Technology Quarterly: Q4 2011
Indolent or aggressive?
Medicine: A computerized pathologist that can outperform its human counterparts could transform the field of cancer diagnosis
Dec 3rd 2011 | from the print edition
LOOKING for needles in haystacks is boring. But computers do not get bored. Contracting out to machines the tedious business of assessing the dangerousness of cancer cells in histological microscope slides ought thus to be an obvious thing to do. Cervical-cancer smear tests aside, however, such electronic intrusions into the pathology laboratory are limited. Grading cancer cells into “indolent” and “aggressive”, and hazarding an opinion about whether they spell a treatable condition or an untreatable one, has remained the realm of the human expert.
But not for much longer, if Daphne Koller, a computer scientist at Stanford University, and her colleagues have their way. They recently reported in Science Translational Medicine that they have written a program which can distinguish between grades of breast-cancer cell—and in a way that provides a more accurate prognosis than a human pathologist can.
- More than just digital quilting
- A classic invention
- Return of the human computers
- »Indolent or aggressive?
- Spotting the rot
- Sticky fingers
- And the winners were…
- The devil in the details
- Cities on the ocean
- Getting past the guards
- Resistance is futile
- Reinventing the wheel
- Seer of the mirror world
Previous attempts to build a computerized pathologist of this sort required the designers to specify precisely which characteristics of the samples being examined were most important. For example, they would tell the computer to measure the three traits human pathologists use to determine a tumour’s grade: the percentage of its cells that are tubelike; the diversity of appearance of the cell nuclei; and the proportion of cancer cells undergoing division. However, people are excellent at pattern recognition and skilled pathologists rely not just on these relatively easy-to-describe traits, but also on less well-defined characteristics that years of experience have taught them are significant too. Restricting computerised pathologists to the well-characterized bits of the process therefore inevitably results in worse performance than that delivered by their human counterparts.
Dr Koller’s Computational Pathologist (C-Path), by contrast, lets the system work out for itself what the most important features of a tumor are. She and her colleagues started by setting down 6,642 characteristics the program might choose from when it assessed images of biopsies from breast-cancer patients, but did not tell it which to prefer. Some of the characteristics they offered were inherent to the cancer cells. Others were features of the surrounding “stromal” cells, which are not, themselves malignant, but act to support a tumor. And some were not features of individual cells at all but, rather, measured relations between cells (for example, the average distance between cancer-cell nuclei) and the context cells found themselves in (for example, whether they occurred in large clusters or were frequently interspersed with stroma).
The program was initially trained and tested using 248 breast-cancer samples from the Netherlands Cancer Institute. It was fed with images of slides from these patients, together with information on how long each patient had survived after the sample in question had been taken. That done, the software was then tested using a second set of samples, this time from 286 breast-cancer patients at Vancouver General Hospital. The system was able both to grade the slides and to predict, in a way that human pathologists could not, whether each patient would survive for five years after treatment.
When Dr Koller looked at which 11 features were the most robust predictors of survival, she discovered that only eight were characteristic of the tumour cells themselves. The other three were stromal characteristics. The fact that three stromal features were on the list suggests that the surrounding stroma influences whether or not a cancer progresses and kills the patient. That is important information because, hitherto, pathologists have focused on the cancer cells themselves and ignored the stroma. As well as outperforming human pathologists, it seems that C-Path can also teach them a thing or two about cancer biology.
from the print edition | Technology Quarterly