Researcher Database
Last Updated :2024/08/15
researchmap
Profile and Settings
Affiliation
Degree
Profile and Settings
Name (Japanese)
NakamotoName (Kana)
TakahiroName
202101007944462536
Affiliation
Achievement
Research Interests
Research Areas
Research Experience
- 2021/03 - Today Hokkaido University Faculty of Health Sciences
- 2018/04 - 2021/02 日本学術振興会特別研究員PD (東京大学)
- 2017/04 - 2018/03 東京大学医学部附属病院 放射線科 特任研究員
- 2015/04 - 2017/03 日本学術振興会特別研究員DC2 (九州大学)
Education
Awards
- 2024 第127回日本医学物理学会学術大会 Student Encouragement Awards
A feasible study for classification of acute radiation-induced xerostomia risk based on a dosiomics受賞者: Sora Takagi;Takahiro Nakamoto - 2017 医用画像情報学会（MII) 平成29年度内田論文賞
A framework for estimating four-dimensional dose distributions during stereotactic body radiation therapy based on a 2D/3D registration technique with an adaptive transformation parameter approach受賞者: Takahiro Nakamoto;Hidetaka Arimura - 2016 The 22nd International conference on Medical Physics (ICMP) Best Presentation Silver Award of IOMP
2D/3D registration-based framework for estimating 4D dose distributions according to dynamic images of an EPID during SBRT for lung cancer受賞者: Takahiro Nakamoto;Hidetaka Arimura - 2015 日本医学物理学会 第109回日本医学物理学会大会長賞
Improvement of automated monitoring approach of 4D dose distributions during SBRT based on 2D/3D registration with adaptive transformation parameters受賞者: 仲本宗泰;有村秀孝 - 2013 日本医学物理学会 第106回日本医学物理学会大会長賞
肺定位放射線治療時に取得した高エネルギーX線動画像に基づく4次元線量分布自動推定法の開発受賞者: 仲本宗泰;有村秀孝
Published Papers
- Takahiro Nakamoto, Kanabu Nawa, Kei Nishiyama, Kosuke Yoshida, Daizo Saito, Masahito Horiguchi, Yuki Shinya, Takeshi Ohta, Sho Ozaki, Yuki Nozawa, Masanari Minamitani, Toshikazu Imae, Osamu Abe, Hideomi Yamashita, Keiichi NakagawaPhysica Medica 125 103425 - 103425 1120-1797 2024/09 [Refereed]
- Takahiro Nakamoto, Hideomi Yamashita, Haruka Jinnouchi, Kanabu Nawa, Toshikazu Imae, Shigeharu Takenaka, Atsushi Aoki, Takeshi Ohta, Sho Ozaki, Yuki Nozawa, Keiichi NakagawaPhysica Medica 117 103182 - 103182 1120-1797 2024/01 [Refereed]
- Jinnouchi Haruka, Yamashita Hideomi, Nozawa Yuki, Nakamoto Takahiro, Sawayanagi Subaru, Katano AtsutoJournal of Cancer Research and Therapeutics 2023/04/07 [Refereed]
- Erika Yamazawa, Satoshi Takahashi, Masahiro Shin, Shota Tanaka, Wataru Takahashi, Takahiro Nakamoto, Yuichi Suzuki, Hirokazu Takami, Nobuhito SaitoCancers 14 (13) 3264 - 3264 2022/07/03 [Refereed]
Chordoma and chondrosarcoma share common radiographic characteristics yet are distinct clinically. A radiomic machine learning model differentiating these tumors preoperatively would help plan surgery. MR images were acquired from 57 consecutive patients with chordoma (N = 32) or chondrosarcoma (N = 25) treated at the University of Tokyo Hospital between September 2012 and February 2020. Preoperative T1-weighted images with gadolinium enhancement (GdT1) and T2-weighted images were analyzed. Datasets from the first 47 cases were used for model creation, and those from the subsequent 10 cases were used for validation. Feature extraction was performed semi-automatically, and 2438 features were obtained per image sequence. Machine learning models with logistic regression and a support vector machine were created. The model with the highest accuracy incorporated seven features extracted from GdT1 in the logistic regression. The average area under the curve was 0.93 ± 0.06, and accuracy was 0.90 (9/10) in the validation dataset. The same validation dataset was assessed by 20 board-certified neurosurgeons. Diagnostic accuracy ranged from 0.50 to 0.80 (median 0.60, 95% confidence interval 0.60 ± 0.06%), which was inferior to that of the machine learning model (p = 0.03), although there are some limitations, such as the risk of overfitting and the lack of an extramural cohort for truly independent final validation. In summary, we created a novel MRI-based machine learning model to differentiate skull base chordoma and chondrosarcoma from multiparametric signatures. - Sho Ozaki, Shizuo Kaji, Kanabu Nawa, Toshikazu Imae, Atsushi Aoki, Takahiro Nakamoto, Takeshi Ohta, Yuki Nozawa, Hideomi Yamashita, Akihiro Haga, Keiichi NakagawaMedical Physics 49 (6) 3769 - 3782 0094-2405 2022/03/22 [Refereed]
Purpose: In recent years, deep learning–based image processing has emerged as a valuable tool for medical imaging owing to its high performance. However, the quality of deep learning–based methods heavily relies on the amount of training data; the high cost of acquiring a large data set is a limitation to their utilization in medical fields. Herein, based on deep learning, we developed a computed tomography (CT) modality conversion method requiring only a few unsupervised images. Methods: The proposed method is based on cycle-consistency generative adversarial network (CycleGAN) with several extensions tailored for CT images, which aims at preserving the structure in the processed images and reducing the amount of training data. This method was applied to realize the conversion of megavoltage computed tomography (MVCT) to kilovoltage computed tomography (kVCT) images. Training was conducted using several data sets acquired from patients with head and neck cancer. The size of the data sets ranged from 16 slices (two patients) to 2745 slices (137 patients) for MVCT and 2824 slices (98 patients) for kVCT. Results: The required size of the training data was found to be as small as a few hundred slices. By statistical and visual evaluations, the quality improvement and structure preservation of the MVCT images converted by the proposed model were investigated. As a clinical benefit, it was observed by medical doctors that the converted images enhanced the precision of contouring. Conclusions: We developed an MVCT to kVCT conversion model based on deep learning, which can be trained using only a few hundred unpaired images. The stability of the model against changes in data size was demonstrated. This study promotes the reliable use of deep learning in clinical medicine by partially answering commonly asked questions, such as “Is our data sufficient?” and “How much data should we acquire?”. - Satoshi Kida, Shizuo Kaji, Kanabu Nawa, Toshikazu Imae, Takahiro Nakamoto, Sho Ozaki, Takeshi Ohta, Yuki Nozawa, Keiichi NakagawaMedical Physics 47 (3) 998 - 1010 0094-2405 2020 [Refereed]
PURPOSE: Cone-beam computed tomography (CBCT) offers advantages over conventional fan-beam CT in that it requires a shorter time and less exposure to obtain images. However, CBCT images suffer from low soft-tissue contrast, noise, and artifacts compared to conventional fan-beam CT images. Therefore, it is essential to improve the image quality of CBCT. METHODS: In this paper, we propose a synthetic approach to translate CBCT images with deep neural networks. Our method requires only unpaired and unaligned CBCT images and planning fan-beam CT (PlanCT) images for training. The CBCT images and PlanCT images may be obtained from other patients as long as they are acquired with the same scanner settings. Once trained, three-dimensionally reconstructed CBCT images can be directly translated into high-quality PlanCT-like images. RESULTS: We demonstrate the effectiveness of our method with images obtained from 20 prostate patients, and provide a statistical and visual comparison. The image quality of the translated images shows substantial improvement in voxel values, spatial uniformity, and artifact suppression compared to those of the original CBCT. The anatomical structures of the original CBCT images were also well preserved in the translated images. CONCLUSIONS: Our method produces visually PlanCT-like images from CBCT images while preserving anatomical structures. - Sho Ozaki, Akihiro Haga, Edward Chao, Calvin Maurer, Kanabu Nawa, Takeshi Ohta, Takahiro Nakamoto, Yuki Nozawa, Taiki Magome, Masahiro Nakano, Keiichi NakagawaThe Journal of Medical Investigation 67 (1.2) 30 - 39 1343-1420 2020 [Refereed]
Statistical iterative reconstruction is expected to improve the image quality of computed tomography (CT). However, one of the challenges of iterative reconstruction is its large computational cost. The purpose of this review is to summarize a fast iterative reconstruction algorithm by optimizing reconstruction parameters. Megavolt projection data was acquired from a TomoTherapy system and reconstructed using in-house statistical iterative reconstruction algorithm. Total variation was used as the regularization term and the weight of the regularization term was determined by evaluating signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), and visual assessment of spatial resolution using Gammex and Cheese phantoms. Gradient decent with an adaptive convergence parameter, ordered subset expectation maximization (OSEM), and CPU/GPU parallelization were applied in order to accelerate the present reconstruction algorithm. The SNR and CNR of the iterative reconstruction were several times better than that of filtered back projection (FBP). The GPU parallelization code combined with the OSEM algorithm reconstructed an image several hundred times faster than a CPU calculation. With 500 iterations, which provided good convergence, our method produced a 512 × 512 pixel image within a few seconds. The image quality of the present algorithm was much better than that of FBP for patient data. - Toshikazu Imae, Shizuo Kaji, Satoshi Kida, Kanako Matsuda, Shigeharu Takenaka, Atsushi Aoki, Takahiro Nakamoto, Sho Ozaki, Kanabu Nawa, Hideomi Yamashita, Keiichi Nakagawa, Osamu AbeNihon Hoshasen Gijutsu Gakkai zasshi 76 (11) 1173 - 1184 2020 [Refereed]
PURPOSE: Volumetric modulated arc therapy (VMAT) can acquire projection images during rotational irradiation, and cone-beam computed tomography (CBCT) images during VMAT delivery can be reconstructed. The poor quality of CBCT images prevents accurate recognition of organ position during the treatment. The purpose of this study was to improve the image quality of CBCT during the treatment by cycle generative adversarial network (CycleGAN). METHOD: Twenty patients with clinically localized prostate cancer were treated with VMAT, and projection images for intra-treatment CBCT (iCBCT) were acquired. Synthesis of PCT (SynPCT) with improved image quality by CycleGAN requires only unpaired and unaligned iCBCT and planning CT (PCT) images for training. We performed visual and quantitative evaluation to compare iCBCT, SynPCT and PCT deformable image registration (DIR) to confirm the clinical usefulness. RESULT: We demonstrated suitable CycleGAN networks and hyperparameters for SynPCT. The image quality of SynPCT improved visually and quantitatively while preserving anatomical structures of the original iCBCT. The undesirable deformation of PCT was reduced when SynPCT was used as its reference instead of iCBCT. CONCLUSION: We have performed image synthesis with preservation of organ position by CycleGAN for iCBCT and confirmed the clinical usefulness. - Satoshi Takahashi, Wataru Takahashi, Shota Tanaka, Akihiro Haga, Takahiro Nakamoto, Yuichi Suzuki, Akitake Mukasa, Shunsaku Takayanagi, Yosuke Kitagawa, Taijun Hana, Takahide Nejo, Masashi Nomura, Keiichi Nakagawa, Nobuhito SaitoInternational Journal of Radiation Oncology*Biology*Physics 105 (4) 784 - 791 0360-3016 2019 [Refereed]
Purpose: A noninvasive diagnostic method to predict the degree of malignancy accurately would be of great help in glioma management. This study aimed to create a highly accurate machine learning model to perform glioma grading.Methods and Materials: Preoperative magnetic resonance imaging acquired for cases of glioma operated on at our institution from October 2014 through January 2018 were obtained retrospectively. Six types of magnetic resonance imaging sequences (T-2-weighted image, diffusion-weighted image, apparent diffusion coefficient [ADC], fractional anisotropy, and mean kurtosis [MK]) were chosen for analysis; 476 features were extracted semiautomatically for each sequence (2856 features in total). Recursive feature elimination was used to select significant features for a machine learning model that distinguishes glioblastoma from lower-grade glioma (grades 2 and 3).Results: Fifty-five data sets from 54 cases were obtained (14 grade 2 gliomas, 12 grade 3 gliomas, and 29 glioblastomas), of which 44 and 11 data sets were used for machine learning and independent testing, respectively. We detected 504 features with significant differences (false discovery rate <0.05) between glioblastoma and lower-grade glioma. The most accurate machine learning model was created using 6 features extracted from the ADC and MK images. In the logistic regression, the area under the curve was 0.90 +/- 0.05, and the accuracy of the test data set was 0.91 (10 out of 11); using a support vector machine, they were 0.93 +/- 0.03 and 0.91 (10 out of 11), respectively (kernel, radial basis function; c = 1.0).Conclusions: Our machine learning model accurately predicted glioma tumor grade. The ADC and MK sequences produced particularly useful features. (C) 2019 Elsevier Inc. All rights reserved. - Takahiro Nakamoto, Wataru Takahashi, Akihiro Haga, Satoshi Takahashi, Shigeru Kiryu, Kanabu Nawa, Takeshi Ohta, Sho Ozaki, Yuki Nozawa, Shota Tanaka, Akitake Mukasa, Keiichi NakagawaScientific Reports 9 (1) 19411 - 19411 2019 [Refereed]
Abstract We conducted a feasibility study to predict malignant glioma grades via radiomic analysis using contrast-enhanced T1-weighted magnetic resonance images (CE-T1WIs) and T2-weighted magnetic resonance images (T2WIs). We proposed a framework and applied it to CE-T1WIs and T2WIs (with tumor region data) acquired preoperatively from 157 patients with malignant glioma (grade III: 55, grade IV: 102) as the primary dataset and 67 patients with malignant glioma (grade III: 22, grade IV: 45) as the validation dataset. Radiomic features such as size/shape, intensity, histogram, and texture features were extracted from the tumor regions on the CE-T1WIs and T2WIs. The Wilcoxon–Mann–Whitney (WMW) test and least absolute shrinkage and selection operator logistic regression (LASSO-LR) were employed to select the radiomic features. Various machine learning (ML) algorithms were used to construct prediction models for the malignant glioma grades using the selected radiomic features. Leave-one-out cross-validation (LOOCV) was implemented to evaluate the performance of the prediction models in the primary dataset. The selected radiomic features for all folds in the LOOCV of the primary dataset were used to perform an independent validation. As evaluation indices, accuracies, sensitivities, specificities, and values for the area under receiver operating characteristic curve (or simply the area under the curve (AUC)) for all prediction models were calculated. The mean AUC value for all prediction models constructed by the ML algorithms in the LOOCV of the primary dataset was 0.902 ± 0.024 (95% CI (confidence interval), 0.873–0.932). In the independent validation, the mean AUC value for all prediction models was 0.747 ± 0.034 (95% CI, 0.705–0.790). The results of this study suggest that the malignant glioma grades could be sufficiently and easily predicted by preparing the CE-T1WIs, T2WIs, and tumor delineations for each patient. Our proposed framework may be an effective tool for preoperatively grading malignant gliomas. - Mazen Soufi, Hidetaka Arimura, Takahiro Nakamoto, Taka-aki Hirose, Saiji Ohga, Yoshiyuki Umezu, Hiroshi Honda, Tomonari SasakiPhysica Medica 46 32 - 44 1120-1797 2018 [Refereed]
PURPOSE: We aimed to explore the temporal stability of radiomic features in the presence of tumor motion and the prognostic powers of temporally stable features. METHODS: We selected single fraction dynamic electronic portal imaging device (EPID) (n = 275 frames) and static digitally reconstructed radiographs (DRRs) of 11 lung cancer patients, who received stereotactic body radiation therapy (SBRT) under free breathing. Forty-seven statistical radiomic features, which consisted of 14 histogram-based features and 33 texture features derived from the graylevel co-occurrence and graylevel run-length matrices, were computed. The temporal stability was assessed by using a multiplication of the intra-class correlation coefficients (ICCs) between features derived from the EPID and DRR images at three quantization levels. The prognostic powers of the features were investigated using a different database of lung cancer patients (n = 221) based on a Kaplan-Meier survival analysis. RESULTS: Fifteen radiomic features were found to be temporally stable for various quantization levels. Among these features, seven features have shown potentials for prognostic prediction in lung cancer patients. CONCLUSIONS: This study suggests a novel approach to select temporally stable radiomic features, which could hold prognostic powers in lung cancer patients. - Subaru Sawayanagi, Hideomi Yamashita, Mami Ogita, Tomoki Kiritoshi, Takahiro Nakamoto, Osamu Abe, Keiichi NakagawaRadiation Oncology 13 (1) 2018 [Refereed]
- Satoshi Kida, Takahiro Nakamoto, Masahiro Nakano, Kanabu Nawa, Akihiro Haga, Jun'ichi Kotoku, Hideomi Yamashita, Keiichi NakagawaCureus 10 (4) 2168-8184 2018 [Refereed]
IntroductionCone beam computed tomography (CBCT) plays an important role in image-guided radiation therapy (IGRT), while having disadvantages of severe shading artifact caused by the reconstruction using scatter contaminated and truncated projections. The purpose of this study is to develop a deep convolutional neural network (DCNN) method for improving CBCT image quality.MethodsCBCT and planning computed tomography (pCT) image pairs from 20 prostate cancer patients were selected. Subsequently, each pCT volume was pre-aligned to the corresponding CBCT volume by image registration, thereby leading to registered pCT data (pCT(r)). Next, a 39-layer DCNN model was trained to learn a direct mapping from the CBCT to the corresponding pCT(r) images. The trained model was applied to a new CBCT data set to obtain improved CBCT (i-CBCT) images. The resulting i-CBCT images were compared to pCT(r) using the spatial non- uniformity (SNU), the peak-signal-to-noise ratio (PSNR) and the structural similarity index measure (SSIM).ResultsThe image quality of the i-CBCT has shown a substantial improvement on spatial uniformity compared to that of the original CBCT, and a significant improvement on the PSNR and the SSIM compared to that of the original CBCT and the enhanced CBCT by the existing pCT-based correction method.ConclusionWe have developed a DCNN method for improving CBCT image quality. The proposed method may be directly applicable to CBCT images acquired by any commercial CBCT scanner. - Yusuke Shibayama, Hidetaka Arimura, Taka-aki Hirose, Takahiro Nakamoto, Tomonari Sasaki, Saiji Ohga, Norimasa Matsushita, Yoshiyuki Umezu, Yasuhiko Nakamura, Hiroshi HondaMedical Physics 44 (5) 1837 - 1845 0094-2405 2017 [Refereed]
PURPOSE: The setup errors and organ motion errors pertaining to clinical target volume (CTV) have been considered as two major causes of uncertainties in the determination of the CTV-to-planning target volume (PTV) margins for prostate cancer radiation treatment planning. We based our study on the assumption that interfractional target shape variations are not negligible as another source of uncertainty for the determination of precise CTV-to-PTV margins. Thus, we investigated the interfractional shape variations of CTVs based on a point distribution model (PDM) for prostate cancer radiation therapy. MATERIALS AND METHODS: To quantitate the shape variations of CTVs, the PDM was applied for the contours of 4 types of CTV regions (low-risk, intermediate- risk, high-risk CTVs, and prostate plus entire seminal vesicles), which were delineated by considering prostate cancer risk groups on planning computed tomography (CT) and cone beam CT (CBCT) images of 73 fractions of 10 patients. The standard deviations (SDs) of the interfractional random errors for shape variations were obtained from covariance matrices based on the PDMs, which were generated from vertices of triangulated CTV surfaces. The correspondences between CTV surface vertices were determined based on a thin-plate spline robust point matching algorithm. The systematic error for shape variations was defined as the average deviation between surfaces of an average CTV and planning CTVs, and the random error as the average deviation of CTV surface vertices for fractions from an average CTV surface. RESULTS: The means of the SDs of the systematic errors for the four types of CTVs ranged from 1.0 to 2.0 mm along the anterior direction, 1.2 to 2.6 mm along the posterior direction, 1.0 to 2.5 mm along the superior direction, 0.9 to 1.9 mm along the inferior direction, 0.9 to 2.6 mm along the right direction, and 1.0 to 3.0 mm along the left direction. Concerning the random errors, the means of the SDs ranged from 0.9 to 1.2 mm along the anterior direction, 1.0 to 1.4 mm along the posterior direction, 0.9 to 1.3 mm along the superior direction, 0.8 to 1.0 mm along the inferior direction, 0.8 to 0.9 mm along the right direction, and 0.8 to 1.0 mm along the left direction. CONCLUSIONS: Since the shape variations were not negligible for intermediate and high-risk CTVs, they should be taken into account for the determination of the CTV-to-PTV margins in radiation treatment planning of prostate cancer. - Takahiro Nakamoto, Hidetaka Arimura, Tomonari Sasaki, Ken'ichi Morooka, Taka-aki Hirose, Yoshiyuki Umezu, Yasuhiko Nakamura, Hiroshi Honda, Hideki HirataMedical Imaging and Information Sciences 33 (3) 48 - 56 2016 [Refereed]
- Takahiro Nakamoto, Hidetaka Arimura, Katsumasa Nakamura, Yoshiyuki Shioyama, Asumi Mizoguchi, Taka-aki Hirose, Hiroshi Honda, Yoshiyuki Umezu, Yasuhiko Nakamura, Hideki HirataComputerized Medical Imaging and Graphics 40 1 - 12 0895-6111 2015 [Refereed]
A computerized framework for monitoring four-dimensional (4D) dose distributions during stereotactic body radiation therapy based on a portal dose image (PDI)-based 2D/3D registration approach has been proposed in this study. Using the PDI-based registration approach, simulated 4D "treatment" CT images were derived from the deformation of 3D planning CT images so that a 2D planning PDI could be similar to a 2D dynamic clinical PDI at a breathing phase. The planning PDI was calculated by applying a dose calculation algorithm (a pencil beam convolution algorithm) to the geometry of the planning CT image and a virtual water equivalent phantom. The dynamic clinical PDIs were estimated from electronic portal imaging device (EPID) dynamic images including breathing phase data obtained during a treatment. The parameters of the affine transformation matrix were optimized based on an objective function and a gamma pass rate using a Levenberg-Marquardt (LM) algorithm. The proposed framework was applied to the EPID dynamic images of ten lung cancer patients, which included 183 frames (mean: 18.3 per patient). The 4D dose distributions during the treatment time were successfully obtained by applying the dose calculation algorithm to the simulated 4D "treatment" CT images. The mean±standard deviation (SD) of the percentage errors between the prescribed dose and the estimated dose at an isocenter for all cases was 3.25±4.43%. The maximum error for the ten cases was 14.67% (prescribed dose: 1.50Gy, estimated dose: 1.72Gy), and the minimum error was 0.00%. The proposed framework could be feasible for monitoring the 4D dose distribution and dose errors within a patient's body during treatment. - Hidetaka Arimura, Genyu Kakiuchi, Yoshiyuki Shioyama, Shin-ichi Minohara, Takahiro Nakamoto, Katsumasa Nakamura, Hiroshi Honda, Mutsumi Tashiro, Tatsuaki Kanai, Hideki HirataInternational Journal of Intelligent Computing in Medical Sciences & Image Processing 6 (1) 1 - 16 1931-308X 2014 [Refereed]
- 溝口明日実, 有村秀孝, 塩山善之, 仲本宗泰, 吉留郷志, 廣瀬貴章, 穴井重男, 本田浩, 梅津芳幸, 平田秀紀, 大喜雅文, 中村和正, 豊福不可依電子情報通信学会論文誌. D, 情報・システム (一社)電子情報通信学会 96 (4) 813 - 823 1880-4535 2013 [Refereed]
本研究の目的は、放射線照射時のビーム方向像であるEPID(electronic portal imaging device)の動画像に基づき、患者体内での四次元線量分布を推定する手法を開発することである。第一に、治療計画CT画像と線量計算アルゴリズムを用いて治療計画時の患者体領域からの射出線量画像を求めた。第二に、照射時に撮影されたEPID動画像を用いて照射時の射出線量画像を推定した。第三に、治療計画時の射出線量画像を照射時の射出線量画像に線形レジストレーションさせ、体内の呼吸性移動を含むアフィン変換行列を求めた。第四に、そのアフィン変換行列を治療計画CT画像に適用し、"照射時"CT画像を推定した。最後に、EPID動画像のフレームごとに推定した。"照射時"CT画像において線量分布計算を行い、四次元線量分布を推定した。提案手法は治療成績の評価やリスク管理に役立つ可能性がある。(著者抄録)
MISC
- Satoshi Takahashi, Wataru Takahashi, Shota Tanaka, Akihiro Haga, Takahiro Nakamoto, Yuuichi Suzuki, Akitake Mukasa, Shunsaku Takayanagi BRAIN PATHOLOGY 29- 194 -194 2019/02
- Satoshi Kida, Shizuo Kaji, Kanabu Nawa, Toshikazu Imae, Takahiro Nakamoto, Sho Ozaki, Takeshi Ohta, Yuki Nozawa, Keiichi Nakagawa CoRR abs/1901.05773- 2019 [Not refereed][Not invited]
- Satoshi Takahashi, Wataru Takahashi, Shota Tanaka, Akihiro Haga, Takahiro Nakamoto, Akitake Mukasa, Shunsaku Takayanagi, Yuichi Suzuki, Tsukasa Koike, Yosuke Kitagawa, Taijyun Hana, Takahide Nejo, Masashi Nomura, Nobuhito Saito NEURO-ONCOLOGY 20- 188 -188 2018/11
- Radiomics入門仲本宗泰, 芳賀昭弘, 高橋渉 医学物理 38- (3) 129 -134 2018
- Tran Thi Thao Nguyen, Hidetaka Arimura, Yoshifumi Oku, Takahiro Nakamoto, Yusuke Shibayama, Takashi Yoshiura MEDICAL PHYSICS 43- (6) 3475 -3475 2016/06 [Not refereed][Not invited]
- 放射線治療における機械学習の応用有村秀孝, 仲本宗泰 医学物理 36- (1) 35 -38 2016
Books etc
- レディオミクス入門仲本宗泰 (Contributor)
オーム社 2021/10 - Image-Based Computer-Assisted Radiation TherapyTakahiro Nakamoto, Hidetaka Arimura (ContributorVisualization of dose distributions for photon beam radiation therapy during treatment delivery)
Springer 2017
Association Memberships
Research Projects
- 日本学術振興会：科学研究費助成事業Date (from‐to) : 2021/04 -2024/03Author : 今江 禄一, 名和 要武, 鍛冶 静雄, 竹中 重治, 仲本 宗泰, 尾崎 翔, 山下 英臣放射線治療において医用画像は治療前や治療期間内，治療後など多くの場面で用いられている．特に，治療期間内の位置照合時に得られる医用画像情報は情報量が少ない（以下，疎な）一方，治療の効果および副作用に関する生体情報を含有している可能性がある．近年の情報処理技術の発展に伴い，医用画像に対して深層学習を用いた画像処理や解析（以下，深層画像処理）が適用され始めているものの，処理の自由度が高いために解析の安定条件に課題があり，汎用的に利用されていないのが現状である．本研究では，放射線治療で得られる疎な医用画像情報に着目し，安全かつ有効に利用可能な深層画像処理の要件を勘案した上で，深層画像処理の安定化を図ることを目的とする．当該年度は以下のことを実施した． (1) 本研究では高精度放射線治療を実施する患者を対象とし，その基本情報の取得と蓄積を行った．深層画像処理の実施にはワークステーションおよびストレージといった計算環境の整備が必要であるため，円滑な研究の遂行が可能な環境の整備を行った． (2) 放射線治療の実施直前に取得されるメガボルトCT（megavoltage CT： MVCT）画像の画質は，位置合わせを主たる目的とするため，治療計画用キロボルトCT（kilovoltage CT： kVCT）よりも劣るのが現状であった．MVCT画像に対して画質改善を行う深層学習処理を提案し，適切なデータ数の探索を行った．得られた知見について論文投稿を行い，採択された． (3) 放射線治療を実施するためには，医用画像を用いて標的や正常組織を判別することが必要である．当該年度では深層画像処理を用いた組織の判別（セグメンテーション）について，その具体的手法や学習数，適切なハイパーパラメータについて，探索を行った．
- Japan Society for the Promotion of Science：Grants-in-Aid for Scientific ResearchDate (from‐to) : 2020/04 -2023/03Author : 中川 恵一, 名和 要武, 鍛冶 静雄, 野沢 勇樹, 仲本 宗泰, 太田 岳史, 尾崎 翔, 山下 英臣, 今江 禄一現代の医療において医用画像は不可欠であり，放射線治療においても治療前や治療期間内，治療後の多くの場面で用いられている．近年の情報処理技術の発展に伴い，医用画像に対して深層学習を用いた画像生成や解析（以下，深層画像処理）が適用され始めている．本研究では，放射線治療に用いられる医用画像に対して深層画像処理を施した上で，生成画像や解析結果を放射線治療の様々な状況において安全かつ有効に利用する方法を確立することを目的とした．当該年度は以下のことを実施した． (1) 深層学習用サーバーの環境セットアップを行った．このサーバーは高性能GPUを4枚搭載しており，複数の深層学習を同時に計算可能な環境を整えた． (2) 深層学習を用いた医用画像の変換・ノイズ除去・病変セグメンテーションについてまとめたレヴュー論文が2021年度 Radiological Physics and TechnologyのMost Citation Award に選出された． (3) 治療期間内に得られるメガボルトCT（megavoltage CT： MVCT）から画質が良好なキロボルトCT（kilovoltage CT： kVCT）の画像変換において，数百枚程度の従来よりも少ないデータ数で画像変換が可能である手法を提案，かつ，臨床的有用性を示した．得られた知見について論文投稿を行い，採択された． (4) 治療期間内に得られるコーンビームCT（cone-beam CT： CBCT）画像から抽出した定量的特徴量を用いて放射線治療後の予後予測モデルを構築した． (5) 深層画像処理を用いたセグメンテーションの開発を行った．対象は骨盤部とし，2次元から2.5次元，3次元へと拡張することに取り組んだ．
- 日本学術振興会：科学研究費助成事業Date (from‐to) : 2019/04 -2022/03Author : 長谷川 洋敬, 仲本 宗泰, 河島 真理子, 高橋 渉, 辛 正廣
- Japan Society for the Promotion of Science：Grants-in-Aid for Scientific ResearchDate (from‐to) : 2018/04 -2022/03Author : Nakamoto TakahiroThe purpose of the study was to develop a system for predicting radiation therapy prognosis based on a radiomics with patient’s variability in the treatment. Radiomic features with patient’s variability were extracted from multi-modal medical images acquired in and before the radiation therapy. We developed system for predicting radiation therapy prognosis and factors related to the prognosis using the radiomic features with patient’s variability. We have been suggested that the radiomics-based system with the patient’s variability would be feasible for predicting the radiation therapy prognosis.
- 日本学術振興会：科学研究費助成事業Date (from‐to) : 2018/04 -2021/03Author : 仲本 宗泰本年度は，放射線治療計画を目的として取得されるcomputed tomography (CT) 画像および治療期間中に位置合わせを目的として取得されるcone-beam CT (CBCT) 画像を収集し，放射線治療における患者統計的変動を含んだ多元的医用画像データベースへの構築を行なった．また画像データベースに加えて治療計画データである肉眼的腫瘍体積 (gross tumor volume: GTV) 輪郭情報および全生存期間(overall survive: OS)や無病生存期間 (disease-free survival: DFS) などの予後に関するデータの収集を行なった．そして多元的医用画像データベース内の症例毎のCBCT画像シリーズから治療計画CT画像およびGTV輪郭情報を用いて腫瘍内の統計的変動を含んだ画像特徴量を抽出した．さらに，腫瘍内の局所領域毎に画像特徴量を計算し腫瘍内の特徴量分布を直感的に把握できる特徴量マップ作成法の開発を行なった． 今後，CBCT画像から抽出した患者変動を含んだ画像特徴量と予後データに基づいて多元的潜在構造・因子 (multi-disciplinary potential structure and factor: MPSF) モデルを構築し治療期間中の患者統計変動を考慮した予後予測法について検討する．また特徴量マップで腫瘍内において相対的に特徴量の値が高い (もしくは低い) 領域のみで特徴量を抽出し，最終的に予後予測精度の向上へ繋がるか検討を行なう必要がある．そして，得られた研究成果を原著論文としてまとめ，海外の学術雑誌に投稿する予定である．
- Japan Society for the Promotion of Science：Grants-in-Aid for Scientific ResearchDate (from‐to) : 2016/08 -2018/03Author : Nakano Masahiro, IMAE Toshikazu, NAWA Kanabu, HAGA Akihiro, HANAOKA Shouhei, NAKAMOTO Takahiro, DEMACHI Kazuyuki, TAKAHASHI Wataru, NAKAGAWA Keiichi, HASHIMOTO Masatoshi, YOSHIOKA Yasuo, OGUCHI MasahikoThe purpose of this study was to predict head and neck cancer patients' body shrinkage and weight loss throughout radiotherapy duration, from daily cone-beam CT (CBCT) image sets. In the first year of the research, hyper-parameter optimization of deformable image registration (DIR), in order to acquire accurate deformation vector fields (DVFs) which contain information about patients' body shrinkage, was implemented. Then, in the second year, principal component vectors (PCVs) were acquired using principal component analysis (PCA) of ten DVFs from earlier 11 treatment fractions, and predicted DVF from the first fraction to the 23rd has successfully reconstructed using an average DVF and nine PCVs.
- 日本学術振興会：科学研究費助成事業Date (from‐to) : 2015/04 -2017/03Author : 仲本 宗泰本年度は患者の治療期間中に撮像された全てのEPID動画像を収集し，前年度に構築した治療中の変動を考慮した統計的多次元放射線治療症例データベースに治療総期間における統計変動についても組み込んだ．そして，治療期間中全ての変動を含んだ患者各々の統計的線量分布を予測した． 統計的線量分布はデータベース内に存在する治療期間中全ての統計的変動を含んだCT画像に線量計算アルゴリズム (AAA: analytical anisotropic algorithm) を適用することによって予測した．この統計的変動を含んだCT画像は治療計画CT画像を治療期間中に撮像された全てのEPID動画像のフレーム毎に変形させることにより計算した．変形には，前年度に提案したATP法を組み込んだ最適化法により決定される変形パラメーターに基づいた非線形レジストレーション法を用いた． 予測した統計的線量分布と治療計画時の線量分布を比較・評価するため，DVHを計算し，標的及びリスク臓器 (肺，脊髄) に対する様々な指標 (D95，V10，V20，mean dose，maximum dose，homogeneity index，conformity index) を求めた．さらに，生物学的指標であるTCP，NTCPについても求めた． 研究成果は2回の国際会議と1回の国内会議で報告し，大きな反響を得られた．そして，ICMP 2016では研究成果が高く評価されBest Presentation Silver Award of IOMPを受賞した．さらに，研究成果は英文原著論文として国内の医用画像分野の学術雑誌 Medical Imaging and Information Sciences誌 に掲載された．そして，本論文は高く評価され論文賞の受賞が内定した．最後に研究成果及び当該研究に関連する知識・技術を英文著書として纏めた．