Dr Anil Handoo, Senior Director, Hospital Laboratory Services and Senior Consultant-Hematology, BLK-Max Super Speciality Hospital in an interaction with Express Healthcare talks about the role of technology in Hematology
You have almost two decades of experience in the field of Hematology. According to you, what are the major factors that have changed the face of Hematology today?
Technological advancements in the cell counter/Hematology analyser science, along with automation of majority of the critical steps in analysis, have been the major changes in recent years, which have led to facilitation of cellular analysis, and improved the way we work in lab. Right from improvements in technology, with better flag optimisation, todays’ hematology analyzers are able to not only differentiate mature cells, but also have the ability to accurately pick up the presence of abnormal cells. With addition of digital morphology platform and Artificial Intelligence (AI) with Machine Learning (ML) algorithms, digitised morphology has further enhanced the ability of picking of incidental findings on the smears, which sometimes were missed by manual microscopy for the sheer lack of time and focus on every slide.
If you have to explain the concept of digital morphology to a layman, how will you do that?
Digital Morphology is nothing, but visualisation of peripheral blood smear microscopy using camera, and automating the image capture with projection on to a computer, instead of manually seeing a slide on the microscope. Digitisation and AI-based pre-classification of cells on the smear ensure less amount of time spent on every smear with enhanced workflow efficiency. There is also an advantage of having stored digital image data not only for reviews at a later date, but also a possibility of remote viewing and opinion seeking.
How has digital morphology revolutionised the field of Hematology?
Digital morphology, with AI-based algorithms, has ensured highly reproducible cellular classification, quantification of abnormalities, not only for the white blood cells, but also for red cells and platelets. Possibility of scanning the smear on edges for presence of platelet clumps, or any other abnormalities enhance pick up rates of such abnormalities. Additionally, review and valued opinion of an expert from a remote location, with real-time incorporation into the report has been made possible.
What is the role of AI in digital morphology? Can you explain with one/two examples?
AI is adding new dimensions in morphology analysis, and predictive capability of the analysers has been increased many notches up. For example, let’s say if there is an abnormal cell or parasite which is present, but in very small numbers, human eye has a chance to miss, but since digital morphology analyser is going to look into lot more cells than human microscopy, chances of it getting picked up are higher.
Will AI replace the role of humans in pathology laboratories?
I do not see that happening in near future. However, AI-assisted cell analysis helps humans in obtaining more precision and accuracy.
To what extent can we rely on the results delivered by AI?
As long as instrument is well-maintained and smear is appropriately stained, I would trust AI to do a reasonably good job most of the times.
How can digital morphology be made affordable so that it can reach the masses? What are your suggestions?
Increasing the awareness regarding use of digital morphology would ensure increased demand. In parallel, miniaturisation of devices with smaller footprint and reduction of production costs would help make affordable to masses.
According to you, what are the aspects in digital morphology that need to change or evolve?
High-speed scanning with upgradation of AI and constant machine learning for pre-classification of cells would help improve the output. Also, remote web-based access with cloud-based data storage is the way to go to help us utilise the system better.
Can you please share your experience of using Mindray’s Hematology System in your laboratory?
Adoption of newer methodologies gets you across newer challenges; Digital Morphology Platform (DMP) integration into our routine workflow was no exception. Our initial experience has been that of learning and change management. However, once we settled, the experience has been exhilarating. We have had some very interesting and beautiful pick-ups by the DMP, and many beautiful images of cells with characteristic morphologies. While we bask in the glory of digital morphology, I still feel there are some areas of improvement and I trust Mindray Team is constantly working to improve the system.