{"id":"9c3b7307-8b5b-4166-9a93-ebe54d016d79","arxiv_id":"2507.09697","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A 48-camera re-imaging microscope captures 630-megapixel images over 16.3 x 18.8 mm in one shot at claimed 0.84 um half-pitch resolution, with per-camera focus for curved samples.","lead":"PANORAMA is a microscope that photographs a 16 by 19 millimeter area at sub-micron detail in a single shot, without moving the sample. It adjusts each of 48 lenses independently to stay focused on curved or uneven surfaces, which could speed up large-area tissue imaging.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Resolution claim rests on an unverified off-design assumption: the 436 nm Zeiss S-Planar lens is treated as diffraction-limited at 510 nm over the full field, with only USAF group 8 evidence (~1.1–1.2 µm half-pitch) shown.","rationale":"The reader's weakest assumption identifies exactly the load-bearing link in the paper: the claimed 0.84 µm half-pitch resolution depends on the 436 nm photolithography lens being diffraction-limited at 510 nm over the full 16.3×18.8 mm field. I agree that this is unverified. The paper's own wording switches from \"theoretical resolution of 0.82 µm\" (from the Rayleigh formula) to an abstract claim of 0.84 µm without presenting a measured MTF or wavefront characterization at the operating wavelength. The USAF data in Fig. 1, described as showing groups G8E5/G8E6, correspond to half-pitches near 1.1–1.2 µm rather than 0.84 µm, so they cannot confirm the headline figure. The supplement is cited as containing more complete characterization, but it is not present in the submitted text, and data availability is restricted. Because the resolution claim is central to the paper's contribution and the submitted evidence does not verify it, the conditional verdict is appropriate. My concern does not move the verdict; it reinforces the reader's conditional assessment. If the requested MTF/field-dependent resolution data were provided and showed 10% contrast at 595 lp/mm across the full FOV, the concern would be resolved.","tokens_in":6709,"tokens_out":7438,"duration_ms":84700,"concrete_test":"Acquire a chrome-on-glass resolution target with line/space features spanning 0.7–1.5 µm half-pitch and image it through PANORAMA at 510 nm; report contrast at 595 lp/mm (0.84 µm half-pitch) at the center and at all four corners of the 16.3×18.8 mm FOV. In parallel, measure the Zeiss S-Planar lens wavefront error or through-focus MTF at 510 nm across the full field. If contrast at 595 lp/mm falls below the stated resolution criterion (e.g., 10%) anywhere in the FOV, the abstract's 0.84 µm claim is not supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that PANORAMA achieves 0.84 µm half-pitch resolution over a 16.3×18.8 mm² FOV requires the primary lens to be diffraction-limited at the operating wavelength and across the entire field. The paper relies on a Carl Zeiss S-Planar 436 nm photolithography lens (NA 0.38) used with a 510/10 nm bandpass filter and cites a \"theoretical resolution of 0.82 µm at 510 nm\" from the Rayleigh formula. That formula is only valid if aberrations are negligible. The lens is designed and corrected for 436 nm, and the paper presents no MTF, wavefront error, or through-focus measurement at 510 nm for any field position. The only experimental resolution evidence in Fig. 1(b4–b6) is a USAF target whose smallest displayed elements, G8E5/G8E6, correspond to roughly 1.2 and 1.1 µm half-pitch (about 400–460 lp/mm), not 0.84 µm half-pitch (about 595 lp/mm). Resolving those groups does not demonstrate the claimed resolution because the target does not contain features fine enough to verify it. The numerical resolution claim therefore rests on an assumption that is both load-bearing and unverified. The photolithography lens could have chromatic focal shift, field curvature, or higher-order aberrations at 510 nm away from its design wavelength/configuration, making the actual resolution markedly coarser than the Rayleigh limit. The supplement is cited as containing \"comprehensive characterization,\" but it is not included in the preprint, and the data availability statement says underlying data are not public. If the supplement contains full MTF curves, the concern would be answerable; as submitted, the central resolution figure is not supported by the presented evidence.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports PANORAMA, a re-imaging microscope built from a Zeiss S-Planar 436 nm photolithography objective, a large-aperture tube lens, and a 6×8 array of independently focusable micro-cameras. The authors claim single-shot gigapixel imaging over a 16.3×18.8 mm field at 0.84 µm half-pitch resolution, with per-camera refocusing to keep curved or uneven samples in focus without mechanical scanning. Demonstrations include a USAF resolution target, a rat brain slice, and curved onion epidermis in both brightfield and fluorescence modes, with a direct comparison between flat-focus and multi-focus configurations.","tokens_in":6968,"tokens_out":6202,"duration_ms":75520,"significance":"If fully supported, this is a substantial practical advance: it would permit centimeter-scale, submicron lateral resolution in a single exposure and would add a simple per-camera refocusing mechanism for non-flat samples. The engineering description is concrete and the onion experiment is a useful controlled comparison. I also credit the paper for not using fitted parameters in its central performance claims and for reporting system numbers (FOV, overlap, DOF, magnification range) that are in principle checkable. The main weakness is that the headline 0.84 µm half-pitch resolution is not experimentally established by the data shown in the manuscript.","major_comments":[{"comment":"The claimed 0.84 µm half-pitch resolution is not supported by the USAF target data shown. The finest displayed elements, G8E5 and G8E6, have line widths of approximately 1.23 µm and 1.10 µm (roughly 406 and 456 lp/mm), whereas a 0.84 µm half-pitch corresponds to about 595 lp/mm. The contrast profile in Fig. 1(b6) therefore only demonstrates resolution coarser than the headline value. To support the claim, the authors should show a target with features finer than 0.84 µm (e.g., USAF group 9 elements or an equivalent high-frequency grating), or provide a measured MTF or knife-edge response at 510 nm; otherwise the stated resolution should be revised to the value actually demonstrated.","section":"Fig. 1(b4–b6) and system characterization paragraph"},{"comment":"The resolution claim assumes that the 436 nm photolithography lens is diffraction-limited at the operating wavelength of 510 nm across the full 16.3×18.8 mm field. The paper introduces a 510/10 nm bandpass filter to reduce chromatic aberration, but it presents no MTF, wavefront, or through-focus measurement at 510 nm for any field position. The Rayleigh limit with NA=0.38 and λ=510 nm is only valid if residual aberrations are negligible over the entire field. Please provide such measurements, or explicitly label the 0.84 µm figure as a design estimate rather than a measured system specification.","section":"Primary lens description (Letter 2, 'we used the Carl Zeiss S-Planar 436 nm photolithography lens...')"},{"comment":"The manuscript repeatedly cites Supplement 1 (S1, S2) for 'comprehensive characterization' and for stitching details, but the supplement is not included with the posted preprint, and the data availability statement says the data are not publicly available. I have no way to verify the claimed resolution, the exact stitching procedure, or the stated overlap-based magnification calibration from the material provided. The authors should make the supplement and at least representative raw or unstitched sub-images available for review; if the supplement already exists in the journal version, it should be supplied with any resubmission.","section":"Supplemental document and Data availability statements"}],"minor_comments":[{"comment":"The phrase '16.3×18.8 mm2 FOV' uses 'mm2' as if it were a unit of area; it should read '16.3 mm × 18.8 mm FOV'.","section":"Abstract"},{"comment":"The caption says 'The insects indicate the regions examined in detail'; this should be 'insets'.","section":"Fig. 4 caption"},{"comment":"The text says 'the entire 2 cm2 section was captured'; this is ambiguous and should be 'a 2 cm × 2 cm section' or 'a sample of area 2 cm²'.","section":"Letter 3, brightfield demonstration"},{"comment":"The resolution values in Table 1 are listed without specifying whether they are theoretical or experimentally measured for each system; adding a footnote with the measurement method would make the comparison more meaningful.","section":"Table 1"},{"comment":"The contrast profile plot lacks axis scales and a clear indication of which group elements are being traced; adding a spatial-frequency or line-width scale would let the reader check the resolution claim directly.","section":"Fig. 1(b6)"}],"recommendation":"major_revision","confidential_remarks":"The key issue is evidentiary rather than conceptual: the system design is plausible and the demonstrations are promising, but the central resolution number is not shown to be met by the USAF data, and the supplement that supposedly contains the detailed characterization is absent from the posted version. I would ask the editor to require the supplement and the additional resolution measurement before considering publication. The author disclosures with respect to Ramona Optics are stated and do not, by themselves, concern me."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a real advance in re-imaging microscope design, and the curvature-adaptive demonstration is convincing. But the headline 0.84 µm half-pitch resolution is not backed by the USAF data presented, and the central figure rests on an unverified assumption about the photolithography lens.\n\nWhat's new: the combination of a telecentric photolithography primary lens, a flat 6x8 camera array, and per-camera axial refocusing in a single-shot architecture. That yields seamless gigapixel coverage over 16.3x18.8 mm² with submicron-ish resolution and demonstrable focus on curved samples. Compared to their own M-FAST, this is a clear improvement in resolution and throughput; compared to MCAM, it avoids tiling. The rat-brain and onion images are nice, and the fluorescence version works too. The paper is honest about the practical bounds: DOF ~45 µm, max curvature 6.6 m⁻¹, inter-camera magnification mismatch up to ~10–15% before stitching artifacts.\n\nSoft spots: the resolution evidence. The USAF close-ups stop at groups G8E5/G8E6, which correspond to ~1.1–1.2 µm half-pitch, not 0.84 µm. The theoretical Rayleigh value presumes the Zeiss S-Planar lens is diffraction-limited at 510 nm across the full field; it is designed for 436 nm, and no MTF, wavefront, or through-focus data at 510 nm are in the preprint. The supplement is cited but not present, and data are not public. That makes the strongest claim in the abstract unverified as submitted. The stitching accuracy and DOF numbers also lack error bars or a clear measurement description, though those are secondary.\n\nI don't think this is a fatal flaw: the lens may well perform near the Rayleigh limit at 510 nm, and the supplement might contain the missing MTF curves. But as it stands, a referee should ask for that evidence before accepting the resolution claim. The system concept and its qualitative demonstrations are solid, and the limitations discussion is refreshingly explicit.\n\nWho it's for: anyone working on large-area microscopy, histology, or computational imaging. Worth serious peer review, with the request for MTF data and higher-frequency USAF targets.","headline":"PANORAMA is a genuine engineering step toward single-shot cm-scale submicron imaging with per-camera refocusing, but the 0.84 µm resolution claim outruns the evidence shown.","tokens_in":7629,"tokens_out":2498,"would_cite":false,"duration_ms":27614,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper claims that a single-shot re-imaging microscope with a 6×8 array of focus-adjustable cameras can capture a seamless 630-megapixel image of a 16.3 × 18.8 mm² area at 0.84 µm half-pitch resolution without mechanical scanning…","keywords":["gigapixel microscopy","multi-camera array microscope","re-imaging optics","field curvature compensation","single-shot imaging","curved sample imaging","fluorescence microscopy","space-bandwidth product"],"falsifier":"Image a USAF 1951 target across the full field and record the finest group resolved in both center and corner, or measure the modulation transfer function of the primary lens at 510 nm. If G8E6 (roughly 1.1 µm half-pitch) is not clearly resolved everywhere, or if the MTF at 0.84 µm half-pitch stays below the detection contrast threshold, the stated resolution claim is not supported.","tokens_in":6454,"feed_emoji":"🔬","tokens_out":9072,"duration_ms":87733,"temperature":0.7,"pith_summary":"This paper claims that a microscope built from a large telecentric photolithography lens and a flat 6×8 array of small cameras, each with independently adjustable focus, can photograph a 16.3 × 18.8 mm² area at 0.84 µm half-pitch resolution in a single exposure. The system, called PANORAMA, produces 630-megapixel stitched images without mechanical scanning. By refocusing individual cameras, the effective focal surface can be curved to match either the lens's field curvature or the shape of a non-flat sample. If correct, this removes the usual trade-off between field of view and resolution for centimeter-scale samples and eliminates the need to tile, stitch, or focus-stack for many biological and materials imaging tasks.","feed_headline":"One snapshot captures 630-megapixel images at submicron resolution","feed_subtitle":"Per-camera focus lets a flat sensor array follow curved samples, replacing scanning and focus stacking.","key_machinery":"The machinery is a re-imaging architecture. A commercial telecentric photolithography lens with numerical aperture 0.38, designed for 436 nm and operated at 510 nm through a 10 nm bandpass filter, forms a large intermediate image of the sample. A 6×8 array of micro-cameras, each with a 14.64 mm focal-length lens and a monochrome CMOS sensor, relays local patches of that intermediate image onto the flat sensor array. Each micro-camera can be moved axially, which changes the array's effective focal surface from a plane to a curved shell; the shell can be set to cancel the objective's field curvature or to follow a curved specimen. Neighboring fields overlap by about 10–30%, and the overlaps are used to register and stitch the sub-images into one seamless gigapixel composite.","core_discovery":"The paper's central claim is that field curvature—normally a defect—can be turned into a feature: a high-NA telecentric objective forms a curved intermediate image, and a dense array of 48 small cameras is focused patch-by-patch onto that curved surface, so a flat sensor array captures the whole field sharply in one shot. With flat samples, individually setting each camera's focus compensates the lens's field curvature; with curved samples, the same per-camera refocusing matches the sample's own surface. The authors report a 16.31 × 18.84 mm² field, 0.84 µm half-pitch resolution (0.82 µm theoretical at 510 nm), 630 MP per snapshot, and demonstrate the system on a rat brain slice and on a curved onion epidermis in both brightfield and fluorescence. The paper's comparison table places PANORAMA as the only listed system with both single-shot acquisition and curvature adaptability at sub-micrometer resolution.","pith_inferences":["If the 0.84 µm claim holds, PANORAMA would close the gap between whole-slide scanners (fast but often slower or coarser) and high-NA microscopes (diffraction-limited but tiny FOV), enabling gigapixel histology as a snapshot.","The authors do not state it, but the per-camera focus control could be automated and combined with a depth sensor to create a dynamically reconfigurable focal surface for moving or deformable samples.","The overlapping fields of view could be leveraged for stereo or light-field depth estimation, extending the system from 2D panoramas to 3D surface measurements without hardware changes.","The central resolution number depends on the primary lens; if an independent MTF test shows the lens is not diffraction-limited at 510 nm across the full field, the real half-pitch may be closer to the roughly 1.1 µm demonstrated by the USAF target in the figure, a useful but more modest result."],"forward_implications":["An entire 2 cm² brain section can be captured at cellular resolution in one shot, removing the need for XY scanning in histology-style imaging.","Curved or unflattenable specimens, such as tissue on curved substrates, can be imaged sharply in a single exposure in both brightfield and fluorescence, eliminating focus stacks.","Fluorescence and brightfield can share the same optics at the same resolution, with only the illumination and emission filter changed.","The single-shot, no-moving-parts design increases throughput relative to tiled scanning systems and avoids stitching artifacts from sample drift between tiles.","The architecture scales: adding more cameras or larger sensors directly enlarges the field of view at the same resolution."],"supporting_citations":[{"why":"Describes the prior M-FAST cascaded microscope design that PANORAMA extends, providing the baseline resolution and FOV numbers.","marker":"[13]"},{"why":"Introduced a re-imaging architecture with a curved intermediate image captured by a camera array, the design lineage PANORAMA follows.","marker":"[3]"},{"why":"Defines the multi-camera array microscope concept used to circumvent the space-bandwidth product limit.","marker":"[10]"},{"why":"Demonstrated multiscale gigapixel photography via re-imaging, the conceptual basis for forming and capturing a large intermediate image.","marker":"[9]"},{"why":"Describes the MCAS histology scanner (0.6 µm, tile scanning) that PANORAMA claims to surpass with single-shot acquisition.","marker":"[2]"},{"why":"RA-WiFi mesoscopy system listed in the comparison table as a prior curvature-capable but lower-resolution baseline.","marker":"[7]"},{"why":"AMATERAS-2 wide-field imaging system listed as a direct wide-field comparison point in Table 1.","marker":"[8]"},{"why":"Mesolens scanning-based large-FOV microscope used as a comparison baseline for resolution and acquisition mode.","marker":"[6]"}],"fun_headline_variants":["Adaptive camera array focuses on curved samples in one shot","630-megapixel single shot adapts to curved surfaces","Curvature-adaptive gigapixel microscopy without scanning","One-shot gigapixel imaging that bends focus to fit curves","Flat sensor array, curved samples: 630 MP in one capture"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The claim rests on the primary photolithography lens being diffraction-limited at 510 nm over the full 16.3 × 18.8 mm field; the paper does not supply an MTF or wavefront measurement at that wavelength, and the included USAF target only proves about 1.1 µm half-pitch, coarser than the headline 0.84 µm.","fun_headline_variants_meta":{"raw":{"variants":["Adaptive camera array focuses on curved samples in one shot","630-megapixel single shot adapts to curved surfaces","Curvature-adaptive gigapixel microscopy without scanning","One-shot gigapixel imaging that bends focus to fit curves","Flat sensor array, curved samples: 630 MP in one capture"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000697,"raw_usage":{"total_tokens":3134,"prompt_tokens":910,"completion_tokens":2224,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":526,"completion_tokens_details":{"reasoning_tokens":2139}},"tokens_in":526,"tokens_out":2224,"duration_ms":16169,"temperature":1.0,"reasoning_tokens":2139,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T17:49:50.297221+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Image a USAF 1951 target across the full field and record the finest group resolved in both center and corner, or measure the modulation transfer function of the primary lens at 510 nm. If G8E6 (roughly 1.1 µm half-pitch) is not clearly resolved everywhere, or if the MTF at 0.84 µm half-pitch stays below the detection contrast threshold, the stated resolution claim is not supported.","supporting_citations":[{"cited_title":"Y ang, M","cited_arxiv_id":null,"evidence_quote":"Describes the prior M-FAST cascaded microscope design that PANORAMA extends, providing the baseline resolution and FOV numbers."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduced a re-imaging architecture with a curved intermediate image captured by a camera array, the design lineage PANORAMA follows."},{"cited_title":"Harfouche, K","cited_arxiv_id":null,"evidence_quote":"Defines the multi-camera array microscope concept used to circumvent the space-bandwidth product limit."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Demonstrated multiscale gigapixel photography via re-imaging, the conceptual basis for forming and capturing a large intermediate image."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Describes the MCAS histology scanner (0.6 µm, tile scanning) that PANORAMA claims to surpass with single-shot acquisition."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"RA-WiFi mesoscopy system listed in the comparison table as a prior curvature-capable but lower-resolution baseline."},{"cited_title":"Ichimura, T","cited_arxiv_id":null,"evidence_quote":"AMATERAS-2 wide-field imaging system listed as a direct wide-field comparison point in Table 1."},{"cited_title":"McConnell, J","cited_arxiv_id":null,"evidence_quote":"Mesolens scanning-based large-FOV microscope used as a comparison baseline for resolution and acquisition mode."}],"review_version":1}