研究者詳細

顔写真

ハリガイ アヤカ
針谷 綾花
Ayaka Harigai
所属
大学院医学系研究科 医科学専攻 内科病態学講座(放射線診断学分野)
職名
助教
学位
  • 博士(医学)(東北大学)

  • M.A.(University of California Los Angeles)

e-Rad 研究者番号
80977265

受賞 5

  1. 2026年度新津記念賞 最優秀賞

    2026年9月 公益財団法人 艮陵医学振興会 治療抵抗性高血圧に対するCTガイド下腎デナベーションの実現に向けた腎動脈周囲局所加温の基礎的検討

  2. 七星賞 最優秀賞

    2025年5月 東北大学大学院医学系研究科 任意の方向からの低侵襲な経皮的アプローチを可能にする高汎用性ナビゲーションシステムの開発

  3. 東北大学病院辛酉会論文奨励賞

    2026年3月 東北大学病院辛酉会 癒着胎盤スペクトラムの診断のための半自動深層学習モデル(COMPAS):放射線科医の読影との比較

  4. 東北大学病院辛酉会論文奨励賞

    2025年3月 東北大学病院辛酉会

  5. Young Researchers Presentation Award

    2024年12月 Ressearch Center for Biomedical Engineering Visualization of Needle Deflection in Out-of-Plane CT-Guided Punctures: A Preliminary Study

論文 9

  1. Response accuracy of GPT-4 across languages: insights from an expert-level diagnostic radiology examination in Japan. 査読有り

    Ayaka Harigai, Yoshitaka Toyama, Mitsutoshi Nagano, Mirei Abe, Masahiro Kawabata, Li Li, Jin Yamamura, Kei Takase

    Japanese journal of radiology 43 (2) 319-329 2025年2月

    DOI: 10.1007/s11604-024-01673-6  

    詳細を見る 詳細を閉じる

    PURPOSE: This study aims to investigate the effects of language selection and translation quality on Generative Pre-trained Transformer-4 (GPT-4)'s response accuracy to expert-level diagnostic radiology questions. MATERIALS AND METHODS: We analyzed 146 diagnostic radiology questions from the Japan Radiology Board Examination (2020-2022), with consensus answers provided by two board-certified radiologists. The questions, originally in Japanese, were translated into English by GPT-4 and DeepL and into German and Chinese by GPT-4. Responses were generated by GPT-4 five times per question set per language. Response accuracy was compared between languages using one-way ANOVA with Bonferroni correction or the Mann-Whitney U test. Scores on selected English questions translated by a professional service and GPT-4 were also compared. The impact of translation quality on GPT-4's performance was assessed by linear regression analysis. RESULTS: The median scores (interquartile range) for the 146 questions were 70 (68-72) (Japanese), 89 (84.5-95.5) (GPT-4 English), 64 (55.5-67) (Chinese), and 56 (46.5-67.5) (German). Significant differences were found between Japanese and English (p = 0.002) and between Japanese and German (p = 0.022). The counts of correct responses across five attempts for each question were significantly associated with the quality of translation into English (GPT-4, DeepL) and German (GPT-4). In a subset of 31 questions where English translations yielded fewer correct responses than Japanese originals, professionally translated questions yielded better scores than those translated by GPT-4 (13 versus 8 points, p = 0.0079). CONCLUSION: GPT-4 exhibits higher accuracy when responding to English-translated questions compared to original Japanese questions, a trend not observed with German or Chinese translations. Accuracy improves with higher-quality English translations, underscoring the importance of high-quality translations in improving GPT-4's response accuracy to diagnostic radiology questions in non-English languages and aiding non-native English speakers in obtaining accurate answers from large language models.

  2. Reply to "The R.E.N.A.L. nephrometry scoring from CT reports with ChatGPT: example with proofs".

    Ayaka Harigai, Yoshitaka Toyama, Kei Takase

    Japanese journal of radiology 42 (8) 932-933 2024年8月

    DOI: 10.1007/s11604-024-01599-z  

  3. Reply to "Assessing the diagnostic performance of large language models with European Diploma in Musculoskeletal Radiology (EDiMSK) examination sample questions".

    Ayaka Harigai, Yoshitaka Toyama, Kei Takase

    Japanese journal of radiology 42 (6) 675-676 2024年6月

    DOI: 10.1007/s11604-024-01556-w  

  4. Performance evaluation of ChatGPT, GPT-4, and Bard on the official board examination of the Japan Radiology Society. 査読有り

    Yoshitaka Toyama, Ayaka Harigai, Mirei Abe, Mitsutoshi Nagano, Masahiro Kawabata, Yasuhiro Seki, Kei Takase

    Japanese journal of radiology 42 (2) 201-207 2024年2月

    DOI: 10.1007/s11604-023-01491-2  

    詳細を見る 詳細を閉じる

    PURPOSE: Herein, we assessed the accuracy of large language models (LLMs) in generating responses to questions in clinical radiology practice. We compared the performance of ChatGPT, GPT-4, and Google Bard using questions from the Japan Radiology Board Examination (JRBE). MATERIALS AND METHODS: In total, 103 questions from the JRBE 2022 were used with permission from the Japan Radiological Society. These questions were categorized by pattern, required level of thinking, and topic. McNemar's test was used to compare the proportion of correct responses between the LLMs. Fisher's exact test was used to assess the performance of GPT-4 for each topic category. RESULTS: ChatGPT, GPT-4, and Google Bard correctly answered 40.8% (42 of 103), 65.0% (67 of 103), and 38.8% (40 of 103) of the questions, respectively. GPT-4 significantly outperformed ChatGPT by 24.2% (p < 0.001) and Google Bard by 26.2% (p < 0.001). In the categorical analysis by level of thinking, GPT-4 correctly answered 79.7% of the lower-order questions, which was significantly higher than ChatGPT or Google Bard (p < 0.001). The categorical analysis by question pattern revealed GPT-4's superiority over ChatGPT (67.4% vs. 46.5%, p = 0.004) and Google Bard (39.5%, p < 0.001) in the single-answer questions. The categorical analysis by topic revealed that GPT-4 outperformed ChatGPT (40%, p = 0.013) and Google Bard (26.7%, p = 0.004). No significant differences were observed between the LLMs in the categories not mentioned above. The performance of GPT-4 was significantly better in nuclear medicine (93.3%) than in diagnostic radiology (55.8%; p < 0.001). GPT-4 also performed better on lower-order questions than on higher-order questions (79.7% vs. 45.5%, p < 0.001). CONCLUSION: ChatGPTplus based on GPT-4 scored 65% when answering Japanese questions from the JRBE, outperforming ChatGPT and Google Bard. This highlights the potential of using LLMs to address advanced clinical questions in the field of radiology in Japan.

  5. Letter to the editor response to "ChatGPT, GPT-4, and bard and official board examination: comment".

    Ayaka Harigai, Yoshitaka Toyama, Kei Takase

    Japanese journal of radiology 42 (2) 214-215 2024年2月

    DOI: 10.1007/s11604-023-01515-x  

  6. Type 2 Endoleak Detection After Endovascular Aneurysm Repair: Can Low-Flow Endoleaks Also Be Visualized in Silicon-Photomultiplier-Based Positron Emission Tomography/Computed Tomography Scans? 国際誌 査読有り

    Ayaka Harigai, Yoshitaka Toyama, Michihisa Umetsu, Kentaro Takanami, Kei Takase

    Journal of endovascular therapy : an official journal of the International Society of Endovascular Specialists 15266028231185237-15266028231185237 2023年7月2日

    DOI: 10.1177/15266028231185237  

    詳細を見る 詳細を閉じる

    PURPOSE: To present a novel clinical application of silicon-photomultiplier-based positron emission tomography (SiPM-based PET)/computed tomography (CT), detecting a type II endoleak 5 years after endovascular aneurysm repair (EVAR). TECHNIQUE: SiPM-based PET/CT scans with a standard whole-body protocol were performed for a 73-year-old man with a past medical history of abdominal aortic aneurysms treated with EVAR and currently under investigation of his duodenal papillary carcinoma. The PET/CT demonstrated 18F-fluorodeoxyglucose (FDG) accumulation outside the stent graft in the native sac of the aneurysm. The site of accumulation corresponded to that of the contrast enhancement depicted in the CT angiography taken 1 month earlier. Another CT scan performed 3 months later revealed enlargement of the aneurysm. CONCLUSION: SiPM-based PET/CT, with its superior sensitivity and spatial resolution over conventional PET/CT, can detect type II low-flow endoleaks. CLINICAL IMPACT: Abnormal intra-aneurysmal FDG activity incidentally detected on SiPM-based PET/CT is worthy of attention because it may be indicative of endoleaks. Additional imaging using different modalities should be considered so that the patient would not miss the additional treatment opportunity upon observing sac enlargement. For patients with contraindications for iodine CT contrast media, SiPM-based PET/CT would serve as a suitable alternative.

  7. Favorable response after radiation therapy for intraductal papillary mucinous neoplasms manifesting as acute recurrent pancreatitis: A case report. 国際誌 査読有り

    Ayaka Harigai, Kiyoshi Kume, Noriyoshi Takahashi, So Omata, Rei Umezawa, Keiichi Jingu, Atsushi Masamune

    World journal of clinical cases 10 (30) 11116-11121 2022年10月26日

    DOI: 10.12998/wjcc.v10.i30.11116  

    詳細を見る 詳細を閉じる

    BACKGROUND: There has been an increasing number of elderly patients with intraductal papillary mucinous neoplasm (IPMN), who are surgically intolerant and require less invasive treatment options, which are limited. In the present study, we report a case of IPMN presenting with acute recurrent pancreatitis (ARP), in which radiation therapy effectively prevented further attacks of ARP and reduced tumor volume. CASE SUMMARY: An 83-year-old man was referred to our hospital with an asymptomatic incidental pancreatic cyst. Endoscopic ultrasound imaging and magnetic resonance cholangiopancreatography revealed a multiloculated tumor in the head of the pancreas, with dilated pancreatic ducts and mural nodules. The patient was diagnosed with mixed-type IPMN, and five years later, he developed ARP. Several endoscopic pancreatic ductal balloon dilatations failed to prevent further ARP attacks. Surgery was considered clinically inappropriate because of his old age and comorbidities. He was referred to our department for radiation therapy targeted at those lesions causing intraductal hypertension and radiation was administered at a dose of 50 Gy. An magnetic resonance imaging scan taken ten weeks after treatment revealed a decrease in tumor size and improvement of pancreatic duct dilatation. Fourteen months later, he remains symptom-free from ARP. CONCLUSION: This case highlights the important role of radiation therapy in mitigating the signs and symptoms of ARP in patients with inoperable IPMN.

  8. The benefit of concurrent chemotherapy with radiotherapy for esophageal cancer is limited in Asian patients aged 80 years or older: a SEER database analysis. 査読有り

    Keiichi Jingu, Noriyoshi Takahashi, Rei Umezawa, Takaya Yamamoto, Kazuya Takeda, Yu Suzuki, Keita Kishida, So Omata, Yuta Sato, Hinako Harada, Ayaka Harigai

    Esophagus : official journal of the Japan Esophageal Society 19 (4) 653-659 2022年10月

    DOI: 10.1007/s10388-022-00938-1  

    詳細を見る 詳細を閉じる

    PURPOSE: To evaluate the benefit of concurrent chemotherapy with radiotherapy (RT) for esophageal cancer in Asian patients aged ≥ 80 years using the Surveillance, Epidemiology, and End Results (SEER) database. MATERIALS AND METHODS: Among more than 7000 patients with squamous cell carcinoma or adenocarcinoma who were treated by RT without surgery for esophageal cancer in the SEER database, 2047 patients aged ≥ 80 years were analyzed. Patients who received chemoradiotherapy (CRT group) and patients who received RT alone (RT alone group) were matched with a propensity score. RESULTS: The median observation period for survivors was 57 months. The 3-year and 5-year overall survival rates in all patients were 15.2% and 8.5%, respectively. The 3-year and 5-year cause-specific survival rates in all patients were 20.8% and 14.5%, respectively. After propensity score matching, the overall survival rate in the CRT group was significantly higher than that in the RT alone group (5-year overall survival rates: 11.9% and 3.2%, respectively, p < 0.001). In 108 Asian or Pacific Islander patients, there was no significant difference (5-year overall survival rates: 13.5% and 0%, respectively, p = 0.291), although the overall survival rate in the CRT group was significantly higher than that in the RT alone group in any other race. CONCLUSIONS: It is controversial whether CRT is beneficial for Asian or Pacific Islander patients aged 80 years or older with esophageal cancer based on Analysis of data in SEER database.

  9. The prognostic value of 18F-FDG PET/CT taken immediately after completion of radiotherapy for lung cancer treated with concurrent chemoradiotherapy: A pilot study. 国際誌 査読有り

    A Harigai, A I Saito, T Inoue, M Suzuki, Y Namba, Y Suzuki, F Makino, O Nagashima, S Sasaki, K Sasai

    Cancer radiotherapie : journal de la Societe francaise de radiotherapie oncologique 26 (5) 711-716 2022年9月

    DOI: 10.1016/j.canrad.2022.01.006  

    詳細を見る 詳細を閉じる

    PURPOSE: The prognostic value of F-18 fluorodeoxyglucose positron emission tomography/computed tomography (PET/CT) taken immediately after completion of radiotherapy in lung cancer patients is not well known. The purpose of this study is to assess the prognostic value of PET/CT taken immediately after completion of radiotherapy in lung cancer patients. MATERIALS AND METHODS: Patients with primary lung cancer planned to undergo concurrent chemoradiotherapy were enrolled. Patients underwent PET/CT scans at 3 time points: before radiotherapy, within 24hours of completing radiotherapy (im-PET/CT), and 2-9 months after radiotherapy (post-PET/CT). Maximum standardized uptake value (SUVmax) was obtained. A post-PET/CT-SUVmax cut-off of 2.5 was determined as radiotherapy success. RESULTS: Nineteen patients were enrolled. im-PET/CT-SUVmax for patients in the high post-PET/CT-SUVmax group was significantly higher than that of the low group (P=0.004). Receiver operator curve analysis indicated that im-PET/CT-SUVmax of 4.35 was an optimal cut-off value to discriminate between the two groups. Multivariable analysis showed that a high im-PET/CT-SUVmax was significantly associated with a high post-PET/CT-SUVmax (P=0.003). CONCLUSION: PET/CT-SUVmax taken immediately following radiotherapy was associated with that evaluated 2-9 months after radiotherapy.

︎全件表示 ︎最初の5件までを表示

講演・口頭発表等 14

  1. 若手医師が産学連携で医療機器を形にするまで:放射線診断科からの実践共有 招待有り

    針谷 綾花, 中川 敦寛

    STROKE2026 2026年3月12日

  2. 「未来に向けたMOVE & SHAKE.」最前線を変えていく 招待有り

    針谷綾花

    東北大学国際卓越研究大学シンポジウム 2025年2月6日

  3. Optimizing Interventional Radiology with DX sensing: a Clinical Perspective 招待有り

    Ayaka Harigai, Hiroki Kase, Hiroshi Ishihata, Kenji Kikuchi, Kei Takase, Toru Aoki

    The 31st International Display Workshops (IDW2024) 2024年12月6日

  4. 急がば回って二刀流 and more! 招待有り

    針谷綾花, 高瀬圭

    第52回日本IVR学会総会 2023年5月18日

  5. 任意の方向からの低侵襲な経皮的アプローチを可能にする高汎用性ナビゲーションシステムの開発 招待有り

    針谷 綾花

    国立研究開発法人日本医療研究開発機構 令和6年度 医療機器・ヘルスケアプロジェクト 成果報告会 2024年7月2日

  6. CTガイド下面外穿刺を支援するナビゲーションシステム開発

    針谷 綾花, 山本新之助, 齋藤一創, 菊地謙次, 田中元, 岡野華奈, 平野仁也, 外山由貴, 山口啓太, 大森悠斗, 白鳥和敏, 茅野信吾, 石幡浩志, 大田英揮, 高瀬圭

    日本IVR学会総会 2026年5月23日

  7. Development of a Virtual CT-Based Puncture Simulation System: Concept and Early Feasibility

    Ayaka Harigai, Hiroki Kase, Hiroshi Ishihata, Yuto Omori, Kei Takase, Toru Aoki

    The 10th International Symposium on Biomedical Engineering (ISBE2025) 2025年10月30日

  8. 仮想 CT ベース穿刺シミュレーションシステムの開発

    針谷綾花, 加瀬裕貴, 山本 新之助, 石幡浩志, 高瀬 圭, 青木 徹

    第10回日本穿刺ドレナージ研究会 2025年7月19日

  9. 人体内へのCTガイド下針穿刺の安全性向上を目指した 穿刺針のたわみ解析 招待有り

    針谷綾花, 山本新之助, 齊藤一創, 石幡浩志, 大森悠人, 菊地謙次, 高瀬圭, 加瀬裕貴, 青木徹

    2024年度生体医歯工学共同研究拠点成果報告会 2025年3月3日

  10. デザイン思考を取り入れた東北大学らしいAI研究の紹介

    針谷綾花

    第3回 Clinical AI アニュアルシンポジウム~産学連携と人材育成の継続に向けて~ 2023年2月2日

  11. 心臓移植後のFDG-PETで 体外型人工心臓の刺入部痕に高集積が見られた一例

    針谷綾花, 外山由貴, 阿部未玲, 益田淳朗, 前川由依, 齊藤美穂子, 髙浪健太郎, 高瀬圭

    第147回日本医学放射線学会北日本地方会・第92回日本核医学会北日本地方会 2022年11月15日

  12. 深層学習を用いた穿刺手技支援のための針位置検知システムの基礎的検討

    田中元, 針谷綾花, 山本新之助, 石幡浩志, 高瀬圭, 加瀬裕貴, 青木徹

    2025年度生体医歯工学共同研究拠点成果報告会 2026年3月3日

  13. 教育用シミュレーションシステムを用いたCTガイド下穿刺トレーニング

    針谷 綾花, 加瀬 裕貴, 石幡 浩志, 岡野華奈, 田中元, 清水弘明, 山本新之助, 高瀬圭, 青木徹

    第38回関東IVR研究会 2026年1月24日

  14. Visualization of Needle Deflection in Out-of-Plane CT-Guided Punctures: A Preliminary Study

    Ayaka Harigai, Shinnosuke Yamamoto, Hiroki Kase, Yuto Omori, Kenji Kikuchi, Kei Takase, Toru Aoki

    The 9th International Symposium on Biomedical Engineering (ISBE2024) 2024年12月3日

︎全件表示 ︎最初の5件までを表示

共同研究・競争的資金等の研究課題 6

  1. 針穿刺経路設定のための腹腔臓器等の呼吸性変動の仮想現実による可視化

    針谷 綾花

    提供機関:Japan Society for the Promotion of Science

    制度名:Grants-in-Aid for Scientific Research

    研究種目:Grant-in-Aid for Early-Career Scientists

    研究機関:Tohoku University

    2025年4月1日 ~ 2027年3月31日

  2. 任意の方向からの低侵襲な経皮的アプローチを可能にする高汎用性ナビ ゲーションシステムの開発(開発実践)

    高瀬圭, 石幡浩志, 菊地謙次, 青木徹, 加瀬裕貴, 都木克之, 谷口彰

    提供機関:Japan Agency for Medical Research and Development

    研究機関:Tohoku University

    2024年4月 ~ 2027年3月

  3. 治療抵抗性高血圧に対するCTガイド下腎デナベーションの実現に向けた腎動脈周囲局所加温の基礎的検討

    2026年9月 ~

  4. 診断・治療針穿刺の安全性向上を目的とした穿刺経路分析・提示システム開発(2025)

    針谷 綾花, 田中元, 高瀬圭, 加瀬 裕貴, 青木徹

    2025年4月 ~ 2026年3月

  5. 診断・治療針穿刺の安全性向上を目的とした穿刺経路分析・提示システム開発(2024)

    針谷綾花, 高瀬圭, 加瀬 裕貴, 青木徹

    2024年4月 ~ 2025年3月

  6. 任意の方向からの低侵襲な経皮的アプローチを可能にする高汎用性ナビゲーションシステムの開発(チャレンジ)

    提供機関:Japan Agency for Medical Research and Development

    研究機関:Tohoku University

    2023年5月 ~ 2024年3月

︎全件表示 ︎最初の5件までを表示

社会貢献活動 1

  1. 東北大学オープンキャンパス 放射線診断科特別企画

    東北大学オープンキャンパス2025

    2025年7月30日 ~ 2025年7月31日