Using imaging software could significantly speed up breast cancer diagnosis. Without the need for tissue preparation by specialists, it cuts down on time as it maintains a 90 percent accuracy rate, according to new research published Thursday in the journal Breast Cancer Research.
The study, led by Jessica Dobbs and Rebecca Richards-Kortum of Rice University, used images of freshly-excised breast tissue specimens from 34 patients and analyzed the tissue using high-speed optical microscopy. They then used a decision tree model that included a total of 33 different parameters to classify if the tissue was benign or malignant.
Benign features were classified in tissue specimens belonging to 30 of the patients, while 22 of them were found to have malignant features, the authors said. Overall, their model achieved 81 percent sensitivity and 93 percent specificity, corresponding to an overall accuracy rating of 90 percent. Moreover, this new method did not require lengthy tissue preparation.
In a statement, Richards-Kortum explained that being able to evaluate fresh breast tissue at the point of care could drastically change the current practice of pathology, adding that she had developed a faster way to classify breast tissue as benign or malignant, hoping to improve management of the disease in countries lacking regular access to pathologists.
Currently, the breast cancer diagnosis process requires tissue to be obtained, prepared, and then assessed, and pathologists need to undergo a complex method to prepare these samples. After the preparation is complete, there is a lengthy diagnosis process that follows. The use of microscopy, however, eliminates this need for complex sample preparation and assessment.
Method promising, but not yet ready for clinical use
“We performed our analysis without tissue fixation, cutting and staining, and achieved comparable classification with current methods,” said Richards-Kortum. “This cuts out the tissue preparation process and allows for rapid diagnosis. It is also reliant on measurable criteria, which could reduce subjectivity in the evaluation of breast histology.”
However, before this method can be used clinically, the team acknowledges that there are some obstacles that need to be overcome. For instance, some criteria rely upon the user’s observations, which could potentially result in incorrect classification of breast tissue. In addition, the cost and high-maintenance requirements of optical microscopy limit its use in patient care.
“Our method is an improvement because it can be done rapidly, potentially at the point of care since it doesn’t require extensive tissue processing,” Richards-Kortum told redOrbit via email.
“One of the places that the technique we developed could be most useful is in low-resource settings that currently don’t have sufficient infrastructure or human resources to implement traditional diagnostic techniques.”
(Image credit: Thinkstock)
Comments