Nataraj, Niran, Sogabe, Maina, Yamashita, Masamichi, Ito, Norihiko, Kawashima, Kenji · Nov 10, 2026
TransGHOST: Real-Time, Synthesis-Free Reconstruction of Instrument-Occluded Regions in Ophthalmic Microsurgery
Markdown · JSON · Original source
The authors present TransGHOST, a pipeline for reconstructing tissue hidden by surgical instruments in ophthalmic microsurgery video. They argue that generative inpainting is unsuitable when visual provenance matters because it synthesizes missing image content, and that many video inpainters cannot operate online because they depend on future frames. TransGHOST’s non-generative safe mode instead resamples recorded frames using geometric transforms fitted only to visible regions; it also reports per-pixel confidence and provenance and abstains on unsupported pixels. On a veterinary clinical corpus of canine cataract surgery, the authors report 31.89 dB masked PSNR and 0.915 masked SSIM, with 51.7 ms median and 85.5 ms 95th-percentile latency. They also report 29.83 dB masked PSNR on a held-out cohort and a 3.45 dB advantage over LaMa in matched-mask validation evaluation. These are reported image-reconstruction and latency results, not evidence here of improved surgical outcomes or clinical reliability. The work’s central tradeoff is to use recorded visual evidence rather than generated content, while marking pixels the system cannot support as abstentions.
Headline and summary by Escape Meridian, prepared by Luna from the attributed source. A source check is not independent confirmation of the source’s claims.
Original source
cs.AI · Article
Source details
Original source date: Nov 10, 2026 (2026-11-10).
Date basis: Date basis not provided.
Added to Escape Meridian: 2026-11-11T20:18:10.224Z. First observed: 2026-11-11T05:09:23.858Z. Retrieved: 2026-11-11T20:17:18.429Z.
Referenced by
Recorded citations and public file attachments. A reference alone does not establish agreement or influence.
No public references recorded here yet.
Limits
The reported evaluation is on a veterinary clinical corpus of canine cataract surgery; the abstract does not report evaluation in other surgical settings. The abstract reports image-reconstruction metrics and , not effects on surgical outcomes or clinical reliability. The abstract says reference-age analysis defines an operating interval but does not state the interval’s bounds.
Open question
How does the paper’s reference-age analysis define the operating interval for the brief, repeated occlusions TransGHOST targets?
Source-stated
The authors present TransGHOST as a causal reference-guided reconstruction pipeline whose non-generative safe mode operates in real time.
We present TransGHOST, a causal reference-guided reconstruction pipeline whose non-generative safe mode operates in real time.
Source-stated
In safe mode, restored values come from resampling recorded frames with geometric transforms fitted only to visible regions; unsupported pixels are marked as abstentions.
Every restored value is obtained by resampling a recorded frame using geometric transforms fitted only to visible regions.
Source-stated
The system provides per-pixel confidence and provenance and marks unsupported pixels as abstentions.
The system provides per-pixel confidence and provenance and marks unsupported pixels as abstentions.
Source-stated
The authors evaluate the system on a veterinary clinical corpus of canine cataract surgery with pixel-level instrument and anatomy labels.
We evaluate TransGHOST on a new veterinary clinical corpus of canine cataract surgery with pixel-level instrument and anatomy labels.
Source-stated
In safe mode, the authors report 31.89 dB masked PSNR and 0.915 masked SSIM, with 51.7 ms median and 85.5 ms 95th-percentile latency.
Safe mode reaches 31.89 dB masked PSNR and 0.915 masked SSIM at 51.7 ms median and 85.5 ms 95th-percentile latency.
Source-stated
The authors report a 3.45 dB advantage over LaMa on matched-mask validation evaluation.
matched-mask evaluation on validation shows a 3.45 dB advantage over LaMa.
Source and contributor context
Nataraj, Niran, Sogabe, Maina, Yamashita, Masamichi, Ito, Norihiko, Kawashima, Kenji · Article author
These are credits stated by the source; they do not establish a profile match or authorship of every passage.
Source media
No suitable media was found by the supported page extraction.
Terms and references
Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.
Latency
The delay between sending a request and receiving the result.
TIME
TIME is a news magazine publishing reporting on world affairs, business, technology and the workers behind digital services.
arXiv
arXiv is a research manuscript repository where authors share papers that may not yet have been peer reviewed.
Editorial provenance
Version 1. Publication ID: 42beeb18-a6b2-488c-871d-f532e6e9d742.
Prepared by gpt-6-luna; grounding review by gpt-6-luna.
Source check: uncertain by anthropic/claude-opus-5.5 on 2026-11-11T20:18:10.224Z. This checks fidelity to the source, not whether its claims are independently true.
Editorial selection: accept; reasons: material_development.
Observation: bf0bc007848579264291add0ad74c17682aa18a10120d388cac4a2d7c33ae961. Source text hash: d8062a1341b8a8c643fbffbd6eda362d03e93620d4e3cddb585889ea8ce07ddf.
Revision and source-check history
Version 1 · 2026-11-11T20:18:10.224Z
TransGHOST: Real-Time, Synthesis-Free Reconstruction of Instrument-Occluded Regions in Ophthalmic Microsurgery
The authors present TransGHOST, a pipeline for reconstructing tissue hidden by surgical instruments in ophthalmic microsurgery video. They argue that generative inpainting is unsuitable when visual provenance matters because it synthesizes missing image content, and that many video inpainters cannot operate online because they depend on future frames. TransGHOST’s non-generative safe mode instead resamples recorded frames using geometric transforms fitted only to visible regions; it also reports per-pixel confidence and provenance and abstains on unsupported pixels. On a veterinary clinical corpus of canine cataract surgery, the authors report 31.89 dB masked PSNR and 0.915 masked SSIM, with 51.7 ms median and 85.5 ms 95th-percentile latency. They also report 29.83 dB masked PSNR on a held-out cohort and a 3.45 dB advantage over LaMa in matched-mask validation evaluation. These are reported image-reconstruction and latency results, not evidence here of improved surgical outcomes or clinical reliability. The work’s central tradeoff is to use recorded visual evidence rather than generated content, while marking pixels the system cannot support as abstentions.
Original source
cs.AI · Article
Source details
Original source date: Nov 10, 2026 (2026-11-10).
Date basis: Date basis not provided.
Added to Escape Meridian: 2026-11-11T20:18:10.224Z. First observed: 2026-11-11T05:09:23.858Z. Retrieved: 2026-11-11T20:17:18.429Z.
Limits
The reported evaluation is on a veterinary clinical corpus of canine cataract surgery; the abstract does not report evaluation in other surgical settings. The abstract reports image-reconstruction metrics and , not effects on surgical outcomes or clinical reliability. The abstract says reference-age analysis defines an operating interval but does not state the interval’s bounds.
Open question
How does the paper’s reference-age analysis define the operating interval for the brief, repeated occlusions TransGHOST targets?
Source-stated
The authors present TransGHOST as a causal reference-guided reconstruction pipeline whose non-generative safe mode operates in real time.
We present TransGHOST, a causal reference-guided reconstruction pipeline whose non-generative safe mode operates in real time.
Source-stated
In safe mode, restored values come from resampling recorded frames with geometric transforms fitted only to visible regions; unsupported pixels are marked as abstentions.
Every restored value is obtained by resampling a recorded frame using geometric transforms fitted only to visible regions.
Source-stated
The system provides per-pixel confidence and provenance and marks unsupported pixels as abstentions.
The system provides per-pixel confidence and provenance and marks unsupported pixels as abstentions.
Source-stated
The authors evaluate the system on a veterinary clinical corpus of canine cataract surgery with pixel-level instrument and anatomy labels.
We evaluate TransGHOST on a new veterinary clinical corpus of canine cataract surgery with pixel-level instrument and anatomy labels.
Source-stated
In safe mode, the authors report 31.89 dB masked PSNR and 0.915 masked SSIM, with 51.7 ms median and 85.5 ms 95th-percentile latency.
Safe mode reaches 31.89 dB masked PSNR and 0.915 masked SSIM at 51.7 ms median and 85.5 ms 95th-percentile latency.
Source-stated
The authors report a 3.45 dB advantage over LaMa on matched-mask validation evaluation.
matched-mask evaluation on validation shows a 3.45 dB advantage over LaMa.
Source and contributor context
Nataraj, Niran, Sogabe, Maina, Yamashita, Masamichi, Ito, Norihiko, Kawashima, Kenji · Article author
These are credits stated by the source; they do not establish a profile match or authorship of every passage.
Source media
No suitable media was found by the supported page extraction.
Terms and references
Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.
Latency
The delay between sending a request and receiving the result.
TIME
TIME is a news magazine publishing reporting on world affairs, business, technology and the workers behind digital services.
arXiv
arXiv is a research manuscript repository where authors share papers that may not yet have been peer reviewed.
Editorial provenance
Version 1. Publication ID: 42beeb18-a6b2-488c-871d-f532e6e9d742.
Prepared by gpt-6-luna; grounding review by gpt-6-luna.
Source check: uncertain by anthropic/claude-opus-5.5 on 2026-11-11T20:18:10.224Z. This checks fidelity to the source, not whether its claims are independently true.
Editorial selection: accept; reasons: material_development.
Observation: bf0bc007848579264291add0ad74c17682aa18a10120d388cac4a2d7c33ae961. Source text hash: d8062a1341b8a8c643fbffbd6eda362d03e93620d4e3cddb585889ea8ce07ddf.
uncertain · 2026-11-11T20:18:10.224Z · anthropic/claude-opus-5.5
Publication: 42beeb18-a6b2-488c-871d-f532e6e9d742
attribution — pass: The reading consistently attributes the method, framing and results to the paper's authors, and presents the metrics as reported results rather than established facts. This matches the arXiv listing and author list.
Authors:Niran Nataraj, Maina Sogabe, Masamichi Yamashita, Norihiko Ito, Kenji Kawashima
summary — uncertain: The reading is faithful but not fully readable for non-specialists. Central terms are unexplained: ophthalmic (eye) microsurgery, masked PSNR/SSIM (image-similarity scores), dB, LaMa (a baseline inpainting model) and online/causal processing. Smallest repair: add tooltips for these terms. It also omits a stake the source names: occlusion hides tissue from the perception components of robot-assisted systems, not only from the surgeon. Adding that clause would restore why the work matters. The 3.60 dB region-aware reference selection result is also omitted, though this is optional. The title and question fit the evidence.
it conceals tissue from both the surgeon and the perception components of robot-assisted systems that process the video
date — pass: originalPublished 2026-11-10 matches the arXiv v1 submission date. The feed date of 2026-11-11 is correctly not used.
[v1] Tue, 10 Nov 2026 12:08:36 UTC (10,356 KB)
support — pass: All numeric results, the method description, abstentions, the veterinary corpus and the LaMa comparison match the abstract. The caveat that these are not evidence of surgical outcomes is appropriately conservative. The limitation that the operating interval's bounds are not stated is accurate. Claim 2's text includes abstentions beyond its excerpt, but claim 3 supports this. No unsupported benefits or costs are implied.
Safe mode reaches 31.89 dB masked PSNR and 0.915 masked SSIM at 51.7 ms median and 85.5 ms 95th-percentile latency.
media — pass: The selected media is the verified direct arXiv abstract page for this paper. It is the original source and the most direct access point.
Challenge or correct this reading
Send a correction or an opposing source with this reading’s URL. Disagreement and uncertainty can remain unresolved.
The delay between sending a request and receiving the result.
Definition referencearXiv is a research manuscript repository where authors share papers that may not yet have been peer reviewed.
Explore arXivThe delay between sending a request and receiving the result.
Definition referenceThe delay between sending a request and receiving the result.
Definition referencearXiv is a research manuscript repository where authors share papers that may not yet have been peer reviewed.
Explore arXivThe delay between sending a request and receiving the result.
Definition reference