https://medicinainternaucv.files.wordpress.com/2013/02/exceso-de-confianza-como-causa-de-error-diagnc3b3stico.pdf

Overconfidence as a cause of diagnostic error in medicine.

The International Symposium on Biomedical Imaging (ISBI) held a grand challenge to evaluate computational systems for the automated detection of metastatic breast cancer in whole slide images of sentinel lymph node biopsies. Our team won both competitions in the grand challenge, obtaining an area under the receiver operating curve (AUC) of 0.925 for the task of whole slide image classification and a score of 0.7051 for the tumor localization task. A pathologist independently reviewed the same images, obtaining a whole slide image classification AUC of 0.966 and a tumor localization score of 0.733. Combining our deep learning system's predictions with the human pathologist's diagnoses increased the pathologist's AUC to 0.995, representing an approximately 85 percent reduction in human error rate. These results demonstrate the power of using deep learning to produce significant improvements in the accuracy of pathological diagnoses.

Deep Learning for Identifying Metastatic Breast Cancer

Lung cancer is the most prevalent cancer worldwide, and histopathological assessment is indispensable for its diagnosis. However, human evaluation of pathology slides cannot accurately predict patients' prognoses. In this study, we obtain 2,186 haematoxylin and eosin stained histopathology whole-slide images of lung adenocarcinoma and squamous cell carcinoma patients from The Cancer Genome Atlas (TCGA), and 294 additional images from Stanford Tissue Microarray (TMA) Database. We extract 9,879 quantitative image features and use regularized machine-learning methods to select the top features and to distinguish shorter-term survivors from longer-term survivors with stage I adenocarcinoma (P<0.003) or squamous cell carcinoma (P=0.023) in the TCGA data set. We validate the survival prediction framework with the TMA cohort (P<0.036 for both tumour types). Our results suggest that automatically derived image features can predict the prognosis of lung cancer patients and thereby contribute to precision oncology. Our methods are extensible to histopathology images of other organs.

/pdf/predicting-non-small-cell-lung-cancer-prognosis-by-fully-4qxfx4kc7z.pdf

Predicting non-small cell lung cancer prognosis by fully automated microscopic pathology image features.

A wide variety of research studies suggest that breakdowns in the diagnostic process result in a staggering toll of harm and patient deaths. These include autopsy studies, case reviews, surveys of patient and physicians, voluntary reporting systems, using standardised patients, second reviews, diagnostic testing audits and closed claims reviews. Although these different approaches provide important information and unique insights regarding diagnostic errors, each has limitations and none is well suited to establishing the incidence of diagnostic error in actual practice, or the aggregate rate of error and harm. We argue that being able to measure the incidence of diagnostic error is essential to enable research studies on diagnostic error, and to initiate quality improvement projects aimed at reducing the risk of error and harm. Three approaches appear most promising in this regard: (1) using ‘trigger tools’ to identify from electronic health records cases at high risk for diagnostic error; (2) using standardised patients (secret shoppers) to study the rate of error in practice; (3) encouraging both patients and physicians to voluntarily report errors they encounter, and facilitating this process.

https://qualitysafety.bmj.com/content/qhc/22/Suppl_2/ii21.full.pdf

The incidence of diagnostic error in medicine

Deep Learning for Classification of Colorectal Polyps on Whole-slide Images.

BACKGROUND

To the authors' knowledge, the frequency and clinical impact of errors in the anatomic pathology diagnosis of cancer have been poorly characterized to date.



METHODS

The authors examined errors in patients who underwent anatomic pathology tests to determine the presence or absence of cancer or precancerous lesions in four hospitals. They analyzed 1 year of retrospective errors detected through a standardized cytologic–histologic correlation process (in which patient same-site cytologic and histologic specimens were compared). Medical record reviews were performed to determine patient outcomes. The authors also measured the institutional frequency, cause (i.e., pathologist interpretation or sampling), and clinical impact of diagnostic cancer errors.



RESULTS

The frequency of errors in cancer diagnosis was found to be dependent on the institution (P < 0.001) and ranged from 1.79–9.42% and from 4.87–11.8% of all correlated gynecologic and nongynecologic cases, respectively. A statistically significant association was found between institution and error cause (P < 0.001); the cause of errors resulting from pathologic misinterpretation ranged from 5.0–50.7% (the remainder were due to clinical sampling). A statistically significant association was found between institution and assignment of the clinical impact of error (P < 0.001); the aggregated data demonstrated that for gynecologic and nongynecologic errors, 45% and 39%, respectively, were associated with harm. The pairwise kappa statistic for interobserver agreement on cause of error ranged from 0.118–0.737.



CONCLUSIONS

Errors in cancer diagnosis are reported to occur in up to 11.8% of all reviewed cytologic-histologic specimen pairs. To the authors' knowledge, little agreement exists regarding whether pathology errors are secondary to misinterpretation or poor clinical sampling of tissues and whether pathology errors result in serious harm. Cancer 2005. © 2005 American Cancer Society.

Frederick A. Meier

Papers

Clinical impact and frequency of anatomic pathology errors in cancer diagnoses.