How surgery and mental-health treatment can benefit
手术和心理治疗如何能受益

渐入佳境 Getting better-书迷号 shumihao.com

A  SOLDIER WATCHES a car approaching a check-point on a hot, dusty road. As the vehicle slows to a stop in front of him, he asks the driver to get out and show his identification. Seconds later, the rattle of gunfire pierces the air, followed by a bang and an intense, searing flash. Knocked to the ground and scrambling to safety, the soldier turns to see a flaming wreck where the car had been just moments before.
在炙热而尘土飞扬的道路上,士兵看着一辆车驶近检查站。当车在他面前减速停下时,他要求驾驶员下车并出示身份证件。几秒钟后,一阵枪声划破长空,接着是一声巨响和强烈而灼热的闪光。士兵被掀翻在地。他奋力爬向安全处,转身看到那车已变成一团燃烧的残骸。

The scene pauses. A voice in the soldier’s ear says: “Let’s rewind the simulation to the seconds just before the explosion—describe exactly what happened.” The voice is a therapist, speaking to a veteran who is placed in a virtual environment. The simulation they are watching has been modelled on the veteran’s own experiences in a war zone, events that have led him to develop post-traumatic stress disorder (PTSD).
场景暂停。士兵耳边的声音说:“让我们把模拟倒退到爆炸发生前的几秒钟,准确地描述发生了什么。” 这是治疗师的声音,正在与被置入虚拟环境中的退伍军人交谈。他们正在观看的模拟场景是仿照这位退伍军人自己在战区的经历建立的,这些事件使他患上了创伤后应激障碍(PTSD)。

This is the Bravemind system, developed in 2005 by Albert “Skip” Rizzo and Arno Hartholt, experts in medical virtual reality at the University of Southern California, to treat soldiers returning home from the wars in Iraq and Afghanistan. Immersed in a virtual environment that mimics their traumatic experiences, veterans narrate the scene to a therapist, who can control how the events in the simulation unfold. The sounds, time of day and number of people or vehicles on the scene can all be customised. Over several sessions, the veteran is exposed to increasingly intense scenarios that get closer to reliving the memory of the original trauma. The aim of the therapy is to steadily dampen the veteran’s negative reactions to the memory. Bravemind is now used in around 60 treatment centres around the world.
这就是南加州大学医学虚拟现实专家艾伯特·“斯基普”·里佐(Albert “Skip” Rizzo)和阿尔诺·哈斯霍尔特(Arno Hartholt)在2005年开发的“Bravemind”系统,用于治疗从伊拉克和阿富汗战争中返回家园的士兵。退伍军人沉浸在模仿他们的创伤经历的虚拟环境中,向治疗师讲述场景,而治疗师可以控制模拟事件的推进方式。声音、发生的时间以及现场的人数或车辆数都可以自定义。在几次治疗中,这位退伍军人目睹越来越激烈的场景,它们越来越接近于重现原始创伤的记忆。该疗法的目的是逐步减轻退伍军人对记忆的负面反应。现在全球已有约60个治疗中心使用Bravemind。

Bravemind builds on a well-established psychological technique known as exposure therapy, in which people are brought to face their fears in a controlled way. VR adds a way of creating detailed, carefully tuned scenarios that can elicit different levels of fear. It works because, even when people know they are watching computer graphics, their brains nonetheless react to virtual environments as if they were real.
Bravemind是建立在一套被称为“暴露疗法”的成熟的心理学方法之上,这种技术让人们有控制地面对自己的恐惧。虚拟现实(VR)提供了一种途径来创建细致入微、精心调试的场景,可以触发不同程度的恐惧。它之所以有效,是因为即使人们知道自己看到的只是计算机画面,大脑仍然会对虚拟环境做出反应,就好像它们是真实的一样。

Someone who is afraid of heights will find that their heartbeat quickens and palms get clammy even if the precipitous drop they can see is clearly a computer graphic in a VR headset. This is because the brain’s limbic system, which controls the fight-or-flight response, activates within milliseconds in response to potential threats, long before the logical part of the brain—which knows the VR experience is not physically real—can intervene.
恐高的人会发现自己心跳加快、手掌出汗,哪怕他们看到的急坠明显是VR头盔中的计算机图形。这是因为控制着战斗或逃跑反应的大脑边缘系统会在几毫秒内被激活来应对潜在威胁,此时大脑的逻辑部分(它知道VR体验并不真实)还远远来不及介入。

May contain graphic content
可能包含逼真内容

Scientists have used VR systems to create and control complex, multi-sensory, 3D worlds for volunteers in their labs since the 1990s. Rather as an aircraft simulator can train and test pilots in a wide variety of settings, virtual worlds allow psychologists and neuroscientists to watch people’s cognitive and emotional responses in situations that are difficult to set up or control in the real world. But the technology has usually been too clunky and expensive for widespread clinical use.
自1990年代以来,科学家一直在使用VR系统为实验室中的志愿者创建和控制复杂的多感官3D世界。就像飞机模拟器可以训练和测试飞行员应对各种情形一样,虚拟世界让心理学家和神经科学家创建难以在现实世界中设立或控制的场景,并在其中观察人们的认知和情感反应。但这项技术以前往往过于笨重且昂贵,无法广泛应用于临床。

That has started to change, thanks to the falling costs of computing and the increasing capability of the new generation of VR systems. At the same time, the scientific evidence base for the clinical uses of VR has grown. The technology has been successfully applied to tackling schizophrenia, depression and phobias (including the fear of flight, arachnophobia, social anxiety and claustrophobia), and reducing pain in cancer patients undergoing chemotherapy. It can help train spatial-navigation skills in children and adults with motor impairments and assist in rehabilitation after a stroke or traumatic brain injury. The kit can also be used to monitor people and identify medical problems: VR has been used to diagnose attention-deficit hyperactivity disorder (ADHD) and Parkinson’s and Alzheimer’s diseases.
由于计算成本的日益下降和新一代VR系统的功能越来越强,这种情况已经开始改变。同时,将VR用于临床的科学证据也在累积。这项技术已成功地用于治疗精神分裂症、抑郁症和恐惧症(包括飞行恐惧、蜘蛛恐惧、社交焦虑和幽闭恐惧),并减轻接受化疗的癌症患者的痛苦。它可以帮助训练有运动障碍的儿童和成人的空间导航能力,并协助中风或脑外伤后的康复。这套工具还可用于监测人员并发现健康问题:VR已被用于诊断注意力不足过动症(ADHD)、帕金森氏症和阿尔茨海默氏症。

Though each condition is unique, researchers have found common ground rules for designing virtual experiences that work: therapists need to be in control of the scene, deciding what a patient sees and hears in order to modify the strength of the fearful stimulus; the therapy works best when the patient is embodied within an avatar, rather than floating, so that they feel present within the scene; and the patient needs agency, so that they can leave the scene if it gets too overwhelming for them. All this adds up to giving the patient the illusion of control and makes the VR experience feel psychologically “real”.
尽管每种情境都是独特的,但研究人员发现了一些通用的基本规则来设计有效的虚拟体验:治疗师需要控制现场,决定患者看到什么、听到什么,以调节恐惧刺激的强度;当患者存在于某个化身里而不是如游魂般飘荡时,他们感觉自己身处场景之中,治疗效果最佳;并且患者需要服务人员在侧,以便在场景变得不堪忍受时离开。所有这些加在一起,给患者带来了控制的幻觉,并使VR体验在心理上感觉“真实”。

In some cases the therapeutic regime is so robust that, instead of a real-life therapist guiding a patient through an anxiety-inducing simulation, an animated avatar can do the job instead. A clinical trial showed that such an automated system, designed by Daniel Freeman, a psychiatrist at the University of Oxford, helped people reduce their fear of heights. In the simulation, a virtual counsellor guided patients up a virtual ten-storey office complex, where the upper floors overlooked a central atrium. At each floor, the counsellor set the patient tasks designed to test and help them manage their fear responses, such as walking to the edge of a balcony while the safety barrier was lowered or riding on a moving platform over the space above the atrium.
在某些情况下,治疗方案已经非常可靠,可以使用动画化身代替真人治疗师来引导患者完成诱发焦虑的模拟。一项临床试验表明,由牛津大学精神病医生丹尼尔·弗里曼(Daniel Freeman)设计的一套这样的自动化系统可以帮助人们减轻恐高症。在模拟中,一名虚拟咨询师将患者引导到一栋虚拟的10层办公大楼中,其高层俯瞰中央天井。咨询师在每个楼层为患者设置任务来测试恐惧反应并帮助他们应对,例如在安全栅栏降低后走到阳台边缘,或坐在天井上方的移动平台上。

Dr Freeman found that six sessions of virtual, automated therapy over two weeks significantly reduced people’s fear of heights, compared with people who had no therapy. A similar automated virtual therapy for arachnophobia, developed by Philip Lindner at Stockholm University, helped patients eventually touch spiders. The reduction in fear was still apparent when the participants were followed up a year later.
弗里曼发现,与没有接受治疗的人相比,在两周内接受六次虚拟自动治疗大大降低了人们的恐高情绪。斯德哥尔摩大学的菲利普·林德纳(Philip Lindner)开发了一种类似的针对恐惧症的自动虚拟疗法,帮助患者最终触摸了蜘蛛。一年后随访参与者发现他们恐惧的减轻仍然很显著。

For doctors, virtual environments also provide a risk-free way to practise important procedures. Surgeons operate in high-pressure environments with a lot of cognitive demands. “You’ve got to learn very rapidly, and you’ve got to make decisions under time pressure, with millimetre precision,” says Faisal Mushtaq, a cognitive neuroscientist at the University of Leeds in England.
对于医生来说,虚拟环境还提供了一种无风险的方法来练习重要的手术。外科医生工作的高压环境有很多认知方面的要求。英格兰利兹大学的认知神经科学家费萨尔·穆斯塔克(Faisal Mushtaq)说:“你必须非常迅速地学习,并且在时间压力下做出毫米精度的决策。”

Practising with computer simulations can help. In the NeuroVR system, developed by a group of Canadian hospitals and universities, surgeons can use MRI scans from their patients to rehearse removing brain tumours before going in with the knife for real. The surgeon gets a 3D view of the tumour on screens and practises cuts and movements by manipulating instruments attached to a robotic arm that responds with haptic feedback. This allows users to sense whether they are cutting through hard or soft material, or through a tumour versus healthy tissue. An advantage of such a system is that, once a doctor is trained, the technology can be used to perform remote surgery. Both virtual training and remote procedures for patients are useful at a time when covid-19 has forced health-care systems around the world to keep doctors and non-emergency patients apart.
用计算机模拟来练习会有帮助。在由加拿大多家医院和大学开发的NeuroVR系统中,外科医生可以在真正动刀子之前使用患者的MRI扫描来练习切除脑部肿瘤。外科医生可在屏幕上看到肿瘤的三维视图,并通过操纵连接在有触觉反馈的机械手上的仪器练习切割和移动。这让用户可以感觉到自己正在切开硬质还是软质材料、是肿瘤还是健康组织。这种系统的优势在于,一旦医生培训好了,就可以用这种技术操作远程手术。当新冠疫情迫使世界各地的医疗系统把医生和非急诊患者分隔开时,为患者提供虚拟治疗和远程手术都非常有用。

In operation
手术中

When surgeons try to reconstruct a limb, a key problem is identifying important blood vessels that need to be protected during the surgery. In the past a surgeon would try to identify those vessels using an ultrasound probe, but the process is lengthy and imprecise. So James Kinross, a consultant surgeon at Imperial College London, has been experimenting with Microsoft’s HoloLens, an augmented-reality headset, which can overlay computer-generated text and images onto the real world.
当外科医生尝试修复肢体时,一个关键问题是识别在手术期间需要保护的重要血管。过去,外科医生会尝试使用超声探头来识别那些血管,但过程冗长且不精确。因此,伦敦帝国学院的顾问外科医生詹姆斯·金罗斯(James Kinross)一直在试用微软的增强现实头盔HoloLens,它可以把计算机生成的文本和图像叠加到现实世界中。

Dr Kinross has used a CT scan of a patient’s limb to highlight the most important blood vessels. He reconstructed that scan as a 3D model in Unity, a games engine. The HoloLens then overlaid that simulation onto the patient’s real limb in the operating theatre during treatment. “What it meant was that the surgeon could immediately visualize, and very precisely map, the anatomy of these blood vessels, and very quickly identify them and protect them,” says Dr Kinross, who has also used this technique during cancer surgery to help surgeons identify and protect healthy tissue. The adoption of the technology has proceeded very smoothly, he adds, because it is easy to learn and provides “an immediate and very obvious advantage to the clinician”.
金罗斯使用了患者肢体的CT扫描来突出显示最重要的血管。他在游戏引擎Unity中将扫描重建为三维模型。然后,HoloLens在手术过程中会将这一模拟叠加到手术室中患者的真实肢体上。金罗斯说:“这意味着外科医生可以立即看到并非常精确地描绘这些血管的解剖结构,迅速识别并保护它们。”他在癌症手术中也使用了这种技术来帮助外科医生识别并保护健康组织。他补充说,采纳这项技术的进展非常顺利,因为它易于学习,并且“为临床医生提供了直接且非常明显的优势”。

He thinks the technology could be pushed much further and wants to try some real-time collaboration with his colleagues during a surgical procedure. “So if you’re running an operation that’s challenging, or you want to have a discussion with a peer, it’s very easy to do and they can have a first-person view of what you’re looking at,” he says.
他认为这项技术的应用还可以深远得多,并希望在手术过程中尝试与同事实时协作。他说:“因此,如果你正在做一台高难度的手术,或者想和某个同行讨论一下,这就非常容易操作了,他们可以以第一人称视角看到你看到的东西。”

Medical uses for computer simulations are promising, but how useful they are will take time to evaluate. That will require robust clinical trials and discussions of frameworks for data protection on technologies that could, if their potential is achieved, become a new type of medical device.
计算机模拟的医学应用前景广阔,但到底多有用需要时间来评估。这需要可靠的临床试验,并探讨相关技术的数据保护框架——如果这些技术能够发挥其潜力,可能会成为一种新型的医疗设备。

“We don’t want to poison the well,” says Dr Mushtaq. “We don’t want to put out systems that are ineffective, that are going to cost our health-care system, and that are going to negatively impact on the growth of this sector.” His research focuses on closing some of those knowledge gaps by examining how the lessons users learn from practising on virtual simulators translate into skills in the real world. Surprisingly, the fidelity of the images to real surgery is not so important. “Something can look very, very, flashy…it’s got all the blood spewing everywhere and so on,” he says. “But it doesn’t necessarily translate to better learning.”
“我们不想把自己的名声搞坏,”穆斯塔克说,“我们不想推出不好用的系统,这会损害我们的医疗体系,会对这个部门的增长产生负面影响。”他的研究重点是审视用户从在虚拟模拟器上的练习中学到的经验如何转化为现实世界中的技能,从而补上一些知识缺口。出人意料的是,图像对真实手术的保真度并不是那么重要。 “有些东西看起来非常、非常炫……鲜血喷得到处都是之类的,”他说,“但这不一定会让人学得更好。”

Defining the validity of a simulator can take several forms. The most basic is “face validity”, which reflects how well a simulation looks like the task in the real world. “Construct validity” is a way of comparing performance differences on the simulation between experts and novices. Finally, “predictive validity” is most useful, because it measures how well a person’s performance on a simulator predicts their ability to do the same task in the real world.
对模拟器效度的定义可以有多种形式。最基本的是“表面效度”,它反映了模拟看起来有多像现实世界中的任务。“建构效度”是一种比较专家和新手在模拟中的表现差异的方法。最后,“预测效度”是最有用的,因为它衡量的是一个人在模拟器上的表现是否能预测他们在现实世界中完成相同任务的能力。

Just like real life
栩栩如生

This can also be used to flag when learners are struggling, and provide early intervention and support. Dr Mushtaq and his colleagues have demonstrated both construct and predictive validity for the Nissin (formerly Moog) Simodont dental-surgery simulator, used by the University of Leeds to train its students. In research published in 2019, they found that scores on the simulator predicted someone’s performance in a clinic two years later.
这也可以用来标记学习者何时感到吃力,并提供早期干预和支持。穆斯塔克和他的同事们已经证明了利兹大学用来训练学生的Nissin(原Moog)品牌的Simodont牙科手术模拟器的建构和预测效度。在2019年发表的研究中,他们发现某人在模拟器上的分数可预测两年后在诊所中的表现。

Video-game engines have made face validity easier to achieve for simulators. The next step is to measure construct and predictive validity more robustly. Unfortunately, precious little of this kind of validation work is undertaken by academics or companies selling simulators. To help grease the wheels and encourage researchers to build a body of knowledge, Dr Mushtaq and his colleagues recently created a set of tools and protocols that streamline human-behaviour research and make use of the Unity game engine as a platform. This Unity Experiment Framework takes care of the tedious programming steps—downloading files that track all of a user’s movements, for example, or anonymising participants—needed to turn the game engine into an environment optimised for studying people.
电子游戏引擎使模拟器更容易实现表面效度。下一步是更稳健地衡量建构和预测效度。遗憾的是,此类验证工作很少由学界或销售模拟器的公司承担。为了帮助推进这项工作,并鼓励研究人员建立知识体系,穆斯塔克和同事最近创建了一套工具和协议,可以简化人类行为研究并使用Unity游戏引擎作为平台。这个Unity实验框架处理了繁琐的编程步骤(例如下载跟踪用户所有动作的文件,或是把参与者匿名),这些都是将游戏引擎转变为针对人员研究而优化的环境所需要的。

Mark Mon-Williams, a cognitive psychologist at the University of Leeds who has worked with VR for more than two decades, reckons simulated worlds have huge potential for improving education and physical and mental health. “But if you’re going to make the most of that powerful set of tools,” he says, “then use the scientific process to ensure that it’s done properly.”■
利兹大学的认知心理学家马克·蒙-威廉姆斯(Mark Mon-Williams)在VR领域工作了20多年,他认为模拟世界在改善教育和身心健康方面具有巨大的潜力。“但如果你要充分利用这些强大的工具,”他说,“那就需要科学流程来确保你做得恰当。”