PACS with a new system which also comes with value added features such as AI algorithms in-built into the RIS-PACS. Other factors which needs to be considered before switching to a new system include the data migration costs, training costs, opportunity costs, related hardware upgradations costs, regulatory environment., cost of upgradation versus going with a new system, all these questions have to be answered.  As per expert estimate, transition and migration to a new system may take from 30-90 days. The challenge is also therefore to identify a vendor which can fast track the transition to maximum 15 days while ensuring that the ongoing work is not affected.

Finding the appropriate RIS-PACS can be a big challenge. Few important questions that needs to be asked regarding a new RIS/PACS is that it must be affordable, faster, should have prior studies uploading feature, voice recording feature, ability to integrate with third part voice recognition system such as dragon application or PowerScribe, and ability to be integrated with the hospital information system amongst the few important pre-requisites. Growing number of RIS-PACS providers are placing emphasis on reporting on the cloud feature.

What is an Integrated Radiology Information System (RIS) PACS?

PACS, or picture archiving and communication system, is a medical imaging technology used for storing, retrieving, presenting and sharing images produced by various medical hardware modalities, such as X-ray, CT scan, MRI and ultrasound machines. While digital medical imaging has brought in enormous savings for the imaging centres in terms of archival, storing, retrieval and sharing. It is the Radiology Information System (RIS), which helps manage the radiology workflow and the business.

Older RIS/PACS consisted of disparate systems – one for archiving patient images and one for storing patient records. Often, it would be noticed that the patient data in the PACS database may not be same as the data entered in the RIS database. If there is a mismatch between patient’s name or other demographic details entered in the PACS and RIS databases, then the system will not be able to correctly access all relevant records.  Such discrepancies can cause unwanted inconvenience to patients and referring physicians while also expose the facilities to unwarranted risks and legal liabilities.

Functions of RIS

Some of the key functions of RIS includes order entry; patient scheduling, assigning studies; tracking number of exams; assistance in billing etc.  A combination of the two (PACS and RIS) is termed as an integrated RIS-PACS. An integrated RIS-PACS gives radiologists or the administrators access to evidence-based insights such as which modalities are being used the most, which referring physicians are sending maximum exams, radiologist productivity, turn-around time, busiest time, days or week of the month etc.

An imaging facility stands to make enormous gain in terms of patient and financial outcomes if they choose an integrated RIS-PACS. RADSpa is one of the best Integrated RIS-PACS available on the shelf and deployable in various situations and for different kinds of facilities.

Now, what is AI-enabled RIS-PACS?

AI -Enabled RIS PACS is a platform on which resides numerous AI algorithms developed by the RIS-PACS providers themselves or those provided by niche AI companies. For example, the RADSpa RIS-PACS platform is AI enabled. So, when you go with RADSpa, you also get access to a number of algorithms. What makes this system very interesting is the fact that there are no initial upfront investments on AI part. You pay for the algorithms, only when you use them. That, too, it is pay-per-use system.

To replace or retain?

There was a time when the average longevity of a RIS-PACS would be around 7-10 years. But now in the current fast developing diagnostic imaging space where not only the volume of imaging and its complexities are growing by leap and bound but also the regulatory and privacy requirements such as HIPAA and GDPR, which is forcing radiologists to be always hard pressed for time and for quality reporting.  Any delay or hesitation in decision making can be hazardous to imaging business.

But hesitation presents hazards

Imaging facilities need to appreciate that failing behind on modernization and working with an out-of-date RIS-PACS can have serious consequences for clinical efficiency and financial health of the centre.  In addition, sticking to system which has far outlived its lifespan can also make it tough for a facility to keep up with the expectations of referring physicians, affiliated organizations and patients.

Telerad Tech, the global health IT company and one of the leading providers of integrated RIS-PACS Workflow strongly recommends that while analyzing current RIS-PACS and its vendor, an imaging facility should do proper audit in the key areas of operations such as – administrative, clinical, information technology (IT), regulatory environment and the market.

Important points to consider

For instance, if your existing PACS or RIS-PACS chokes your competitiveness, is unable to give you important insights regarding productivity of manpower, modality machines, have a confusing user interface, or simply doesn’t give you the next generation workflow tools such as workflow orchestration that optimize your productivity, you should consider a new system. Also, you should take into account the regulatory environment such as HIPAA and GDPR and go with systems which is capable of anonymizing patient data and is able to give you patient security framework (PSF) gateway, if needed. Teleradiology companies who are either reporting or have an ambition to serve defense hospital establishments, should ideally look for RIS-PACS which comes with PSF Gateway feature. RADSpa, which comes with PSF feature, is deployed at multiple hospitals under the Navy establishment in Mexico.

Making the Switch

Finally, having decided to switchover to a new RIS-PACS, a facility should first clearly define its requirement. Talk to vendors who can help you assess your current and future growth requirements.

First Time RIS-PACS buyers

For a facility which is considering acquiring PACS for the first time, they should look for a solution which offers latest productivity tools, can integrate with the existing DICOM compliant modalities amongst other features which are explained below.

Listing the requirements

Begin your process for procuring RIS-PACS, like for any other product, by listing the requirements. If possible, involve the Radiologists, Technicians, IT Team, Operations Team, and Finance Team in the process. Your list of requirements can include:

  • RIS-PACS should easily and fast integrate with existing modality machines;
  • System should be able to integrate with existing Hospital Information System (HIS) through Health Level 7 (HL7) protocol;
  • Should be able to reduce the turn around time (TAT) and increase productivity;
  • Should be able to integrate with existing PACS, without requiring any major overhaul of the system;
  • For teleradiology companies which have complex workflow and QA requirements, the facility should look for systems which has multi-read workflow management features and whose QA and peer review module facilitates collaborations as per ACR guidelines;
  • The new system should have smart features like workflow orchestration, real-time work lists, CD burning feature, multi-monitor support;
  • The RIS-PACS should meet regulatory requirements of FDA and should be CE certified, HIPAA and GDPR compliant;
  • Should have advanced 3D DICOM Viewer Features such as Minimum Intensity Projection (MIP), Maximum Intensity Projection (MIP), Multi Planar Reconstruction (MPR), and sculpting tools;
  • Radiologists/facilities often want to customize layout in the viewport using custom feature which helps standardize the workflows as per their specific requirements. So, look for such features in the system proposed by your vendor;
  • Facilities should also look for RIS-PACS which offers hanging protocol features for each specific modality machine;
  • Vendor Neutral Archive (VNAs) technology is today a game changer. VNA is enabling imaging facilities to archive and retrieve millions of medical images generated by disparate modalities from many different vendors. So, look out for RIS-PACS which is VNA compliant;
  • If you are a new and starting small, look for solutions which will be able to grow with you, i.e., look for a solution which is scalable;
  • Depending upon your specific choice, you may go with a solution which is pure cloud so that you can jump start radiology without any major investments in IT infrastructure;
  • You may go with an on-premise solution, if your already have the IT infra and manpower in place or have the capital to invest in on-premise solutions;
  • Products like RADSpa also offers something called Hybrid solution that stores images on site in the local system, while RIS is available on cloud which gives the flexibility to report from anywhere. This type of systems can potentially reduce your investments and recurring bandwidth expenditure by up to 30%.


Regardless of the fact that whether you are switching to a new system or a first-time buyer of RIS-PACS, it is the timing which is most important. The transition should be such that the normal operations are not affected. You can also go for a trial run for about a month so that your team is well familiarized with the software. RADSpa offers free trial to most of its prospects after properly assessing the seriousness of the customer.  So, go on, and go for your new system to leapfrog your facility to an integrated and AI-Enabled RIS-PACS environment.

14 Dec 2018
Dr Arjun Kalyanpur

Artificial Intelligence in Teleradiology

A Report, based on the Keynote Speech delivered
– By –

Dr. Arjun Kalyanpur, Chief Radiologist and CEO, Teleradiology Solutions
at Artificial Intelligence in Radiology 2018 Symposium, November 10, 2018
Organized by Telerad Tech and Image Core Lab

Dr Arjun Kalyanpur began his speech recalling how Teleradiology Solutions for the past few years has been engaged in the field of artificial intelligence (AI) with the health-IT company Telerad Tech and drug trial company Image Core Lab. He also¬ congratulated Telerad Tech for launching MammoAssist – its AI software for breast cancer detection and also for establishing a new dedicated AI lab facility at the same time.

Dr. Kalyanpur noted that his company Teleradiology Solutions had reached an inflection point in its journey in the direction of artificial intelligence and it was an opportune moment for their company to be part of an event like this and be amongst the distinguished speakers in the field of AI.

AI and Teleradiology – synergistic and disruptive technologies

As per Dr. Kalyanpur, both AI and Teleradiology are very closely linked to each other and are extremely synergistic. They both address the same fundamental clinical issues about shortages of radiologists and both provide technology-enabled solutions for the same clinical issues.

Revisiting their journey of how teleradiology brought them to the current point where AI is seen as the direction forward, Dr. Kalyanpur said that the drivers that set them on this path 15 /16 years ago was the shortages of radiologists with most of them being focused in metros, and emergency radiology being the most severely affected area and still continues being one today.

He noted, that the numbers of radiologists haven’t changed much in the past 20 years and that there was still a dramatic shortage of radiologists the world over. Drawing up a scenario, Dr. Kalyanpur highlighted that even a developed country like the US with the best radiologist to population ratio, considered itself of being in the deficit with India being in the middle of the curve somewhere, whereas Tanzania on the other end had 1 radiologist for over 1 million population.

AI is important for addressing the increasing burden of diseases

He cautioned that, there is an increase in chronic diseases, due to progressively unhealthy lifestyles and living conditions, all of which required repeated imaging at multiple times in a patient’s life, leading to an increase in the number of images and their complexity in studying them. He added that in the 123 years, between the first X-Ray being produced in Roentgen’s Lab, till today, advancements have led to a dizzying array of imaging modalities of extremely high resolution, in all shades and hues being available to the radiologists, increasing work for them analyzing the data.

Understanding the gaps, Dr. Kalyanpur noted that, 15 years ago teleradiology evolved as a solution to these factors delivering the value proposition of not needing to have a radiologist at every hospital. It was technology driven solution with work flow and efficiency being at the core. It also had a quality benefit because the images were being brought to the most qualified radiologist, spawning the term “nighthawk”. The company was serendipitously at the forefront in the sector, providing night shift emergency services to hospitals at different parts of the world resulting in better performance and better quality services.

Emphasizing again, Dr. Kalyanpur said that, both teleradiology and AI were essentially disruptive technologies. Having been through and seen the evolution of teleradiology, he felt that, AI was seeing the same disruption as they had witnessed in teleradiology. Citing similarities between the two, he noted, that the resistance pattern was similar, and AI also faced the same concerns and fears that teleradiology had previously faced. Depending on how one would see them, both could either be seen as a threat or as a benefit. The same drivers were present for both innovations. Since both are IT enabled solutions and follow Moore’s law principles, one could only expect them to be faster, more effective and cheaper in the coming years. He also noted that both had the potential to transform healthcare by simultaneously increasing radiologist quality and productivity – a double whammy in terms of their benefit and impact on healthcare.

Key trends in radiology

Describing radiology as a quantitative science, Dr. Kalyanpur said that radiology today was beyond a simple image description. Studying a few trends in radiology prevailing today, Dr. Kalyanpur chose to highlight 3 of them, sub-specialization being one. Further, he noted that there was an increase in focus on report turn-around time as all of medicine was becoming metrics- driven. Turnaround time was becoming the basis of judging the performance of a radiologist or a radiology group. He also pointed out that the evolution of cloud-based workflows also had made a tremendous impact in the field.

Citing stroke as an example, Dr. Arjun said that, in the 15 years that they had been practicing teleradiology the specialty had evolved to the point where it was no longer just a matter of looking at the CAT scan of the brain and determining whether it was a stroke or not. Today they had a CT angiogram or a perfusion CT to analyze as well, increasing the number of imaging modalities available to assist in clinical decision making. Quantitative imaging today helped in quantifying the size of the stroke, determine the collateralization of the brain and the impact on patient outcome. Therefore, radiology was no longer just a diagnostic science but had evolved into being a prognostic tool while Quantitative imaging was available in every sphere – Whether it was tumor detection, tumor analysis or quantification of vascular disease.

Looking at it from the operational standpoint, Dr. Kalyanpur noted that turnaround time was another factor that entities needed to be conscious of. He mentioned that, every tele-radiologist in the organisation practiced medicine under the scrutiny of the stopwatch. Citing it as a challenge, he said that, finding a fine balance between quality and speed was important as the faster a radiologist reads an image, the greater was the chance of him/her missing a finding. As per him, teleradiology had made the important aspect of work flow possible by allowing decentralized reporting and being a cloud-based solution, a radiologist could read for hospitals in any part of the world, with the understanding that compliance with licensure and certification was adhered too.

Making a radiologist more efficient and productive

Taking a step back and speaking on his role as a radiologist, Dr. Kalyanpur said that, usually a radiologist detected and quantified the lesion, analyzed the finding, communicated and recommended the follow up. Tackling the criteria on making the radiologist more efficient and productive, he suggested making the process of reporting, the work flow smooth by providing all the relevant information to the radiologist and assisting them by creating a templated report. Speaking on the productivity, Dr. Kalyanpur said that, the average productivity of a radiologist in the world was between 6-10 RVUs (relative value unit) an hour. That could probably be increased by about 15-20% by using these techniques. However, AI could potentially double the productivity of a radiologist, thereby helping in solve the radiologist shortage issue of half the world.

Highlighting the larger issue of radiologist errors, Dr. Kalyanpur mentioned that as per some studies, 3-5% of radiologist reports contain errors, such as for example small bleeds in the brain, small clots in the lungs or very subtle fractures in the spine. Considering that a billion scans were performed around the world every year, it meant that around 30 million scans would have reporting errors – a gravely concerning huge number. He opined that radiologist accuracy and patient outcome could be improved by allowing AI to detect these findings.

Predicting on the estimates on the size of the industry, Dr. Kalyanpur said that, the initial figures in teleradiology suggested it being a billion dollar industry, however it had grown beyond 4 billion currently and he was confident that it would further grow. Similarly, he was optimistic about AI growing beyond the current 2 billion in the future and wished best of luck to Telerad Tech for its endeavours.

Developing AI in radiology

Speaking on the importance of developing AI in radiology, Dr. Kalyanpur said that, it was important to have communication between radiologists, scientists and engineers. This would enable data with labelled information help in developing the tools and would require radiologists to validate them once they were developed to confirm that they actually worked. He suggested that this could be integrated into the workflow most effectively by integrating it into the teleradiology workflow which allowed these technologies to be deployed at scale across the globe, where the entire world population of radiologists could be the consumer.

The synergies between teleradiology and AI was that at every stage of the AI cycle teleradiology had a valuable role to play and offer – whether it was data acquisition from around the world (as many as over 200 hospitals in 20 countries) forming a huge repository for analysis, testing an algorithm in a teleradiology environment and the teleradiology workflow to be used as a distribution tool for AI software.

Simplifying it, Dr. Kalyanpur noted that, teleradiology could be seen as being a shopping mall, in which each AI algorithm was an individual store or teleradiology could be an I-phone on which each individual app resided and allowed the consumer to connect with the product. He expressed his satisfaction to note that Telerad Tech, apart from developing its own software algorithm, was also partnering with other entities in the same space. And allowed their workflow to be used as a distribution module for teleradiology algorithms or for AI algorithms to radiologists around the world allowing all possibilities of detection, quantification triage for smaller lesions and detection for the larger lesions.

In his concluding remarks, Dr. Kalyanpur said that, it was all about perspective and focus. Quoting radicals such as Professor Hinton, who opined that since AI would have taken over in the future, training of radiologists should be stopped because we wouldn’t need them anymore on one hand to moderates such as Professor Langlot, Stanford, on the other who believed that radiologists who used AI will replace radiologists who wouldn’t in the future. He suggested that it would be wiser to have a pragmatic approach since it was an evolutionary area. He was of the view that one needs to learn to adapt and understand the space so that one could utilize it better. Citing Darwin he wrapped up his speech by saying that only the most responsive to change would survive over the more stronger or the more intelligent.