Distraction index, part II: inflammation in nongynecologic cytology

Searching Images for Consensus: Can AI Remove Observer Variability in Pathology?

One of the major obstacles in reaching diagnostic consensus is observer variability. With the recent success of artificial intelligence, particularly the deep networks, the question emerges as to whether the fundamental challenge of diagnostic imaging can now be resolved. This article briefly reviews the problem and how eventually both supervised and unsupervised AI technologies could help to overcome it.

Comparing methods to measure error in gynecologic cytology and surgical pathology

CONTEXT

Both gynecologic cytology and surgical pathology use similar methods to measure diagnostic error, but differences exist between how these methods have been applied in the 2 fields.

OBJECTIVE

To compare the application of methods of error detection in gynecologic cytology and surgical pathology.

DATA SOURCES

Review of the literature.

CONCLUSIONS

There are several different approaches to measuring error, all of which have limitations. Measuring error using reproducibility as the gold standard is a common method to determine error. While error rates in gynecologic cytology are well characterized and methods for objectively assessing error in the legal setting have been developed, meaningful methods to measure error rates in clinical practice are not commonly used and little is known about the error rates in this setting. In contrast, in surgical pathology the error rates are not as well characterized and methods for assessing error in the legal setting are not as well defined, but methods to measure error in actual clinical practice have been characterized and preliminary data from these methods are now available concerning the error rates in this setting.

Cytologic features of squamous cell carcinoma in ThinPrep slides: evaluation of cases that performed poorly versus those that performed well in the College of American Pathologists Interlaboratory Comparison Program in Cervicovaginal Cytology

CONTEXT

Although the cytologic features of squamous cell carcinoma in ThinPrep specimens are well known, whether these features are different in cases that are easily identified than in cases that are more difficult to identify is not known.

OBJECTIVE

To determine the cytologic features of squamous cell carcinoma in ThinPrep specimens that are easy to identify versus those that are difficult.

DESIGN

The cytologic features of 6 cases of squamous cell carcinoma that performed poorly in the College of American Pathologists Interlaboratory Comparison Program were compared with 14 cases that performed extremely well.

RESULTS

After evaluation of multiple criteria, 7 different cytologic features were analyzed based on review by a consensus panel blinded to the performance of the cases. The feature that was most strongly associated with cases that performed poorly was the presence of Trichomonas vaginalis (5/6 [83%] vs 0/14; P <.001). The presence of marked nuclear pleomorphism was more common in cases that performed well (4/14 [28%] vs 0/6; P =.27), but was not significant. The number of tumor cells, the number of normal cells, and the presence of keratinization, pleomorphism, nucleoli, and diathesis were not significant. The most common misdiagnosis after Trichomonas vaginalis was reparative change.

CONCLUSIONS

The presence of Trichomonas is characteristic of cases of squamous cell carcinoma in ThinPrep slides that are often misdiagnosed in this program. While Trichomonas is identified by participants in some of these cases, a significant percentage of participants interpreted the findings as reparative, without identifying the organism. These results emphasize the importance of distracting factors, whether identified or not, in evaluating gynecologic cytology.