Iain Carmichael

Background: Tumor invasion and malignant transformation are spatiotemporally coordinated processes that shape clinical behavior, diagnostic classification, and treatment response. Malignant peripheral nerve sheath tumors (MPNSTs) are aggressive tumors that can arise sporadically or in people with neurofibromatosis type 1 (NF-1), yet how distinct spatial niches underlie MPNST intratumor heterogeneity, local invasion, or malignant transformation from benign precursors is not fully understood. Methods: We integrate spatial analysis of 50,896 transcriptomes from 15 human MPNSTs with deep learning image analysis to define the interplay between genetic alterations, the immune microenvironment, and epigenetic cell fate regulation in MPNST invasion and transformation. Results: Spatial transcriptomics reveal distinct infiltrating immune cells and immunomodulatory spatial clusters in sporadic compared to NF-1–associated MPNSTs. Local invasion is marked by spatially defined immunomodulatory and dedifferentiation programs. Analysis of 2 NF-1–associated MPNSTs with contiguous morphologically defined neurofibroma and MPNST components demonstrates that NF-1–associated tumors exhibit a similar malignant transformation process associated with dynamic immune infiltration, chromosome 9p loss, and dedifferentiation concomitant with increased cell proliferation. Finally, deep learning morphology encoders were trained on spatial transcriptomic data to predict spatial transcriptomic cluster identity and chromosome 9p loss from hematoxylin and eosin–stained images although the generalizability of such encoders remains a challenge. Conclusions: Our data provide a spatial framework for peripheral nerve sheath tumor transformation although the limited sample size of the present cohort will require additional validation of the proposed model.

Background: Histologic assessment of tumor tissue is fundamental to the diagnosis and management of bladder cancer. As digital pathology becomes more widely adopted and artificial intelligence (AI) tools rapidly advance, computational pathology has emerged as a promising avenue for extracting clinically meaningful information from pathology images. Objectives: This scoping review aims to provide a comprehensive overview of the current landscape of computational pathology in bladder cancer. We describe the commonly used AI and machine learning approaches for analyzing whole slide images (WSIs) and summarize their use in diagnosis, grading, staging, molecular classification, prognostication, and treatment response prediction in bladder cancer. Methods: This review was conducted in accordance with PRISMA-ScR 2018 guidelines. We searched PubMed/MEDLINE for primary research articles focused on computational or AI-based analysis of human bladder cancer histopathology images across predefined predictive tasks. Eligible studies underwent full-text review. Findings were synthesized narratively by task. Results: A total of 2055 abstracts were screened. Fifty-three studies met eligibility criteria and were included in the review. Conclusions: Computational pathology offers substantial promise for augmenting pathologist workflows and advancing personalized care in bladder cancer. Current studies remain largely retrospective, but there is a rapid pace of innovation and a clear opportunity to work towards real-world clinical implementation. Establishing shared multi-institutional datasets and standardized evaluation frameworks could help support future research and expedite clinical implementation of computational pathology tools for bladder cancer.

Cytomorphological analysis of the bone marrow aspirate (BMA) is pivotal for the diagnostic workup of a broad range of hematological disorders. However, this skill is error prone, highly complex, and time consuming. Deep learning–based models for the automatic classification of bone marrow cell morphology demonstrate the potential to improve diagnostic efficiency and accuracy. However, existing deep learning approaches in this field fall short of expert-level performance and lack generalizability beyond a single dataset. Working with multiple hematopathologists, we curated a dataset from the University of California, San Francisco, which included a training set of 30,394 images from 40 patients with morphologically normal marrows and a test set of 8507 images from 10 different patients, all derived from 400×-equivalent whole-slide images (WSIs). We then developed DeepHeme, a snapshot ensemble deep learning classifier, which outperformed previous models in accuracy while expanding the total number of differentiable cell classes. We externally validated DeepHeme using an independent dataset from the Memorial Sloan Kettering Cancer Center, which included 2694 images from 10 morphologically normal patients and 11,076 images from 655 patients with normal or diseased marrow, scanned using a different WSI system, demonstrating robust generalizability. At the level of individual cell classifications, we systematically compared DeepHeme’s diagnostic performance with that of three medical experts from different academic hospitals, demonstrating that DeepHeme achieved accuracy comparable to, or exceeding, that of human experts. Accurate and generalizable cell classification represents a step toward automated analysis of hematopathology slides and the development of quantitative, morphology-based, predictive markers.

For measuring the strength of visually observed subpopulation differences, the Population Difference Criterion is proposed to assess the statistical significance of visually observed subpopulation differences. It addresses the following challenges: in high-dimensional contexts, distributional models can be dubious; in high-signal contexts, conventional permutation tests give poor pairwise comparisons. We also make two other contributions: Based on a careful analysis we find that a balanced permutation approach is more powerful in high-signal contexts than conventional permutations. Another contribution is the quantification of uncertainty due to permutation variation via a bootstrap confidence interval. The practical usefulness of these ideas is illustrated in the comparison of subpopulations of modern cancer data.

The support vector machine (SVM) is a powerful and widely used classification algorithm. This paper uses the Karush-Kuhn-Tucker conditions to provide rigorous mathematical proof for new insights into the behavior of SVM. These insights provide perhaps unexpected relationships between SVM and two other linear classifiers: the mean difference and the maximal data piling direction. For example, we show that in many cases SVM can be viewed as a cropped version of these classifiers. By carefully exploring these connections we show how SVM tuning behavior is affected by characteristics including: balanced vs. unbalanced classes, low vs. high dimension, separable vs. non-separable data. These results provide further insights into tuning SVM via cross-validation by explaining observed pathological behavior and motivating improved cross-validation methodology. Finally, we also provide new results on the geometry of complete data piling directions in high dimensional space.

A recent paper (Balch et al. (), ‘Satellite conjunction analysis and the false confidence theorem’, arXiv 5 preprint arXiv:1706.08565.) presents the “false confidence theorem,” which has potentially broad implications for statistical inference using Bayesian posterior uncertainty. This theorem says that with an arbitrarily large (sampling/frequentist) probability, there exists a set that does not contain the true parameter value but has an arbitrarily large posterior probability. As the use of Bayesian methods has become increasingly popular in applications of science, engineering and business, it is critically important to understand when Bayesian procedures lead to problematic statistical inferences or interpretations. In this paper, we consider a number of examples demonstrating the paradoxical nature of false confidence to begin to understand the contexts in which the false confidence theorem does (and does not) play a meaningful role in statistical inference. Our examples illustrate that models involving marginalization to non‐linear, not one‐to‐one functions of multiple parameters play a key role in more extreme manifestations of false confidence. © 2018 John Wiley & Sons, Ltd.