{"id":20272,"date":"2026-11-19T05:18:39","date_gmt":"2026-11-19T08:18:39","guid":{"rendered":"https:\/\/rtmedical.com.br\/tmp-en-1795076318454\/"},"modified":"2026-11-19T05:18:46","modified_gmt":"2026-11-19T08:18:46","slug":"mixed-reality-ai-procedures","status":"publish","type":"post","link":"https:\/\/rtmedical.com.br\/en\/mixed-reality-ai-procedures\/","title":{"rendered":"Mixed Reality and AI Bring 3D Imaging Into Procedures"},"content":{"rendered":"<p>Mixed reality and AI-generated 3D maps have reached real procedures: holograms that fuse CT with live ultrasound to guide needles, three-dimensional prostate cancer maps used to plan treatment, and virtual screens that put endoscopic video directly in the surgeon&#8217;s line of sight. That is the picture in a Radiology Today feature by Rebecca Montz (November\/December 2026), built on interviews with physicians at Cleveland Clinic, UPMC in Pittsburgh and a urology practice in Los Angeles.<\/p>\n<p>Both commercial products in the story hold FDA 510(k) clearance, but the published evidence mostly measures technical accuracy in phantoms, cadavers, animals and reader studies. Clinical gains such as fewer complications or less radiation have not been shown yet, and some of the figures come from the companies themselves.<\/p>\n<figure class=\"wp-block-image size-large\"><img decoding=\"async\" class=\"alignright lazyload\" data-src=\"https:\/\/rtmedical.com.br\/wp-content\/uploads\/2026\/11\/realidade-mista-radiologia-intervencionista.jpg\" alt=\"Illustration: interventional radiologist wearing a mixed reality headset and holding an ultrasound probe views a holographic 3D liver model with a planned needle path above a draped patient\" width=\"640\" src=\"data:image\/svg+xml;base64,PHN2ZyB3aWR0aD0iMSIgaGVpZ2h0PSIxIiB4bWxucz0iaHR0cDovL3d3dy53My5vcmcvMjAwMC9zdmciPjwvc3ZnPg==\" style=\"--smush-placeholder-width: 1672px; --smush-placeholder-aspect-ratio: 1672\/941;\"><figcaption>Illustrative AI-generated image; it does not depict a real product or procedure.<\/figcaption><\/figure>\n<h2>XR90: CT and live ultrasound in a single hologram<\/h2>\n<p>The MediView XR90 merges a 3D model from pre-procedure CT with real-time ultrasound and the needle position, both tracked by electromagnetic sensors, and projects the result into an augmented reality headset. FDA cleared it in July 2023 (<a href=\"https:\/\/www.accessdata.fda.gov\/scripts\/cdrh\/cfdocs\/cfpmn\/pmn.cfm?ID=K223125\" target=\"_blank\" rel=\"noopener\">510(k) K223125<\/a>) as an adjunct display workstation for ultrasound-guided needle procedures: the virtual images must always be used alongside conventional monitors, and the system is not intended for primary diagnosis.<\/p>\n<p>At Cleveland Clinic Weston Hospital in Florida, interventional radiologist Kevin Stadtlander calls it &#8220;a way to use two imaging modalities at once,&#8221; with a workflow identical to a procedure done with a handheld ultrasound. Adam Rakestraw, MediView&#8217;s cofounder and chairman, says the technology has been used in more than 500 procedures across more than 20 anatomical locations, and that internal evaluations showed about 47% less ergonomic strain. Both figures are the company&#8217;s own and have not been published or independently verified.<\/p>\n<p>The published record is more modest. In the 510(k) summary, mean target registration error (TRE), the distance between needle tip and target, was 2.5 mm in a phantom, 2.3 mm in cadavers and 2.9 mm in an animal model. In a <a href=\"https:\/\/doi.org\/10.7602\/jmis.2025.28.3.122\" target=\"_blank\" rel=\"noopener\">2025 swine study<\/a> with company-affiliated coauthors, accuracy and first-pass success were similar to ultrasound alone (2.9 vs 3.0 mm; 41.7% vs 39.1%), but mean needle repositioning fell from 3.0 to 0.8. A multicenter soft-tissue biopsy study began enrolling in July 2025; no published results could be found.<\/p>\n<h2>Unfold AI: a 3D map of prostate cancer<\/h2>\n<p>In urologic oncology, the tool in question is Avenda Health&#8217;s Unfold AI. The software combines MRI, biopsy coordinates and results, and clinical data to give every prostate voxel a probability of containing clinically significant cancer (Gleason grade group 2 or higher); the physician picks a threshold and gets a lesion contour plus a confidence score that all of the tumor sits inside it. FDA cleared it in November 2022 (<a href=\"https:\/\/www.accessdata.fda.gov\/scripts\/cdrh\/cfdocs\/cfpmn\/pmn.cfm?ID=K221624\" target=\"_blank\" rel=\"noopener\">510(k) K221624<\/a>), when it was marketed as iQuest.<\/p>\n<p>&#8220;If we have a 3D image of the tumor, then we know which part of the prostate we can go and treat focally,&#8221; S. Adam Ramin, medical director of Urology Cancer Specialists in Los Angeles, told the magazine. In nerve-sparing robotic prostatectomy, he hopes the map will help achieve negative margins while preserving urinary and sexual function, an expectation that has not been measured yet.<\/p>\n<p>The main evidence is a <a href=\"https:\/\/doi.org\/10.1097\/JU.0000000000003960\" target=\"_blank\" rel=\"noopener\">multireader study in The Journal of Urology<\/a> (2024): seven urologists and three radiologists contoured 50 prostatectomy cases with and without the software, with whole-mount histopathology as the reference. Balanced accuracy was 84.7% with AI, versus 67.2% for cognitive contours and 75.9% for hemigland contours; the contour captured all significant tumor in 72.8% of cases with AI and in 1.6% without it. Nearly all authors have ties to Avenda, and the endpoint is contour quality, not recurrence or quality of life. The tool joins other threads in prostate imaging, such as the <a href=\"https:\/\/rtmedical.com.br\/en\/ai-prostate-mri-scarlet-1\/\">open-source AI that matched radiologists on prostate MRI<\/a> and <a href=\"https:\/\/rtmedical.com.br\/en\/psma-pet-biopsy-negative-mri\/\">PSMA PET\/CT-guided biopsy<\/a>.<\/p>\n<h2>UPMC: surgical video inside the headset<\/h2>\n<p>In Pittsburgh, the Surreality Lab, led by neurosurgeon Edward Andrews and computer scientist Jacob Biehl, built a platform that brings endoscope video, navigation, imaging studies and physiologic data into a shared virtual workspace. &#8220;We can expand things, make them larger, smaller, move them here, move them there in our view,&#8221; otolaryngologist Garret Choby told the magazine, adding that the technology changes how surgeons view information, not how they operate.<\/p>\n<p>The first clinical case, announced in February 2026, was an endoscopic endonasal pituitary tumor operation at UPMC Presbyterian, performed by Georgios Zenonos and Choby with an Apple Vision Pro; the <a href=\"https:\/\/doi.org\/10.1055\/a-2945-8171\" target=\"_blank\" rel=\"noopener\">case report<\/a> describes it as completed without intraoperative complications. Earlier, a <a href=\"https:\/\/doi.org\/10.1227\/ons.0000000000002029\" target=\"_blank\" rel=\"noopener\">cadaver study in Operative Neurosurgery<\/a> compared mixed reality streaming with 2D monitors: four neurosurgeons, 16 dissections, no difference in total errors (p = 0.73). The authors also recorded discomfort, eyestrain and difficulty concentrating, although Andrews says the system &#8220;doesn&#8217;t seem to offer any increased cognitive load.&#8221;<\/p>\n<h2>Why fusing images in the field of view matters<\/h2>\n<p>Ultrasound shows the needle in real time without radiation, but its field is narrow and it misses deep targets or those behind gas and bone; CT shows the whole anatomy, but it is static and irradiates patient and staff when used to check the needle. Fusing the two requires spatial registration, and every step adds error: breathing, organ deformation, shifting markers. In a <a href=\"https:\/\/doi.org\/10.1117\/1.JMI.11.6.062602\" target=\"_blank\" rel=\"noopener\">cadaver study in the Journal of Medical Imaging<\/a>, XR90&#8217;s CT-ultrasound fusion error was 4.4 mm, larger than the final needle error, which helps explain why the indication requires live ultrasound. MIT researchers are testing a related idea with an <a href=\"https:\/\/rtmedical.com.br\/en\/mit-augmented-reality-ultrasound\/\">augmented reality ultrasound system<\/a>.<\/p>\n<p>Stadtlander thinks the approach could cut confirmation CT scans and eventually reach procedures now done under fluoroscopy. For hospitals in Latin America and other cost-sensitive settings, where ultrasound-guided biopsy and ablation are routine, the question is cost-effectiveness: headsets, tracking and PACS integration are expensive, clinical use depends on local regulatory approval, and the gains have not been measured in patients. MediView&#8217;s plan to turn every tracked procedure into an AI dataset also raises governance and privacy questions.<\/p>\n<h2>Limitations: plenty of promise, few outcomes<\/h2>\n<p>The feature is a set of interviews, not new data, and the sources have a stake in the technology; one interviewee is a MediView executive. The evidence found is technical and early: accuracy in phantoms, cadavers and pigs, one retrospective reader study and one case report, with no measurement of complications, needle passes in patients, radiation dose or function after prostatectomy. There is also hardware dependence: XR90 was cleared with HoloLens 2, which Microsoft stopped producing in 2024, with security updates through the end of 2027.<\/p>\n<p>The next milestones are the multicenter XR90 study, the clinical trials under way at UPMC, whose platform has been spun out into a startup called SymphonyMR, and prospective Unfold AI data in focal therapy and surgery. Until then, these are tools cleared as adjuncts, with good bench accuracy and a plausible case for ergonomic and planning gains, but no proof yet of patient benefit.<\/p>\n<p><strong>Source:<\/strong> <a href=\"https:\/\/www.radiologytoday.net\/in-the-mix\/\" target=\"_blank\" rel=\"noopener\">Radiology Today \u2014 In the Mix<\/a>; <a href=\"https:\/\/doi.org\/10.1097\/JU.0000000000003960\" target=\"_blank\" rel=\"noopener\">Mota et al., The Journal of Urology, 2024<\/a>.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>Mixed reality and AI now guide needles, map prostate cancer and stream surgical video into headsets. See what the evidence shows so far.<\/p>\n","protected":false},"author":5,"featured_media":20209,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"om_disable_all_campaigns":false,"_monsterinsights_skip_tracking":false,"_monsterinsights_sitenote_active":false,"_monsterinsights_sitenote_note":"","_monsterinsights_sitenote_category":0,"ngg_post_thumbnail":0,"_rt_cluster":"","fifu_image_url":"","fifu_image_alt":"","footnotes":""},"categories":[102,100],"tags":[],"class_list":["post-20272","post","type-post","status-publish","format-standard","has-post-thumbnail","category-ai","category-radiology"],"aioseo_notices":[],"rt_seo":{"title":"","description":"Mixed reality and AI now guide needles, map prostate cancer and stream surgical video into headsets. See what the evidence shows so far.","canonical":"","og_image":"","robots":"index,follow","schema_type":"Article","include_in_llms":true,"llms_label":"Mixed reality and AI in image-guided procedures","llms_summary":"Radiology Today feature on mixed reality and AI in procedures: MediView XR90 (holographic CT plus live ultrasound, FDA 510(k) K223125), Unfold AI 3D prostate cancer maps (510(k) K221624) and UPMC's mixed reality platform in endonasal surgery. Evidence is still mostly technical accuracy; vendor figures attributed.","faq_items":[],"video":[],"gtin":"","mpn":"","brand":"","aggregate_rating":[]},"_links":{"self":[{"href":"https:\/\/rtmedical.com.br\/en\/wp-json\/wp\/v2\/posts\/20272\/"}],"collection":[{"href":"https:\/\/rtmedical.com.br\/en\/wp-json\/wp\/v2\/posts\/"}],"about":[{"href":"https:\/\/rtmedical.com.br\/en\/wp-json\/wp\/v2\/types\/post\/"}],"author":[{"embeddable":true,"href":"https:\/\/rtmedical.com.br\/en\/wp-json\/wp\/v2\/users\/5\/"}],"replies":[{"embeddable":true,"href":"https:\/\/rtmedical.com.br\/en\/wp-json\/wp\/v2\/comments\/?post=20272"}],"version-history":[{"count":1,"href":"https:\/\/rtmedical.com.br\/en\/wp-json\/wp\/v2\/posts\/20272\/revisions\/"}],"predecessor-version":[{"id":20274,"href":"https:\/\/rtmedical.com.br\/en\/wp-json\/wp\/v2\/posts\/20272\/revisions\/20274\/"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/rtmedical.com.br\/en\/wp-json\/wp\/v2\/media\/20209\/"}],"wp:attachment":[{"href":"https:\/\/rtmedical.com.br\/en\/wp-json\/wp\/v2\/media\/?parent=20272"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/rtmedical.com.br\/en\/wp-json\/wp\/v2\/categories\/?post=20272"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/rtmedical.com.br\/en\/wp-json\/wp\/v2\/tags\/?post=20272"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}