A5082167694- Breast Cancer Treatment Studies
- Radiomics and Machine Learning in Medical Imaging
- Breast Lesions and Carcinomas
- Medical Imaging Techniques and Applications
- Cancer Diagnosis and Treatment
- Breast Implant and Reconstruction
- AI in cancer detection
- Reconstructive Surgery and Microvascular Techniques
- MRI in cancer diagnosis
- Cancer and Skin Lesions
- Metastasis and carcinoma case studies
- Advanced MRI Techniques and Applications
- Ultrasound Imaging and Elastography
- Digital Radiography and Breast Imaging
- Advanced Breast Cancer Therapies
- Salivary Gland Tumors Diagnosis and Treatment
- HER2/EGFR in Cancer Research
- Lymphoma Diagnosis and Treatment
- Photoacoustic and Ultrasonic Imaging
- Ultrasound and Hyperthermia Applications
- Proteoglycans and glycosaminoglycans research
- Cancer Genomics and Diagnostics
- Ultrasonics and Acoustic Wave Propagation
- Neutropenia and Cancer Infections
- Viral-associated cancers and disorders
Tokyo Medical and Dental University
2015-2025
Tokyo Medical and Dental University Hospital
2019-2024
Institute of Science Tokyo
2024
Tokyo Metropolitan Hiroo Hospital
2012
Breast Center
2011
We aimed to evaluate an artificial intelligence (AI) system that can detect and diagnose lesions of maximum intensity projection (MIP) in dynamic contrast-enhanced (DCE) breast magnetic resonance imaging (MRI). retrospectively gathered MIPs DCE MRI for training validation data from 30 7 normal individuals, 49 20 benign cases, 135 45 malignant respectively. Breast were indicated with a bounding box labeled as or by radiologist, while the AI was trained calculate possibilities malignancy using...
We aimed to use deep learning with convolutional neural networks (CNNs) discriminate images of benign and malignant breast masses on ultrasound shear wave elastography (SWE). retrospectively gathered 158 146 as training data for SWE. A model was constructed using several CNN architectures (Xception, InceptionV3, InceptionResNetV2, DenseNet121, DenseNet169, NASNetMobile) 50, 100, 200 epochs. analyzed SWE 38 35 test data. Two radiologists interpreted these through a consensus reading 5-point...
Deep convolutional generative adversarial networks (DCGANs) are newly developed tools for generating synthesized images. To determine the clinical utility of images, we generated breast ultrasound images and assessed their quality value. After retrospectively collecting 528 144 benign masses 529 216 malignant in breasts, were using a DCGAN with 50, 100, 200, 500, 1000 epochs. The (n = 20) original 40) evaluated by two radiologists, who scored them overall quality, definition anatomic...
Deep learning (DL) has become a remarkably powerful tool for image processing recently. However, the usefulness of DL in positron emission tomography (PET)/computed (CT) breast cancer (BC) been insufficiently studied. This study investigated whether model using images with multiple degrees PET maximum-intensity projection (MIP) contributes to increase diagnostic accuracy PET/CT classification BC. We retrospectively gathered 400 200 BC and non-BC patients training data. For each image, we...
We aimed to use generative adversarial network (GAN)-based anomaly detection diagnose images of normal tissue, benign masses, or malignant masses on breast ultrasound. retrospectively collected 531 ultrasound from 69 patients. Data augmentation was performed and 6372 (531 × 12) were available for training. Efficient GAN-based used construct a computational model detect anomalous lesions in calculate abnormalities as an score. Images 51 tissues, 48 72 analyzed the test data. The sensitivity,...
Microwave radar-based breast imaging technology utilizes the principle of radar, in which radio waves reflect at interface between target and normal tissues, have different permittivities. This study aims to investigate feasibility safety a portable microwave device clinical practice.We retrospectively collected data ten cancers nine women (median age: 66.0 years; range: 37-78 years) who had undergone examination before surgery. All were Japanese tumor sizes from 4 10 cm. Using five-point...
Tumor-associated macrophages (TAMs) and abnormalities in cancer cells affect progression response to therapy. TAMs are a major component of the tumor microenvironment (TME) breast cancer, with their invasion affecting clinical outcomes. Programmed death-ligand 1 (PD-L1), target immune checkpoint inhibitors, acts as suppressive signal for surrounding system; however, its expression effect on outcome unknown. In this study, we used high-throughput multiple immunohistochemistry spatially...
Objectives We sought to generate realistic synthetic breast ultrasound images and express virtual interpolation of tumors using a deep convolutional generative adversarial network (DCGAN). Methods After retrospective selection 528 benign masses, 529 malignant 583 normal breasts, 20 synthesized each were generated by the DCGAN. Fifteen changing value input vector. A total 60 evaluated 2 readers, who scored them on 5‐point scale (1, very good; 5, poor) then answered whether image was benign,...
In breast cancer (BC), the development of immunotherapy including immune checkpoint inhibitors has progressed. Tumor infiltrating lymphocytes (TILs) is one important factors for an response between tumor cells and in microenvironment, presence TILs been identified as predictors to chemotherapy. However, because complex mechanisms underlies crosstalk cells, relationship profiles microenvironment efficacy blocked unclear. Moreover, many cases cancer, quantitative analysis immuno-modification...
Objective The aim of this study was to evaluate the performance preoperative axillary lymph node assessment in breast cancer using time-of-flight 18F-fluorodeoxyglucose PET/computed tomography (TOF [F-18]FDG-PET/CT). Methods Eighty-two women with (mean age, 59.3 years; range, 30–84 years) underwent TOF [F-18]FDG-PET/CT scanning before surgery between January 2016 and June 2018 at our hospital. Visual analysis FDG uptake maximum standardized value (SUVmax) nodes were compared pathological...
Synthetic MRI reconstructs multiple sequences in a single acquisition. In the present study, we aimed to compare image quality and utility of synthetic with that conventional breast.We retrospectively collected imaging data 37 women (mean age: 55.1 years; range: 20-78 years) who had undergone both T2-weighted, T1-weighted, fat-suppressed (FS)-T2-weighted images. Two independent breast radiologists evaluated overall quality, anatomical sharpness, contrast between tissues, homogeneity,...
Sarcopenia is associated with postoperative complications in patients undergoing digestive surgery. In this study, we investigated the impact of preoperative sarcopenia on breast cancer who underwent total mastectomy. Patients mastectomy were included analysis. The relationship between presence and (e.g., skin flap necrosis seroma) incidence these as well surgical factors was investigated. Moreover, effects recurrence-free survival overall evaluated. psoas muscle index calculated using...
Decorin (DCN), an extracellular matrix proteoglycan found in tumor surrounding tissues, is a natural inhibitor of cell proliferation and invasion. We conducted cross-sectional observation study to evaluate the association pathological stage with levels DCN plasma or tissue. Among 118 patients who underwent breast surgery, 35 were designated as carcinoma situ (Stage 0), 39 Stage I, 44 II III. The stromal expression was quantified using semiquantitative digital image analysis after...
The activation of signal transducer and activator transcription 3 (STAT3) has been reported in several types cancer, where it acts as an oncogene. However, breast the clinical role STAT3 remains unclear. In present study, association between phosphorylated-STAT3 (p-STAT3) expression clinicopathological/biological factors was examined each subtype. p-STAT3 135 cases cancer by immunohistochemistry. not associated with prognosis a complete cohort cases. patients estrogen receptor-negative...
Background and Objectives: This study compares the clinical properties of original breast ultrasound images those synthesized by a generative adversarial network (GAN) to assess usefulness GAN-synthesized images. Materials Methods: We retrospectively collected approximately 200 for each five representative histological tissue types (cyst, fibroadenoma, scirrhous, solid, tubule-forming invasive ductal carcinomas) as training A deep convolutional GAN (DCGAN) image-generation model types. Two...