By John Vartanian
With Philips Healthcare’s breakthrough whole-slide imaging technology cleared by the FDA this past April, the potential applications appear limitless.
One of the most promising developments that will soon impact the day to day lives of pathologists is the integration of artificial intelligence and machine learning algorithms into these powerful scanners. The algorithms will take the complete image of the slide and assess each pixel for worrisome abnormalities at a speed and accuracy impossible using current techniques. This is the promise of computer enhanced pathology, and it is coming not a moment too soon; projections show that the field of pathology is experiencing a staggering 25% decrease in the work force, and patient volumes continue to climb along with the aging population. Even with these irreconcilable trends, pathologists have been resistant to machine learning diagnostics due to reasonable fears of patient safety and job security.
The only effective approach to a reaching a solution is to partner with pathologists rather than replace them. This is the mantra of an exciting startup company coming out of Ann Arbor, Michigan, named MedKairos. Rather than aiming to outright replace the pathologist as other ventures have attempted, they have decided to help pathologists by automating the tedious process of biopsy quality control.
Without this quality control measure in place, one out of every five biopsy procedures are repeated, with the financial costs felt directly by the hospital. This has forced many large cancer centers to adopt the labor-intensive process of biopsy quality control at the point of care. Current practice requires a pathologist to travel with their microscope to the biopsy suite, where they will manually count the cells in every pass of the needle to verify adequate tissue was collected to make a diagnosis. This process adds another task to the already over-worked pathologist’s schedule, and wastes valuable time that could be spent making more diagnoses. MedKairos has an automated platform that was developed for pathologists, by pathologists, that streamlines this process and allows the pathologist to spend their time doing that for which they were trained: making diagnoses.
