當前全球醫療科技領域迎來重大突破。由美國約翰霍普金斯大學領銜、聯合史丹佛與哥倫比亞大學團隊所研發的手術機器人「SRT-H」（Surgical Robot Transformer-Hierarchy），成功在無人介入的情況下，100% 自主完成八例膽囊切除術的核心步驟。這項成就不僅登上國際頂尖期刊《Science Robotics》，也被廣泛視為醫療AI發展史上的里程碑。

這套機器人系統與過往僅能執行單一動作的手術輔助機械截然不同。SRT-H具備多層級AI結構，結合大型語言模型與影像辨識技術，可模擬醫師的決策思維，針對手術過程中多變的情況進行動態判斷。此次實驗雖在豬的離體器官中進行，但手術流程已與人體手術高度接近，包括辨認膽囊與相關血管結構、使用手術器械定位與切除，整個過程皆由AI自主完成，未有任何人類操作介入。研究顯示，SRT-H 的手術路徑更短、動作更流暢，切割過程穩定性甚至超過人類醫師。

這項突破代表的不僅是機械精度的提升，更關鍵的是AI已具備對手術語境與操作步驟的「理解能力」。舉例來說，若術中遇到視野遮蔽或結構異常，SRT-H不但能自主判斷問題，更能即時調整策略，顯示其非死板程式，而是真正進入「智能手術時代」。

不過，業界對此仍持審慎態度。英國皇家外科學會與NHS等專業機構指出，這項技術雖具革命性潛力，但目前仍處於實驗階段，距離進入臨床應用仍有不少挑戰。真正的體內手術牽涉到血液流動、器官移動、患者反應等複雜變數，這些都非AI模擬環境所能完全涵蓋。因此，SRT-H雖展現令人驚豔的技術成熟度，但仍需經過長時間的驗證與監管程序。

在市場層面，該技術的發布迅速引發醫療產業與科技投資界高度關注。分析師普遍認為，AI自主手術將可能徹底改寫全球外科手術格局，特別是在人力資源緊縮與醫療資源分配不均的地區。長遠而言，AI外科助手有望讓一名醫師監控多台自主機器人同時執行標準手術，大幅提升效率與安全性，並降低醫療成本與錯誤率。

然而，也有醫師提醒，手術從來不只是技術的堆疊，更是醫病互動與臨場判斷的藝術。AI若要進一步走入臨床，不只是技術成熟，還需制度、倫理、責任劃分等多方面的同步發展。特別是未來若發生手術風險，責任歸屬與患者知情同意的框架都需嚴格建立。

總的來說，SRT-H的出現，宣示人類正從「人機協作手術」走向「機器人自主手術」的新階段。雖然距離大規模臨床應用仍有距離，但這一研究成果無疑為未來醫療 AI開啟嶄新篇章。科技正在改變醫療現場的樣貌，而我們正站在這個歷史性轉捩點的門檻上。

A groundbreaking development has emerged in the field of global medical technology. A surgical robot named SRT-H (Surgical Robot Transformer-Hierarchy), developed by a research team led by Johns Hopkins University in collaboration with Stanford University and Columbia University, has successfully completed the core steps of eight gallbladder removal surgeries (cholecystectomies) entirely autonomously, without any human intervention. This achievement not only earned a feature in the prestigious journal Science Robotics, but is also widely regarded as a milestone in the history of medical AI.

Unlike previous surgical-assist robots that could only perform basic, predefined actions, SRT-H features a multi-layered AI architecture. It integrates large language models with advanced visual recognition, enabling it to simulate a surgeon’s decision-making process and dynamically respond to complex and variable conditions during surgery. Although the recent experiments were conducted on ex vivo pig organs, the procedures closely mirrored real human surgeries—identifying the gallbladder and related vascular structures, manipulating surgical tools, and executing precise excisions—all carried out independently by the AI, with zero human input. Notably, the study found that SRT-H’s surgical paths were shorter, its movements smoother, and its cutting precision more stable than that of human surgeons.

This breakthrough represents more than just enhanced robotic precision. The most critical advancement lies in the AI’s “understanding” of surgical context and procedural steps. For instance, if the field of vision is obstructed or an anatomical anomaly is detected, SRT-H is capable of autonomously identifying the issue and adjusting its approach in real-time. This reveals that it is not a rigid, pre-programmed tool—but a true step into the era of intelligent surgery.

However, the medical community remains cautiously optimistic. Organizations such as the Royal College of Surgeons (UK) and the UK National Health Service (NHS) emphasize that while the technology is revolutionary, it is still in the experimental phase. Transitioning to clinical use involves considerable hurdles. Real surgeries inside the human body involve complexities such as blood flow, organ movement, and patient reactions—factors that AI in simulated environments cannot yet fully replicate. As such, while SRT-H demonstrates impressive technical maturity, it must undergo extensive validation and regulatory review before clinical deployment.

On the market front, the unveiling of this technology has sparked significant attention from both the medical industry and tech investment sectors. Analysts believe that autonomous AI surgery could fundamentally reshape global surgical practice, especially in regions facing shortages of medical professionals and unequal healthcare resource distribution. In the long term, AI surgical assistants could allow a single physician to monitor multiple autonomous robots performing standardized procedures simultaneously, thereby boosting efficiency and safety while reducing costs and error rates.

Yet, some surgeons caution that surgery is not merely a technical procedure—it is also an art rooted in human interaction and real-time judgment. For AI to integrate into clinical practice, more than just technological maturity is needed. It will require synchronized developments in institutional frameworks, ethics, and liability protocols. In particular, when surgical risks arise in the future, clear accountability and robust informed consent systems must be in place.

In summary, the emergence of SRT-H marks a transition from “human-robot collaborative surgery” to a new era of fully autonomous robotic operations. Although it may still be some time before widespread clinical adoption, this achievement undoubtedly opens a new chapter for medical AI. Technology is transforming the face of modern healthcare, and we now stand at the threshold of this historic turning point.