🧠 EJNMMI Research 每週科學新知（2025/08/19）
1. 利用 TriDFusion 與 nnU‑Net V2 自動分割 PET/CT 中棕色脂肪組織
Brown adipose tissue machine learning nnU‑Net V2 network using TriDFusion (3DF)
Authors: Daniel Lafontaine¹, Stephanie Chahwan², Gustavo Barraza², ...
Affiliations: Memorial Sloan Kettering Cancer Center, Rockefeller University, USA
發表日期：2025‑08‑13
本研究提出一種結合 TriDFusion 與 nnU‑Net V2 深度學習模型的自動分割工具，可精準辨識 PET/CT 影像中的棕色脂肪組織。該方法可提升影像解剖清晰度與判讀效率，對代謝性疾病與癌症影像研究具有實用價值。
This study introduces an automated segmentation tool that combines TriDFusion and nnU‑Net V2 deep learning to accurately identify brown adipose tissue (BAT) in PET/CT images. It enhances anatomical clarity and diagnostic efficiency, which is valuable for metabolic and cancer imaging studies.
🔗 閱讀原文 / Full text
2. 利用⁶⁸Ga‑Exendin 導引手術定位胰島素產生病灶：先天性高胰島素症個案研究
[⁶⁸Ga] labelled Exendin for radioguided surgery of intrapancreatic insulin producing lesions in patients with congenital hyperinsulinism
Authors: Peter Kühnen, Sonal Prasad, Karin Rothe, ...
發表日期：2025‑08‑12
研究評估⁶⁸Ga‑標記 Exendin 在導引手術中定位胰臟病灶的可行性，可提升術中定位精準度並減少不必要切除，對先天性高胰島素症患者具臨床潛力。
This study evaluates the feasibility of using 68Ga-labeled Exendin for radioguided surgery targeting intrapancreatic lesions in congenital hyperinsulinism, aiming to improve surgical precision and reduce unnecessary resection.
🔗 閱讀原文 / Full text
3. 運用深度學習預測 NDUFS1 靶向放射性藥物的蛋白結合潛力
Deep learning‑based radiolabelled compound‑protein interaction prediction for NDUFS1‑targeting radiopharmaceutical discovery
Authors: Muath Almaslamani, Jingyu Yang, ...
Affiliations: Korea Institute of Radiological and Medical Sciences & University of Science & Tech., Republic of Korea
發表日期：2025‑08‑12
本文運用深度學習預測放射性示蹤劑與粒線體蛋白 NDUFS1 的相互作用，可加速新型放射藥物發現與研發流程。
This article employs deep learning to predict interactions between radiolabeled compounds and the mitochondrial protein NDUFS1, potentially accelerating the discovery of novel radiopharmaceuticals.
🔗 閱讀原文 / Full text
4. 評估第2型糖尿病患者之胰島素敏感性影像與測量重現性
Tissue‑specific and whole‑body insulin sensitivity by integrated imaging and hyperinsulinemic euglycemic clamp: A repeatability study in people with T2DM and overweight/obesity
Authors: Iina Laitinen, Helena Litorp, Folke Sjöberg, ...
發表日期：2025‑08‑07
此研究透過 PET 和高胰島素棉蘭普試驗評估第2型糖尿病與超重者的全身與組織層級胰島素敏感性，並驗證其測量穩定性。
This study assesses whole-body and tissue-specific insulin sensitivity in individuals with T2DM and overweight/obesity using PET combined with hyperinsulinemic euglycemic clamp, and validates measurement reproducibility.
🔗 閱讀原文 / Full text
5. 探討不同 linkers 對 GRPR PET 示蹤劑的腫瘤攝取與背景比
A preclinical study on the influence of linkers in [⁶⁸Ga]Ga‑NOTA‑X‑RM26 radiotracers for PET imaging of GRPR expression
Authors: Fan Yu, Xiaoran Li, Yue Zhang, ...
發表日期：2025‑08‑06
本研究比較不同 linker 對 GRPR 靶向 PET 示蹤劑的腫瘤攝取與背景比值影響，對提升前列腺與乳癌成像質量具參考價值。
This preclinical study compares the effects of various linkers on tumor-to-background uptake of GRPR-targeting PET tracer [68Ga]Ga-NOTA-X-RM26, informing strategies to improve imaging quality in prostate and breast cancer.
🔗 閱讀原文 / Full text