Shumpei Ishikawa

Cancer histological images contain rich biological and clinical information, but quantitative representation can be problematic and has prevented the direct comparison and accumulation of large-scale datasets. Here, we show successful universal encoding of cancer histology by deep texture representations (DTRs) produced by a bilinear convolutional neural network. DTR-based, unsupervised histological profiling, which captures the morphological diversity, is applied to cancer biopsies and reveals

Frequent chromosomal aberrations and/or losses of heterozygosity involving the short arm of chromosome 3 in carcinomas of the lung, kidney and other tissues imply that multiple putative tumor suppressor genes may be present on this chromosomal arm. To search for one of these genes, we determined DNA sequences in the genomic region at 3p22-21.3 where we had previously detected a homozygous deletion in a lung cancer cell line. The DNA sequence results of an about 685-kb region indicated that the s

Numerous cancer histopathology specimens have been collected and digitized over the past few decades. A comprehensive evaluation of the distribution of various cells in tumor tissue sections can provide valuable information for understanding cancer. Deep learning is suitable for achieving these goals; however, the collection of extensive, unbiased training data is hindered, thus limiting the production of accurate segmentation models. This study presents SegPath-the largest annotation dataset (>

The prognosis of gastric cancer (GC) is significantly affected by distant metastases and postoperative recurrences. Bone metastasis is one of the worst prognostic metastases in GC; however, its molecular mechanisms and predictive biomarkers remain elusive. In prostate and breast cancers, it has been reported that overexpression of Cadherin 11 (CDH11), a mesenchymal cell-cell contact factor, is known to be correlated with bone metastasis. Overexpression of CDH11 mRNA in bulk GC tissues has also b

Deep texture representations (DTRs) produced from a bilinear convolutional neural network allow objective quantification of tumor histopathology images effectively. They can be used for various analyses, including visualization of morphological correlation between histology images, content-based image retrieval (CBIR), and supervised learning. This protocol describes the simplified workflow to analyze DTRs from data preparation, visualization of the histological profile, and CBIR analysis, to su

Summary 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 r

Abstract 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 se