Tissue Contamination Challenges the Credibility of Machine Learning Models in Real World Digital Pathology.

Machine learning (ML) models are poised to transform surgical pathology practice. The most successful use attention mechanisms to examine whole slides, identify which areas of tissue are diagnostic, and use them to guide diagnosis. Tissue contaminants, such as floaters, represent unexpected tissue. Although human pathologists are extensively trained to consider and detect tissue contaminants, we examined their impact on ML models. We trained 4 whole-slide models. Three operate in placenta for the following functions: (1) detection of decidual arteriopathy, (2) estimation of gestational age, and (3) classification of macroscopic placental lesions. We also developed a model to detect prostate cancer in needle biopsies. We designed experiments wherein patches of contaminant tissue are randomly sampled from known slides and digitally added to patient slides and measured model performance. We measured the proportion of attention given to contaminants and examined the impact of contaminants in the t-distributed stochastic neighbor embedding feature space. Every model showed performance degradation in response to one or more tissue contaminants. Decidual arteriopathy detection--balanced accuracy decreased from 0.74 to 0.69 ± 0.01 with addition of 1 patch of prostate tissue for every 100 patches of placenta (1% contaminant). Bladder, added at 10% contaminant, raised the mean absolute error in estimating gestational age from 1.626 weeks to 2.371 ± 0.003 weeks. Blood, incorporated into placental sections, induced false-negative diagnoses of intervillous thrombi. Addition of bladder to prostate cancer needle biopsies induced false positives, a selection of high-attention patches, representing 0.033 mm2, and resulted in a 97% false-positive rate when added to needle biopsies. Contaminant patches received attention at or above the rate of the average patch of patient tissue. Tissue contaminants induce errors in modern ML models. The high level of attention given to contaminants indicates a failure to encode biological phenomena. Practitioners should move to quantify and ameliorate this problem.

BRAF Exon 15 Mutations in the Evaluation of Well-Differentiated Epithelial Nephroblastic Neoplasms in Children: A Report From the Children's Oncology Group Study AREN03B2.

CONTEXT.­: The distinction between well-differentiated epithelial favorable-histology Wilms tumor (EFHWT) and metanephric adenoma (MA) in children has historically been determined by the required absence of both a fibrous pseudocapsule and mitotic activity in MA. More recently these features have been allowed in adult MA. Mutations in exon 15 of the BRAF gene are reported in up to 88% of MAs but have not been reported in EFHWTs. OBJECTIVE.­: To clarify the pathologic and molecular features used to distinguish between pediatric MA and EFHWT. DESIGN.­: Stage I epithelial tumors classified as EFHWT on central review (36 patients) were identified from the Children's Oncology Group AREN03B2 study. Thirteen tumors had morphologic features overlapping those of MA and 23 lacked such features; 35 of 36 had tissue available for sequencing of BRAF. RESULTS.­: Patients with EFHWTs with MA features (13) were older (mean, 8.4 versus 1.9 years; P < .001), had smaller tumor diameters (mean, 6.0 versus 9.7 cm; P < .001), and had fewer mitoses (mean, 1 versus 48 mitoses per 10 high-power fields; P < .001) than patients with EFHWT lacking MA features (23). All EFHWTs with MA features contained at least a partial fibrous pseudocapsule; 7 of 12 (58%) had BRAF exon 15 mutation. No BRAF exon 15 mutations were identified in 23 EFHWTs lacking MA features. None of the 13 EFHWT patients with MA features have experienced relapse (median follow-up 5.9 years). CONCLUSIONS.­: Pediatric epithelial neoplasms with features of MA that show partial encapsulation and/or modest mitotic activity may be classified as MAs. Although BRAF mutation supports the diagnosis of MA, it is not required for the diagnosis.

Enhancing Automatic Placenta Analysis through Distributional Feature Recomposition in Vision-Language Contrastive Learning.

The placenta is a valuable organ that can aid in understanding adverse events during pregnancy and predicting issues post-birth. Manual pathological examination and report generation, however, are laborious and resource-intensive. Limitations in diagnostic accuracy and model efficiency have impeded previous attempts to automate placenta analysis. This study presents a novel framework for the automatic analysis of placenta images that aims to improve accuracy and efficiency. Building on previous vision-language contrastive learning (VLC) methods, we propose two enhancements, namely Pathology Report Feature Recomposition and Distributional Feature Recomposition, which increase representation robustness and mitigate feature suppression. In addition, we employ efficient neural networks as image encoders to achieve model compression and inference acceleration. Experiments validate that the proposed approach outperforms prior work in both performance and efficiency by significant margins. The benefits of our method, including enhanced efficacy and deployability, may have significant implications for reproductive healthcare, particularly in rural areas or low- and middle-income countries.