Biology subjects

Komura, D.

Publications and source records attributed to Komura, D..

7 recordsLinked to original sources

bioRxiv2025

Label Noise in Pathological Segmentation is Overlooked, Leading to Potential Overestimation of Artificial Intelligence Models

Artificial intelligence (AI) has transformed medical imaging, driving advancements in radiology and endoscopy. Semantic segmentation, a pixel-level technique crucial for delineating pathological features, has become a cornerstone of digital pathology. Pathology segmentation AI models are often trained using annotations generated by pathologists. Despite the meticulous care typically exercised, these annotations frequently contain empirical label noise. However, the specific types of label noise in pathology data and their impact on AI model training remain inadequately explored. This study systematically investigated the effects of label noise on the performance of pathology segmentation models. Using publicly available datasets and a breast cancer semantic segmentation dataset, modules were developed to simulate four types of artificial label noise at varying intensity levels. These datasets were used to train deep learning models with encoder-decoder architectures, and their performance was evaluated using metrics such as the Dice coefficient, precision, recall, and intersection over union. The results indicated that models were highly susceptible to overfitting label noise, particularly boundary-dependent noise such as dilation and shrinkage. Discrepancies were identified between apparent performance scores obtained under real-world conditions and true performance scores derived using clean test data. This overestimation risk was most pronounced for datasets containing boundary-altering noise. Furthermore, random combinations of noise types and levels significantly impaired model generalization. This study underscores the critical importance of addressing label noise in pathology datasets. It is proposed that future efforts focus on developing standardized methods for quantifying and mitigating label noise, along with creating robust benchmarks using noise-inclusive datasets. Enhancing annotation quality and addressing label noise can improve the reliability and generalizability of AI in pathology, facilitating broader clinical adoption.

bioinformatics↗

bioRxiv2022

Single-Cell Transcriptome Analyses Reveal the Cell Diversity and Developmental Features of Human Gastric and Metaplastic Mucosa

The stomach is an important digestive organ with a variety of biological functions. However, due to the complexity of its cellular and glandular composition, the precise cellular biology has yet to be elucidated. In this study, we conducted single-cell RNA sequence analysis of the human stomach and constructed a 137,610-cell dataset, the largest cell atlas reported to date. By integrating this single-cell analysis with spatial cellular distribution analysis, we were able to clarify novel aspects of the developmental and tissue homeostatic ecosystems in the human stomach. We identified LEFTY1+ as a potential stem cell marker in both gastric and intestinal metaplastic glands. We also revealed skewed distribution patterns for PDGFRA+BMP4+WNT5A+ fibroblasts that play pivotal roles in, or even precede, the phenotypic changes from gastric to metaplastic mucosa. Our extensive dataset will function as a fundamental resource in investigations of the stomach, including studies on development, aging, and carcinogenesis.

genomics↗

bioRxiv2022

Beyond pathologist-level annotation of large-scale cancer histology for semantic segmentation using immunofluorescence restaining

Numerous cancer histopathology specimens have been collected and digitised as whole slide images over the past few decades. A comprehensive evaluation of the distribution of various cells in a section of tumour tissue can provide valuable information for understanding cancer and making accurate cancer diagnoses. Deep learning is one of the most suitable techniques to achieve these goals; however, the collection of large, unbiased training data has been a barrier to producing accurate segmentation models. Here, we developed a pipeline to generate SegPath, the largest annotation dataset that is over one order of magnitude larger than publicly available annotations, for the segmentation of haematoxylin and eosin (H&E)-stained sections for eight major cell types. The pipeline used H&E-stained sections that were destained and subsequently immunofluorescence-stained with carefully selected antibodies. The results showed that SegPath is comparable to, or significantly outperforms, conventional pathologist annotations. Moreover, we revealed that annotations by pathologists are biased toward typical morphologies; however, the model trained on SegPath can overcome this limitation. Our results provide foundational datasets for the histopathology machine learning community.

pathology↗

bioRxiv2021

Viola: a structural variant signature extractor with user-defined classifications

Here, we present Viola, a Python package that provides structural variant (SV; large scale genome DNA variations that can result in disease, e.g., cancer) signature analytical functions and utilities for custom SV classification, merging multi-SV-caller output files, and SV annotation. We demonstrate that Viola can extract biologically meaningful SV signatures from publicly available SV data for cancer and we evaluate the computational time necessary for annotation of the data. AvailabilityViola is available on pip (https://pypi.org/project/Viola-SV/) and on GitHub (https://github.com/dermasugita/Viola-SV). Contactkdais-prm@m.u-tokyo.ac.jp

bioinformatics↗

bioRxiv2020

Deep Texture Representations as a Universal Encoder for Pan-cancer Histology

Cancer histological images contain rich biological and clinical information, but quantitative representation can be problematic and has prevented direct comparison and accumulation of large-scale datasets. Here we show that deep texture representations (DTRs) produced by a bilinear Convolutional Neural Network, express cancer morphology well in an unsupervised manner, and work as a universal encoder for cancer histology. DTRs are useful for content-based image retrieval, enabling quick retrieval of histologically similar images from optimally area selected datasets of 7,175 cases from The Cancer Genome Atlas. Via comprehensive comparison with driver and clinically actionable gene mutations, we have successfully predicted 309 combinations of genomic features and cancer types from hematoxylin and eosin-stained images at high accuracy (AUC > 0.70 and q < 0.02). With its mounting capabilities on accessible devices such as smartphones, DTR-based encoding for cancer histology has a potentially strong impact on global equalization for cancer diagnosis and targeted therapies.

cancer biology↗

bioRxiv2018

Luigi: Large-scale histopathological image retrieval system using deep texture representations

BackgroundAs a large number of digital histopathological images have been accumulated, there is a growing demand of content-based image retrieval (CBIR) in pathology for educational, diagnostic, or research purposes. However, no CBIR systems in digital pathology are publicly available.\n\nResultsWe developed a web application, the Luigi system, which retrieves similar histopathological images from various cancer cases. Using deep texture representations computed with a pre-trained convolutional neural network as an image feature in conjunction with an approximate nearest neighbor search method, the Luigi system provides fast and accurate results for any type of tissue or cell without the need for further training. In addition, users can easily submit query images of an appropriate scale into the Luigi system and view the retrieved results using our smartphone application. The cases stored in the Luigi database are obtained from The Cancer Genome Atlas with rich clinical, pathological, and molecular information. We tested the Luigi system by querying typical cancerous regions from four cancer types, and confirmed successful retrieval of relevant images.\n\nConclusionsThe Luigi system will help students, pathologists, and researchers easily retrieve histopathological images of various cancers similar to those of the query image.

pathology↗

bioRxiv2017

Capturing the Difference in Humoral Immunity between Normal and Tumor Environments from RNA Sequences of B-Cell Receptors Using Supervised Machine Learning

The recent success of immunotherapy in treating tumors has attracted increasing interest in research related to the adaptive immune system in the tumor microenvironment. Recent advances in next-generation sequencing technology enabled the sequencing of whole T-cell receptors (TCRs) and B-cell receptors (BCRs)/immunoglobulins (Igs) in the tumor microenvironment. Since BCRs/Igs in tumor tissues have high affinities for tumor-specific antigens, the patterns of their amino acid sequences and other sequence-independent features such as the number of somatic hypermutations (SHMs) may differ between the normal and tumor microenvironments. However, given the high diversity of BCRs/Igs and the rarity of recurrent sequences among individuals, it is far more difficult to capture such differences in BCR/Ig sequences than in TCR sequences. The aim of this study was to explore the possibility of discriminating BCRs/Igs in tumor and in normal tissues, by capturing these differences using supervised machine learning methods applied to RNA sequences of BCRs/Igs.\n\nRNA sequences of BCRs/Igs were obtained from matched normal and tumor specimens from 90 gastric cancer patients. BCR/Ig-features obtained in Rep-Seq were used to classify individual BCR/Ig sequences into normal or tumor classes. Different machine learning models using various features were constructed as well as gradient boosting machine (GBM) classifier combining these models. The results demonstrated that BCR/Ig sequences between normal and tumor microenvironments exhibit their differences. Next, by using a GBM trained to classify individual BCR/Ig sequences, we tried to classify sets of BCR/Ig sequences into normal or tumor classes. As a result, an area under the curve (AUC) value of 0.826 was achieved, suggesting that BCR/Ig repertoires have distinct sequence-level features in normal and tumor tissues.\n\nTo the best of our knowledge, this is the first study to show that BCR/Ig sequences derived from tumor and normal tissues have globally distinct patterns, and that these tissues can be effectively differentiated using BCR/Ig repertoires.

bioinformatics↗