top of page
aaaa.png
ssee.png
unnamed (1).jpg
ssssss.png

service
 Contact Us
                                         

STOP COVID-19　×　cutting-edge technologies Technology

Free_Sample_By_Wix (4).jpeg
mv_bi.jpg

ビッグデータ・BI

Big data/BI

​大数据/BI

mv_ai.jpg

AI（人工知能）人工智能

mv_iot.jpg

 IoT

gff.jpg
核心优势
fddd.png
gff3.jpg
ｆｆｆｆｆｆｆｆｄｄｄ.jpg
mv_robotics.jpg

ロボティクス

Robotics

​机器人

​

mv_payment.jpg

決済

settlement

清算

mv_cloud.jpg

クラウドコンピューティング

Cloud computing

​云计算

ビジネスアプリケーション

CRM（顧客管理）

ERP（統合業務）・経営管理

財務会計・IFRS

防災・危機管理

EDI

文書管理

SCM（サプライチェーン管理）

電子申請

ワークフロー

内部統制

環境・エネルギー

システム基盤設計

ライフサイエンス

テクノロジー

ブロックチェーン

ビッグデータ・BI

クラウドコンピューティング

決済

オープンソースソフトウェア

セキュリティ・認証

ウェブサービス

システム基盤設計

GIS・GPS

AI（人工知能）

IoT

ロボティクス

ITインフラ・運用

ネットワーク

データセンター

セキュリティ・認証

クラウドコンピューティング

アウトソーシング

決済

Business application
CRM (customer management)
ERP (Integrated Business)/Business Management
Financial accounting/IFRS
Disaster prevention/crisis management
EDI
Document management
SCM (Supply Chain Management)
Electronic application
Workflow
Internal control
Environment/Energy
System infrastructure design
life science
technology
Block chain
Big data/BI
Cloud computing
settlement
Open source software
Security/Authentication
Web service
System infrastructure design
GIS/GPS
AI (artificial intelligence)
IoT
Robotics
IT infrastructure/operation
network
Data center
Security/Authentication
Cloud computing
outsourcing
settlement

业务应用
CRM（客户管理）
ERP（综合业务）/业务管理
财务会计/ IFRS
防灾/危机管理
电子数据交换
文件管理
SCM（供应链管理）
电子应用
工作流程
内部控制
环境/能源
系统基础架构设计
生命科学
技术
区块链
大数据/ BI
云计算
沉降
开源软件
安全/认证
网络服务
系统基础架构设计
地理信息系统/全球定位系统
AI（人工智能）
物联网
机器人技术
IT基础架构/运营
网络
数据中心
安全/认证
云计算
外包
清算

SERIVICE- 需要予測- 在庫最適化- 異常検知- 与信評価- 物流最適化- 言語・画像認識- 教育等

AI training
yhyh.jpg
sssddd.jpg
kkjkkj.jpg
88.jpg
kkkk888.png
kooo.jpg
aa.jpg
jjhjhj.jpg
jjj.jpg
jijij.png
jjj0000.jpg

Getting to the root of cognitive impairment

We’re creating personalized digital therapeutics engineered to directly improve cognitive impairments. Our medicines look and feel like high-end interactive action video games but, unlike typical video games, they are informed by decades of neuroscience and built with proprietary technology developed to target and treat cognitive impairments at their specific sources in the brain.

Not only is our medicine designed to target the areas of your brain that need it most, it’s designed specifically for you, adapting based on your individual needs.

Explore the science and technology behind our digital treatments

Developed through the collaboration of world-renowned cognitive neuroscientists and acclaimed entertainment and technology designers, Akili has created a proprietary technology platform that represents an entirely new category of medicine.

Built on extensive peer-reviewed research, Akili’s products are designed to deliver sensory and motor stimuli to selectively target and activate specific cognitive neural systems in the brain. The proprietary technology is engineered to directly generate physiological changes in the brain to improve cognitive function.

A pioneering approach

We strive to make treatments that are not just engaging, but captivating to the point that, in the moment, patients forget they are taking medicine.

Our digital medicine is delivered through creative and immersive action video game experiences to keep patients engaged and immersed in the treatment. Our products leverage the fun, deep engagement, and rewards that make games incredibly compelling but, unlike typical video games, this personalized gameplay experience is engineered with adaptive algorithms designed to improve cognitive function.

EndeavorIpad.png
background.jpg
iStock-506584266-768x513.jpg
medical-scan-monitor-picture-id908045946

Our first program is Sleepio, a digital sleep improvement program

iStock-1027781002.jpg
jjhjhj.jpg
hhggg.jpg
iStock-494219400-768x468.jpg

COVID-19に対して高リスクとなりうる基礎疾患を持つ患者をAIで抽出する試みが、インドのムンバイで展開されている。同国のスタートアップ Radical Health-tech Private Limitedは、網膜画像の微細な変化からAIによって「非眼科疾患を含む併存疾患の有無」を検出するスクリーニング技術を有する。

インドの英字日刊紙 Hindustan Timesが26日に報じたところによると、Radical Health-techは6月から7月にかけてムンバイ市内15ヶ所で3,167人の網膜画像から基礎疾患スクリーニングを行った。1,255人に併存疾患が指摘され、その65%は自身の基礎疾患に気付いていなかった。緑内障と加齢黄斑変性が259人、糖尿病が188人、高血圧が376人、その他5人（喘息・黄疸・結核・尿路結石など）の疾患が発見された。3,167名はCOVID-19の重症化リスクとして、高リスク88名（2.78%)・中等度リスク713名（22.51%）・低リスク2,366名（74.71%）に分類された。

Radical Health-techの共同創設者 Rito Maitra氏によると、インドでは年に一度の健康診断を行わない人が多く、特に低所得者層を中心に自身の基礎疾患に気付いていない例が多数あるという。そのような患者らに、基礎疾患の意味を理解してもらい、今後の治療につながるカウンセリングを受けてもらうことには公衆衛生上の大きな意義がある。ムンバイ市はAIを社会的利益のために活用する非営利団体「Wadhwani Institute for Artificial Intelligence」と連携しており、同様の取り組みを今後も続けていく。

jjhjj.png
hhhh00.jpg
ggg7777.jpg
jijij.png

安全で効果的なワクチンが幅広く利用できるようになるまでは、新型コロナウイルス（SARS-CoV-2）の感染をくい止めることが最も重要となる。そのためには、誰が新型コロナウイルスに感染しているのかを知る必要がある。最近になり、検査能力は向上しているが、感染者の半数近く近くを占める無症状患者の検査に必要なレベルには程遠い。

私たちの研究によると、データサイエンスを活用することで、現在の検査能力を効果的に向上させられることが分かった。機械学習とプーリング検査（複数の人の検体を混ぜ合わせて一度に検査すること）を組み合わせれば、大規模な人口集団が毎週または毎日でも、1日1人当たり3ドルから5ドル程度で検査を受けられる。

つまり、検査ごとにかかる費用はコーヒー1杯分ほどだ。それだけの費用で政府は、新しい研究所を建築したり、新しい薬やワクチンを作ったりしなくても、安全に経済活動を再開させて、今も続いている新型コロナウイルスの感染を止めることが可能となる。

新型コロナウイルスの検査を受けている人々のほとんどは、症状があるか、感染者の濃厚接触者だ。しかし、会社や学校の再開を迫る声が高まる中で、組織は不都合な真実に対処せざるを得なくなっている。それは、症状を頼りに検査をすると、無症状の症例や症状が出る前の症例を見逃すことになり、すべての人を危険に晒してしまうということだ。

しかし、現在の選択肢は魅力的ではない。頻繁ではない検査（多くの提案では毎月がデフォルトになっている）や闇雲なスクリーニングでは、感染者が発症するまでの数週間にウイルスを拡散してしまう。また、費用はいまだに100ドルから200ドル以上と高額だ。

機械学習アルゴリズムに基づいた「プーリング検査」によって、こうした状況を変えることができる。プーリング検査では、多くの人の検体を混ぜて1つにする。そこからウイルスが検出されなければ、検体を混ぜ合わせた集団（「プール」と呼ぶ）内の人は誰も感染していないことになる。たった1度の検査でプール全体の人々の検査結果が判明するのだ。

しかし、落とし穴もある。プール内の誰か1人でも感染していれば、検査結果は陽性反応を示し、誰がウイルスに感染しているのかを特定するためにさらに検査が必要となる。

そのため、プールにひとまとめにして検査をする人々を選ぶ際に、陽性になる可能性の高い人たちを特定し、そうでない人たちと分けることが重要となる。そこで登場するのが機械学習だ。

米 …

こちらは会員限定の記事です。
メールアドレスの登録で続きを読めます。

有料会員にはメリットがいっぱい！

  1. 毎月120本以上更新されるオリジナル記事で、人工知能から遺伝子療法まで、先端テクノロジーの最新動向がわかる。

  2. オリジナル記事をテーマ別に再構成したPDFファイル「eムック」を毎月配信。
    重要テーマが押さえられる。

  3. 各分野のキーパーソンを招いたトークイベント、関連セミナーに優待価格でご招待。

GettyImages-1256395389-web-scaled.jpg
MzY2MTYxNg.jpeg

Welcome to the entrance to the world of acupuncture
Acupuncturist: a profession that will not disappear in the AI ​​era
With the development of AI (artificial intelligence), many existing occupations are expected to disappear. It is said that artificial intelligence has gradually entered the high-level intellectual professional fields such as finance and legal affairs, and has begun to serve people.

But what about acupuncture?

Acupuncture is a delicate and special technique that cannot be easily replaced by computers or machines. It can only be done manually. We often hear about their physical conditions from patients, consider the cause of the current symptoms, and select appropriate treatment points from 361 acupoints (acupoints) for acupuncture. This is a complex and delicate technique. It is impossible for robots. Even if it is possible, it will not be controlled by practitioners.

On the contrary, since it is surrounded by digital devices every day, do you think it needs the warmth, softness and safety of human hands? Acupuncture is a technology that is expected to become more valuable in the future AI era.

The WHO (World Health Organization) has also certified the therapeutic effects of acupuncture. Acupuncture therapy is good at treating symptoms that are difficult to improve by Western medicine, pain that has not been diagnosed as "discomfort" (discomfort) according to the examination results, and pain that cannot be completely cured. It is understandable that in recent years, the upsurge of Oriental medicine is coming, and it is said that this is a very stressful society. There is a growing demand for acupuncturists who can understand and treat patients' pain.

In addition, acupuncturists are long-term participants of men and women. This is not hard work, so even after aging, you can use the skills and intuition accumulated according to your physical strength to stand in the field of treatment. This is a lifelong job, you can improve your skills while you work, and cultivate opportunities for success while reaping rewards.
 

ACCELERATE OUTCOMES WITH A BACKBONE FOR ARTIFICIAL INTELLIGENCE.

makes machine learning operational by connecting models to the real-world decisions they inform. Often, AI/ML algorithms live in experimental vacuums. Foundry provides the end-to-end infrastructure an organization needs to apply AI/ML to real problems and real data:

  • A data foundation.  provides the data engineering capabilities an organization needs to deploy AI/ML models it can trust. Organizations use  to build a solid foundation of sufficient, quality data, then bring that data into daily operations.

  • Production deployment infrastructure. revolutionizes the way organizations build and deploy AI/ML by combining a data foundation with end-to-end algorithm deployment infrastructure. Data scientists and engineers can customize, deploy, assess, and compare across homegrown, open-source, and third-party algorithms. All models are tightly integrated with end-to-end platform capabilities, ranging from feature curation and health checks to model management to inference/serving to outcome monitoring.

  • Faster feedback loops. AI/ML models rarely work on a "set it and forget it" basis.  integrates the full model lifecycle with end user actions and feedback, and with operational decisions and outcomes. This enables operationally oriented model monitoring, management, understanding, selection, and adjustment. The result is more adaptable and ambitious AI/ML, faster.

UNITE THE ORGANIZATION AROUND A SHARED ENVIRONMENT FOR MACHINE LEARNING.

The most effective AI/ML encodes and supercharges an organization's unique expertise. That requires uniting an organization's data scientists, decision-makers, and everyday employees in an environment for collaborating on AI/ML-powered operations.  collaboration infrastructure drives AI/ML that brings the organization together:

  • A unifying ontology. ontology translates an organization's complex data landscape into human-readable concepts. built models reflect how an organization views the world and unite data scientists, engineers, analysts, executives, and operational end users around a common semantic layer.

  • Granular security controls. lets organizations define granular access control policies at the integration stage, then propagates those policies intelligently across the system. Organizations can promote collaboration confidently with granular data security and transparent data governance.

  • Model templates.model templates empower low-code and no-code AI/ML so even non-technical users can use AI to accelerate and enrich their workflows.

MAXIMIZE IMPACT AND MINIMIZE RISK WITH A PLATFORM DESIGNED FOR RESPONSIBLE ENGINEERING.

Our approach to AI/ML in  reflects our foundational belief in augmenting human intelligence, not replacing it. We believe AI/ML algorithms are most effective when they empower humans to ask complex questions, interpret answers, and act on results. At public and private sector organizations around the world, driven AI/ML is accelerating human decision-making by:

  • Surfacing new leads in dark web, weapons trafficking, financial fraud, and drug trafficking investigations so investigators can identify persons of interest more quickly

  • Aggregating and correlating biomedical research data to streamline drug discovery

  • Processing entity resolution suggestions so analysts can focus on making assessments rather than manually sorting and tagging data

  • Analyzing massive-scale sensor data so that engineers can make better aircraft maintenance decisions

  • Rapidly generating simulations while allowing operators to tweak scenario variables, leading to better-informed decisions optimized for different variables (e.g., safety rating and production quantity)

I see our approach to ethical machine learning as being grounded in an appreciation for both the promise and limitations of human-computer collaboration. The promise of AI is in augmenting and enhancing human intelligence, expertise and experience. Think helping a aircraft mechanic make better, more accurate and more timely repairs – not automating the mechanic out of the picture.

— Anthony Bak, Co-Lead of Palantir's Machine Learning team, in an interview with CTOVision

Protecting privacy and preserving civil liberties is fundamental to our mission.ships with technical capabilities that enable ethical engineering and ethical machine learning, including data protection features such as granular access controls, data provenance and lineage tracking, data retention and deletion management, and audit logging.  also enables industry-leading monitoring and validation for AI models. Its flexible, configurable tools let organizations evaluate model bias, in terms of both the data used to train the model and model outcomes.

We recognize that technology alone can't fully mitigate the risks of machine bias. Our dedicated Privacy and Civil Liberties team works with our engineers and our customers to approach building and deploying AI/ML thoughtfully.

Acupuncture significantly reduces AI-associated arthralgias

Publish date: December 8, 2017

By 

Roxanne Nelson

 

 

 

REPORTING FROM SABCS 2017

SAN ANTONIO – Acupuncture significantly reduced joint pain that was associated with the use of aromatase inhibitors (AIs) in women with early breast cancer, according to new findings reported at the San Antonio Breast Cancer Symposium.

The randomized, phase 3 SWOG S1200 clinical trial found that, compared with sham acupuncture and a control group receiving no therapy, women receiving acupuncture reported significantly lower scores on the Brief Pain Inventory–Short Form (BPI).

“We have shown consistently, with multiple measures assessing pain and stiffness, that true acupuncture generated better outcomes than either control group in a large multicenter trial,” said lead author Dawn L. Hershman, MD, leader of the Breast Cancer Program at the Herbert Irving Comprehensive Cancer Center at NewYork-Presbyterian/Columbia University Medical Center. “Acupuncture provides a nonpharmacologic option that can improve symptoms and possibly increase AI adherence and subsequent breast cancer outcomes.”

AIs can reduce both early breast cancer recurrence and mortality. Dr. Hershman noted that these agents are effective in the adjuvant setting and for prevention “but we know that it doesn’t work if you don’t take it. Noncompliance is a major problem among women taking hormonal therapy.”

Noncompliance is multifactorial and one of the main reasons women discontinue their therapy early is because of arthralgias or joint discomfort. “We were interested in a nonpharmacologic intervention, to assess whether or not we could control these symptoms.”

Dr. Hershman pointed out acupuncture provides a safe and effective alternative for patients reluctant to take a prescription medication that can result in other side effects. “Identification of nonopioid options for pain control is a public health priority,” she said.

Acupuncture is a popular nonpharmacologic modality and widely used for a number of indications. Several single-institution studies have suggested that it may be useful for controlling AI-associated arthralgias, while other studies have not demonstrated a benefit.

In this trial, the authors evaluated the efficacy of acupuncture, compared with sham acupuncture or waitlist control, in the treatment of AI associated arthralgia in a large population of patients. The study was conducted at 11 centers.

The cohort comprised 226 postmenopausal women diagnosed with early-stage, hormone receptor–positive breast cancer who were receiving treatment with AIs. The primary endpoint was the decline in joint pain as measured by BPI-SF at 6 weeks, and to assess the duration of the effect, the women were followed for an additional 12 weeks.

Within this group, 110 were randomized to true acupuncture; 59 to sham acupuncture, and 57 to waitlist control (no treatment). Patients receiving true or sham acupuncture had sessions three times a week for 6 weeks followed by one session per week for 6 more weeks. Pain status was reported at baseline, during treatment, and then afterwards, using a variety of measurement tools including the BPI-SF, which is a self-administered 14-item questionnaire that evaluates pain severity on a 0-10 scale, and the impact of pain on activities of daily living.

At 6 weeks, the true acupuncture treatment arm reported significantly lower BPI worst pain scores than those in the sham acupuncture and the waitlist control arms. The mean BPI worst pain for the true acupuncture arm was 0.92 points lower than the sham acupuncture arm (P = .01) and 0.96 points lower than the waitlist control arm (P = .01). The proportion of patients experiencing a large reduction in BPI worst pain (greater than 2) was significantly greater in the true acupuncture arm, compared with the other groups: 58% versus 33% percent and 31%, respectively. The differences continued to remain statistically significant at 24 weeks, even though the treatment only continued for 12 weeks.

Associated adverse effects were minimal with true and sham acupuncture and limited to grade 1 bruising.

The cost of the 12-week intervention was about $1,250 or $65-$75 a session. “We feel that there is now sufficient evidence to support insurance coverage of acupuncture of AI arthralgia.”

In a discussion of the paper, Dr. Anne Partridge, from the Dana Farber Cancer Center, noted that it is imperative to seek new ways to improve outcomes in breast cancer, and AIs are contributing to that. However, she echoed the concern that nonadherence to treatment is a “tremendous problem” and hampers the clinical effectiveness of AI therapy.

The rate of discontinuation during the first year of therapy is 20% within the first year and up to 40% of patients do not take them daily. Both early discontinuation and nonadherence contribute to mortality.

Based on these results from the largest randomized controlled trial looking at acupuncture in this setting, should physicians be recommending acupuncture to patients prescribed AI therapy?

“The short answer is, why not?” said Dr. Partridge, “And that we should be recommending it for some of our patients.”

However, there are a number of issues that need to be addressed, she added. The duration of treatment is not known, and the need for follow-up treatment or the frequency of it is not known. The generalizability of it is also unclear when looking at a larger population, and acupuncture is highly operator dependent.

“There are cost and access issues, and insurance right now offers very limited coverage,” she said.

Importantly, Dr. Partridge emphasized, “We know that it will help symptoms, but will it improve adherence to AI?”

It may improve adherence for some patients, but “side effects are only one factor,” she said. “Adherence behavior is complicated. We need to figure out how to optimize these therapies in our patients.”

This study was supported by the National Institutes of Health National Center for Complementary and Integrative Health and the Office of Research on Women’s Health, and grants from the NIH/National Cancer Institute Division of Cancer Prevention. Dr. Hershman declared no conflicts of interest. Dr. Partridge had no disclosures.

empty-wheelchair-parked-in-hospital-pict
digitwin.png

Illumina, Chinese Children's Hospital to Launch Newborn Sequencing Study

Apr 23, 2019

 

| 

staff reporter

 

NEW YORK (GenomeWeb) – Illumina and the Children's Hospital of Fudan University in China plan to launch a study of whole-genome sequencing in the hospital's neonatal intensive care unit to determine whether it can be used as a diagnostic for critically ill infants, Illumina said this week.

According to Illumina, the researchers plan to enroll 200 patients and compare the diagnostic rate of rapid WGS with genetic diagnostic methods such as microarray analysis and gene panel sequencing.

The researchers will also compare the time it takes to reach a diagnosis, impact on the patient's prognosis, and turnaround time.

Illumina will provide the sequencing reagents and the Children's Hospital of Fudan University will conduct the testing and data analysis and also be responsible for reporting results and providing genetic consultations with family members.

The hospital is also a sponsor of the Newborn Genome Project, which is creating a genome database for newborns in China in order to develop better methods for detecting genetic diseases in newborns and to establish standards for neonatal genetic diseases.

UK Study Details Liquid Biopsy's Ability to Guide Metastatic Breast Cancer Treatment

Dec 13, 2019

 

| 

staff reporter

 

NEW YORK – Researchers from the Institute of Cancer Research, London and the Royal Marsden NHS Foundation Trust have presented data demonstrating that blood-based genotyping assays can accurately detect specific biomarkers and potentially be used to guide targeted treatment of metastatic breast cancer. 

The researchers used Guardant Health's liquid biopsy assay and Bio-Rad Technologies' droplet digital PCR (ddPCR) testing to identify errors — including mutations in the HER2, ESR1 and AKT1 genes — in circulating tumor DNA (ctDNA) that had been shed into the patients' bloodstream. 

During a presentation this week at the San Antonio Breast Cancer Symposium (SABCS), Nicholas Turner, molecular oncology professor at the Institute of Cancer Research, explained that his team wanted to solve the issue of genotyping breast cancer tumors without having to perform multiple biopsies. Turner highlighted the need for prospective studies to assess the accuracy of ctDNA testing in routine practice and the potential of these tools to guide targeted therapy without requiring solid tissue testing. 

With funding from Stand Up To Cancer (a fundraising campaign from Cancer Research UK and Channel 4) Turner's team therefore launched an ongoing prospective study, called "plasmaMATCH," to examine the clinical utility of ctDNA for metastatic breast cancer detection. The group enrolled a total of 1,044 patients with advanced breast cancer who had either progressed on prior drug therapy or relapsed within a year after adjuvant chemotherapy. 

The researchers analyzed ctDNA in blood samples from patients using the Guardant360 sequencing-based assay and Bio-Rad's ddPCR as an orthogonal method. Patients with identified mutations could also enroll in a treatment arm of the study that matched their mutation. 

Turner and his team split the patient population into three major cohorts depending on specific genetic errors found in ctDNA: ESR1 mutations, HER2 mutations, and AKT-1 mutations in estrogen-receptor-positive breast cancer. The researchers also added a fourth cohort that had AKT-1 and PTEN mutations based on both ctDNA and tumor sequencing results, as well as a fifth group with triple-negative breast cancer without mutations. 

While Turner's team's initially aimed to measure the response rate of targeted therapies matched to mutations in ctDNA without tissue testing, the researchers also monitored the frequency of mutations, accuracy of testing, proportion of patients entering a treatment cohort, and the activity in clonality-dominant versus subclonal mutations. Turner's team discovered that 784 patients who had cfDNA testing done with both Guardant360 and ddPCR had "very strong agreement" of between 96 and 99 percent. Comparing the ctDNA to the patients' tumor samples to validate the plasmaMATCH findings, the researchers found that the liquid biopsy assay had an overall sensitivity of 93 percent. 

Guardant360 also identified several more targetable alterations than hotspot testing, including over 35 percent more PIK3CA mutations that can be targeted by US Food and Drug Administration-approved therapies. The assay also detected significantly more ESR1 mutations and found previously undetected microsatellite instability and ERBB2 mutations.   

The researchers provided 142 patients that had specific identified mutations experimental targeted therapies as part of the study, and they plan to test the treatments that showed initial promise in larger clinical trials. 

The researchers now believe that they can use blood testing to identify rare subtypes of breast cancer, in addition to potentially replacing more invasive methods of breast cancer detection. 

"We have now confirmed that blood tests can quickly give us a bigger picture of the mutations [that] are present within multiple tumors throughout the study," Turner said in a statement.  

"We show that circulating tumor DNA testing offers a simple, efficient, and fast method of tumor genotyping," Turner also noted during the presentation at SABCS. "The patients with mutations identified in their ctDNA have efficacy with matching target[ed] therapies." 

The researchers believe the blood-based assays are now reliable enough to be applied routinely by clinicians after receiving regulatory approval.

"As the number of treatment-relevant genomic alterations in metastatic breast cancer continues to grow, it is critical that oncologists have a simple and reliable way to comprehensively test for this information about their patients," Guardant Health Global Chief Medical Officer Rick Lanman said in a statement. "Guardant360 detected changes in the genome picture over time … and also identifies targetable but uncommon genomic biomarkers."

1c8b9b577cc41266ebab5e65ff8f3de8.jpg
090320_COVID_Test_0111-1024x682.jpg
05b50bc878af8f403a01848035fa3a86.jpg
7d54e3018152c4d0b389e55596039404.jpg
4deac0802d4054a8a85b3f3d256911aa.jpg
hocit999 - コピー_ページ_16.jpg
hocit999_ページ_6.jpg
bottom of page