Since inception, advances in Medicine and Technology have always been interlinked. Technology has played a major role in making medical science more accurate and precise, more evolved and powerfully equipped to both diagnose and treat diseases effectively. From new age imaging solutions to highly competent pathology scanners; from surgical robots to extremely precise, dose adjustable radiation machines; technology has brought to reality, almost everything that the medical fraternity could conceive and ideate.
As a pathologist, and someone extremely fascinated with the application of newer technology to improve processes in healthcare, I am highly intrigued by the exponential upsurge seen in the Digital Pathology (DP) and Artificial Intelligence (AI) space globally.
What is digital pathology? Digital pathology is the creation of a dynamic image-based environment that enables the acquisition, management and interpretation of pathology information from a digitised glass slide. For the last 100 years, pathology interpretation was done by pathologists using a microscope to view glass slides that have processed tissue sections taken from a biopsy. Current technologies allow digitization of the glass slide into a high-resolution image that can be viewed on a computer-essentially converting the computer into a microscope. The digitisation of the glass slides not only helps the pathologist make a diagnosis on the computer screen, but the thousands of pixelated data points become a treasure trove of information which researchers can apply artificial intelligence techniques on to retrieve information not normally seen by the human eye. AI techniques have already started to help make diagnosis better and faster.
They are also playing a seminal role in the prognostication of the disease and the prediction of treatment modalities that would work best for the patient. Pharma companies find the digitised slides useful to build models for companion diagnostic drugs, for clinical trials and for predicting underlying molecular/genetic alterations.
From single slide scanners that scan a slide in minutes to multislide scanners which scan from few slides to many hundreds of slides at a rapid pace, the new age scanners cater to a wide range of requirements of a pathologist and a laboratory. Some of these platforms can easily be carried in a customised briefcase while some need reasonable space and special care of a laboratory setting. Most scanners of commercial use offer crystal clear image quality with highest magnification ranging from 40X and 80X to even 100X for the Haematology smears.
With so much variety and mammoth demand, it’s not difficult to be confused and lose sight of the need that one wants to address with the scanner from its utility in a particular diagnostic setting.
The result is that quite often we encounter scenarios where people have invested in a platform or bought multiple platforms of a similar type but haven’t really been able to decipher how to put it to good use. Also, many seem to be sceptical and insecure about the potential of these systems to make pathologists redundant. Through this article myself and Dr Rajendra Singh, one of the global pioneers of the application of DP and AI solutions, aim to address the utility of these technologies in current times and throw light on what the future holds.
Before beginning to adopt these modalities, it is imperative to understand the problem area and need of an individual laboratory which can be addressed through digitization. A reference laboratory set up which caters to hundreds of cytology, biopsy and immunohistochemistry slides in a day, needs more robust multi-slide scanners for various purposes which range from digitization of cases for archival, redistribution of work load to network laboratories with lesser load, to address the needs of high image resolutions for annotations helpful in developing AI algorithms and then validate and use these algorithms to report cases. These systems are expensive, need special handling for installation and usage.
On the other hand, a smaller laboratory catering to a limited number of histopathology and cytology cases with usually a single reporting pathologist, would need simpler and user-friendly equipment with fewer demands of budget, space, care during installation and usage and IT. The requirement which needs to be addressed in these laboratories is mainly of seeking second opinions from the referral centres or scanning of cases in specific situations when the reporting cannot happen in house, like if the pathologist is on leave (and a locum replacement is not available or not desired) or there is a system breakdown in one of the units essential for slide preparation like the tissue processor. These simpler, unostentatious, more mobile systems are also very handy for pathologists working part-time at various centres in case they require expert consult for a difficult case of a particular organ system.
A pivotal role for the application of DP in any setting is played by the Information Technology (IT) backbone of a centre. An adequate upload and download speed and a secure, stable connectivity is mandatory for these systems to function to their optimal capacity. IT also plays a major role in creation of a strong digital pathology network (DPN) which has all check points of security and compliances ticked. Another major role of IT is the integration of systems with the Laboratory Information System (LIS) for seamless and structured reporting as per the laboratory format and standards, capturing the demographics and other patient details like clinical findings and specimen type.
DP has given the power to many pathologists to create teams addressing demands of speciality reporting of various systems from anywhere in the world. This has also enabled pathologists working in limited resource settings across the globe to access these services for difficult cases and provide accurate diagnosis despite the limitations.
With the development of various AI models catering to reporting in fields like cytology and systemic pathology, inter-observer variability and subjective nature of reporting is largely mitigated. Another advantage of some AI tools is that they highlight potential areas/regions of interest (AOI/ROI) in a slide which can sometimes be missed by a pathologist reporting large volumes especially at the end of a heavy working day.
The recent announcement of inclusion of new Current Procedural Terminology (CPT) test codes by College of American Pathologists (CAP), has brought enormous excitement and also realisation of the advent of the ‘Digital Age of Pathology’. We visualise more and more CPT codes being included in the future rationalising the use of digital reporting.
The potential as we see is thus, limitless. The future most definitely belongs to a global adoption of these technologies for various purposes particularly systemic reporting with AI enabled tools, especially for haematology and organs like Breast, Colon and Prostate. This will also see more pathologists being able to work from settings different from the traditional laboratories and even at times when it is not feasible to go to work due to personal commitments, ill health or circumstances like the ones we saw during Covid-19 pandemic and with the Ukrainian crisis. DP definitely emerges as the most defining technology which circumvents the need to travel to work to report cases, ensuring the patients get their diagnosis for instituting proper and timely treatment despite grave and life-threatening circumstances affecting millions of people across nations.
From the perspective of Indian subcontinent, special areas where DP emerges as a game changer, besides the already enumerated aspects, are development of cytology screening tools for specific cancers like those of cervix which show high incidence in these geographies. It also assuages the cardinal crisis caused due to shortage of trained anatomic pathologists in the Tier 2 & Tier 3 cities.
In the end, we would like to conclude by addressing the most dreaded scepticism. Whether these technologies make pathologists redundant. A simple way of answering that would be, autopilot as a technology doesn’t make pilots oblivious, it just increases the efficiency of the flying process and will always need human supervision and intervention. So, technology can only be an aid for amelioration of processes and quality, bringing in more objectivity through automation. It can never replace the nuances that human brain can pick and never have that ‘gut feeling’ which many a times hits the nail on its head.
Dr Kunal Sharma is Associate Director & Head- Histopathology & Centre of Excellence, Lead- DP & AI initiatives, Mumbai Reference Laboratory, SRL Diagnostics, India and Dr Rajendra Singh is Professor, Director of Dermatopathology, and Associate Chair for Digital Pathology, Northwell Health, USA. Views are personal.
Read all the Latest News, Trending News, Cricket News, Bollywood News,
India News and Entertainment News here. Follow us on Facebook, Twitter and Instagram.
