As if the field of pathology were not already thought of as being impersonal, a Stanford team has developed a Computational Pathologist (C-Path) capable of making more accurate diagnoses than its human counterparts. The traditional method for conclusively determining whether a patient has cancer is to perform a biopsy and analyze the tissue under a microscope. Based on factors such as cell size and shape, or number of cells undergoing division, pathologists could diagnose whether the tissue was malignant. Despite significant medical advances, this particular procedure has not changed much since the 1920s.
Working off of this, the team used images of tissue samples from breast cancer patients to train a computer to more accurately differentiate between cancerous and non-cancerous tissue.
According to the press release:
To train C-Path, the researchers used existing tissue samples taken from patients whose prognosis was known. The computers pored over images, measuring various tumor structures and trying to use those structures to predict patient survival. By comparing results against the known data, the computers adapted their models to better predict survival and gradually figured out what features of the cancers matter most and which matter less in predicting survival…
Medical science has long used three specific features for evaluating breast cancer cells — what percentage of the tumor is comprised of tube-like cells, the diversity of the nuclei in the outermost (epithelial) cells of the tumor and the frequency with which those cells divide (a process known as mitosis). These three factors are judged by sight with a microscope and scored qualitatively to stratify breast cancer patients into three groups that predict survival rates…
C-Path, in fact, assesses 6,642 cellular factors. Once trained using one group of patients, C-Path was asked to evaluate tissues of cancer patients it had not checked before and the result was compared against known data. Ultimately, C-Path yielded results that were a statistically significant improvement over human-based evaluation.
What’s more, the computers identified structural features in cancers that matter as much or more than those that pathologists have focused on traditionally. In fact, they discovered that the characteristics of the cancer cells and the surrounding cells, known as the stroma, were both important in predicting patient survival.
We are excited by this trend towards computer-assisted clinical decision making. Systems like C-Path and IBM’s Watson have the potential to parse, analyze, and learn from expansive data sets in order to make the most informed diagnoses and treatment plans. One obvious downside? CSI could become a whole lot less interesting if the attractive pathologists are replaced by a computer.
Press release: Stanford team trains computer to evaluate breast cancer
Science Translational Medicine paper: Systematic Analysis of Breast Cancer Morphology Uncovers Stromal Features Associated with Survival