View in EDS HTML Full Text PDF Full Text

A Comprehensive Guide to High-Fidelity 3D Neuroanatomical Modeling Techniques: A Quantitative Comparison between Photogrammetry and Structured Light Scanning

Saved in:
Bibliographic Details
Title: A Comprehensive Guide to High-Fidelity 3D Neuroanatomical Modeling Techniques: A Quantitative Comparison between Photogrammetry and Structured Light Scanning
Language: English
Authors: Megan M. J. Bauman (ORCID 0000-0003-1910-6617), Amedeo Piazza, Fabio Torregrossa, Charles Wes Price, Jonathan M. Morris, Luciano C. P. C. Leonel, Maria Peris-Celda
Source: Anatomical Sciences Education. 2025 18(7):697-708.
Availability: Wiley. Available from: John Wiley & Sons, Inc. 111 River Street, Hoboken, NJ 07030. Tel: 800-835-6770; e-mail: cs-journals@wiley.com; Web site: https://www.wiley.com/en-us
Peer Reviewed: Y
Page Count: 12
Publication Date: 2025
Document Type: Journal Articles
Reports - Research
Education Level: Higher Education
Postsecondary Education
Descriptors: Anatomy, Neurology, Visual Aids, Fidelity, Photography, Electronic Equipment, Medical Students, Graduate Students
DOI: 10.1002/ase.70062
ISSN: 1935-9772
1935-9780
Abstract: Cadaveric dissections, which are considered the most realistic model to study neuroanatomy, are expensive and not readily available in all centers. Given the surge of technological advances, incorporation of three-dimensional (3D) scanning technologies and 3D models has gained popularity, both in the educational and clinical settings. We present our institutional experience in creating high-fidelity neuroanatomical 3D models using three 3D scanning techniques: structured light 3D scanning, "manual" photogrammetry with a single DSLR camera, and "automatic" photogrammetry using a scanner equipped with five vertically arranged DSLR cameras and an automatic turntable within a square box. A survey study was conducted with 20 neurosurgical residents to assess the quality of the three resulting 3D models. In the study, "manual" photogrammetry was determined to be the most cost-effective technique, while "automatic" photogrammetry was the most time-effective and user-friendly technique. The best visual quality was obtained using "manual" photogrammetry, as determined from survey results of 20 neurosurgical residents. While structured light 3D scanning had the lowest quality of resolution of the texture map, this technique was the most accurate to use for determining measurements, with a fixed accuracy of 0.05 mm. Overall, "manual" photogrammetry can allow for the most detailed 3D models and is the most cost-effective strategy, while structured light 3D scanning is most suitable for obtaining clinically relevant measurements given the high degree of structural accuracy. Alternatively, "automatic" photogrammetry can serve as a hybrid between obtaining relatively high-quality models in a time-effective and user-friendly manner.
Abstractor: As Provided
Entry Date: 2025
Accession Number: EJ1476155
Database: ERIC
Full text is not displayed to guests.
FullText Links:
  – Type: pdflink
    Url: https://content.ebscohost.com/cds/retrieve?content=AQICAHj0k_4E0hTGH8RJwT4gCJyBsGNe_WN95AvKlDbXJGqwxwF929uwWoDtjLo-nNTMelOCAAAA4jCB3wYJKoZIhvcNAQcGoIHRMIHOAgEAMIHIBgkqhkiG9w0BBwEwHgYJYIZIAWUDBAEuMBEEDML6OGCneWD30a_LHwIBEICBmmeFX3t1IErexE9geA_MWC8mqZhtzGti4B_f7H8oisZnxwbh5qHOrvaMpHCCLYZ0wDjkaveOgR7qxNzWMjkhUFckvTYRBoss3FHdp2qMIgbSArhcNW5TRCnkZlijwKBInWAMyUr3729l9zRHjMGnn4MBtOcRDmAvz0AIASSR1lMoDWVkxEPGj2aZll6bosfRp20FODngMIx2IgU=
Text:
  Availability: 1
  Value: <anid>AN0186371264;[8z8k]01jul.25;2025Jul07.02:54;v2.2.500</anid> <title id="AN0186371264-1">A comprehensive guide to high‐fidelity 3D neuroanatomical modeling techniques: A quantitative comparison between photogrammetry and structured light scanning </title> <p>Cadaveric dissections, which are considered the most realistic model to study neuroanatomy, are expensive and not readily available in all centers. Given the surge of technological advances, incorporation of three‐dimensional (3D) scanning technologies and 3D models has gained popularity, both in the educational and clinical settings. We present our institutional experience in creating high‐fidelity neuroanatomical 3D models using three 3D scanning techniques: structured light 3D scanning, "manual" photogrammetry with a single DSLR camera, and "automatic" photogrammetry using a scanner equipped with five vertically arranged DSLR cameras and an automatic turntable within a square box. A survey study was conducted with 20 neurosurgical residents to assess the quality of the three resulting 3D models. In the study, "manual" photogrammetry was determined to be the most cost‐effective technique, while "automatic" photogrammetry was the most time‐effective and user‐friendly technique. The best visual quality was obtained using "manual" photogrammetry, as determined from survey results of 20 neurosurgical residents. While structured light 3D scanning had the lowest quality of resolution of the texture map, this technique was the most accurate to use for determining measurements, with a fixed accuracy of 0.05 mm. Overall, "manual" photogrammetry can allow for the most detailed 3D models and is the most cost‐effective strategy, while structured light 3D scanning is most suitable for obtaining clinically relevant measurements given the high degree of structural accuracy. Alternatively, "automatic" photogrammetry can serve as a hybrid between obtaining relatively high‐quality models in a time‐effective and user‐friendly manner.</p> <p>Keywords: 3D models; neuroanatomical education; neuroanatomy; photogrammetry; structured light 3D scanning</p> <hd id="AN0186371264-2">INTRODUCTION</hd> <p>Within neuroanatomical education and medicine, it is challenging to master three‐dimensional (3D) anatomical relationships and develop spatial awareness of relevant structures given the complexity of the central nervous system (CNS). Cadaveric dissections, which are considered the most realistic model to study neuroanatomy,[<reflink idref="bib1" id="ref1">1</reflink>] are expensive and not readily available in all academic centers in the United States and around the world. Given the surge of technological advances, 3D technologies have gained popularity in order to improve and enhance neurosurgical practices, both in the educational and clinical settings. Stereoscopic images,[[<reflink idref="bib2" id="ref2">2</reflink>]] 3D videos,[[<reflink idref="bib4" id="ref3">4</reflink>]] 3D printing,[[<reflink idref="bib6" id="ref4">6</reflink>]] and interactive 3D models[[<reflink idref="bib8" id="ref5">8</reflink>]] allow the audience to perceive depth and appreciate anatomical relationships, which are of paramount importance in neurosurgical approaches. In the clinical setting, 3D technologies can aid in measurements of patient anatomy for the purposes of operative planning,[<reflink idref="bib10" id="ref6">10</reflink>] along with evaluation of surgical outcomes,[[<reflink idref="bib11" id="ref7">11</reflink>]] among other uses.</p> <p>3D scanning technologies have evolved beyond the initial DICOM‐based 3D model reconstruction,[[<reflink idref="bib13" id="ref8">13</reflink>]] to now possessing the capability of capturing the minute details of cadaveric specimens,[[<reflink idref="bib9" id="ref9">9</reflink>], [<reflink idref="bib15" id="ref10">15</reflink>]] along with creating accurate representations of measured structural relationships.[[<reflink idref="bib16" id="ref11">16</reflink>]] Photogrammetry is a technique that captures an object from all 360° and converts the scan into a 3D digital file.[<reflink idref="bib8" id="ref12">8</reflink>] Similarly, technological advancements have resulted in the development of new methodologies to acquire 3D models, including the use of structured light 3D scanners. These scanners use structured light scanning—measuring the distortion of emitted light bouncing off of an object's surface—to create 3D models.[<reflink idref="bib16" id="ref13">16</reflink>] These models, captured using 3D scanning techniques, can then be manipulated by users in 360° to explore the anatomy and complex approaches in an interactive and realistic format.</p> <p>Given the numerous options for creating 3D neuroanatomical models, the selection of the most appropriate 3D scanning technique becomes challenging and is often influenced by the desired final use of the 3D model within neuroanatomical practice. Therefore, we present our institutional experience in creating high‐fidelity neuroanatomical 3D models and provide a comprehensive guide detailing three commercially available 3D scanning techniques: "manual" one‐camera photogrammetry, "automatic" 5‐camera photogrammetry, and structured light 3D scanning. Further, we compare the benefits and limitations of these three scanning methods relative to one another in order to provide insight into the most optimal use for each technique.</p> <hd id="AN0186371264-3">METHODS</hd> <p></p> <hd id="AN0186371264-4">Ethical considerations</hd> <p>This research was conducted under the Institutional Review Board (IRB) 17‐005898 of Mayo Clinic, Rochester, MN, USA. Explicit permission for the acquisition and use of images was acquired as part of the consent process. All the donors used in this study were provided by the "Mayo Clinic Body Donation Program" in the Department of Clinical Anatomy, Mayo Clinic, Rochester, MN‐US.</p> <hd id="AN0186371264-5">Specimen selection and preparation</hd> <p>One embalmed specimen (10% formalin‐fixed) was used to obtain three 3D neuroanatomical models using three different 3D scanning techniques. The specimen was prepared using the methodology as previously described in Leonel et al.[<reflink idref="bib18" id="ref14">18</reflink>] A six‐vessel latex injection technique was used with blue corresponding to veins and red corresponding to arteries.</p> <p>Following specimen preparation, the dissection was performed in a stepwise manner to expose the entire optic pathway from the orbit to the occipital cortex. Dissections were performed under an operating microscope (Leica M320 F12, Leica Microsystems, Germany; 6–40× magnification) with the aid of microsurgical instruments such as Penfield and Rhoton dissectors, micro‐forceps, and microscissors. In between dissections, the specimen was stored in 10% ethanol to preserve the integrity of the tissues and exposed white matter.</p> <p>As standard in our group's practice, 3D‐photodocumentation (i.e., stereoscopic photography) was completed for each step of the dissection using the technique as previously described in Leonel et al.[<reflink idref="bib18" id="ref15">18</reflink>] Following the final stage of the dissection, the specimen was also captured using three 3D scanning techniques as described below.</p> <hd id="AN0186371264-6">"Manual" photogrammetry</hd> <p>Photogrammetry involves photographing an object of interest from multiple angles and combining the multiple photographs together to build a 3D model. For "manual" photogrammetry, photographs of the dissection were acquired using a single DSLR camera (Canon EOS 6D Mark II DSLR Camera® Canon Inc., Tokyo, Japan) with a 100 mm macro lens (Canon EF 100 mm f/2.8L Macro IS USM Lens®, Canon Inc., Tokyo, Japan). Two hundred and fifty photographs were taken, covering 360° of the specimen, with the aim of capturing the specimen's surface from every angle of view with an average overlap of 70% between each consecutive image. During this process, photos that were out‐of‐focus or blurry were discharged and retaken. The specimen was placed on a rotatable tray in a dark room with artificial ambient lighting set at a color temperature of 52,000 K. The selected photos were processed using AgiSoft Metashape Professional (Version 2.1, https://<ulink href="http://www.agisoft.com/downloads/installer/">www.agisoft.com/downloads/installer/</ulink>) to manually create a 3D model. Background removal from the 3D model was achieved using Blender software (Blender Foundation, Amsterdam, Holland) (Figure 1).</p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/8Z8K/01jul25/ase70062-fig-0001.jpg?ephost1=dGJyMNXb4kSepq84yOvqOLCmsE6epq5Srqa4SK6WxWXS" alt="ase70062-fig-0001.jpg" title="1 "Manual" configuration. (A) The main components this technique consists of a single DSLR camera with 100 mm macro lens placed on a camera tripod mount. The specimen of interest is placed on a rotatable tray facing the camera and is manually rotated as pictures are taken from all angles, with the goal of at least 70% overlap between adjacent images. After photographs of the specimen have been attained from 360°, the height of the camera tripod mount is adjusted to ensure that the superior and inferior aspects of the specimen are captured. (B) For specimens with deeper and/or finer detailed structures, the cameras can be placed closer to the specimen, and additional photos of these structures can be obtained from a closer detail in order to ensure the best image quality of the structure for use in the rendering of the final 3D model." /> </p> <p></p> <hd id="AN0186371264-8">"Automatic" photogrammetry</hd> <p>For "automatic" photogrammetry, our group uses a five‐camera MedCreator® scanner (MedReality, Thyng, Chicago, IL). Similar to the three‐camera scanner previously used by our group and described in de Olivieria et al.,[<reflink idref="bib8" id="ref16">8</reflink>] our five‐camera photogrammetry scanner is equipped with 5 DSLR cameras (Canon EOS 2000D® Canon Inc., Tokyo, Japan) each with 18–400 mm Macrolens lens (Canon Lens Cap E‐67 II®, Canon Inc., Tokyo, Japan) and polarization filters to capture images for the rending of 3D models.[<reflink idref="bib19" id="ref17">19</reflink>] Within the scanning 'box,' an automatic turntable platform rotates the specimen 360° clockwise while photos are simultaneously taken from 71 different angles, for a total of 355 photos. Four fluorescent lights with a color temperature of 52,000 K serve as the lighting source within the scanning box. Customized scanning parameters used to acquire the scan—such as time of ISO, shutter speed, frames per second, etc.—are highlighted in Figure 2. Following positioning of the specimen within the scanner, the remainder of the scanning and reconstruction processes was completed by software packages installed by the manufacturer inside the scanner, allowing for the automatic creation of a 3D model without the need for any direct user input. Specifically, third‐party software—Reality Capture (Reality Capture 1.3, Epic Games, Inc. 620 Crossroads Blvd. Cary, NC 27518)—was employed for the rendering of the 3D model.</p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/8Z8K/01jul25/ase70062-fig-0002.jpg?ephost1=dGJyMNXb4kSepq84yOvqOLCmsE6epq5Srqa4SK6WxWXS" alt="ase70062-fig-0002.jpg" title="2 "Automatic" photogrammetry scanner. (A) The main components of this technique consist of the photogrammetry scanner (a white box which houses the cameras and specimen), computer system with built‐in 3D scanning software, and command panel. (B) Within the scanner, five cameras are arranged in a vertical position with different angulation and can be manually adjusted in order to optimize the frame to capture a given specimen. The specimen is positioned on a platform, which is then placed in the center of an automatic turntable. Four, vertically oriented fluorescent lights with a color temperature of 52,000 K serve as the lighting source within the scanning box. (C) Positioning of the specimen on the platform is achieved through the use of spikes, poles, and markers. Poles are arranged in a triangular shape in the center of the platform and secured through holes built into the platform. Spikes are threaded into the tops of three markers, which are placed on top of the poles. The markers are labeled 1–3 and serve as the registration system for the 3D processing software. Optionally, green coverings can be placed on top of the spikes so that they can be erased during postprocessing of a model. It is important that all aspects of the specimen are above marker 1 as this represents the lower boundary during the image registration process for building the 3D model. (D) The settings our group uses for photogrammetry scans. Briefly (a) ISO represents the sensitivity to light where the higher the number the higher is the light capture, consequently creating a brighter image; (b) the aperture controls the depth of field to be captured by the camera and the amount of flashlight passing through each lens; (c) the shutter speed also controls the exposure to the light where the faster shutter speed results in shorter exposure to light; (d) frames per scan represent the number of pictures captured during scanning by each camera; (e) the pauses during each rotation of the turntable prevent blurry images keeping the specimen steady; (f) automatic reconstruction results in automatic building of the 3D model by the built‐in Reality Capture software; (g) key color modified to 'green' allows the software to recognize and eliminate the markers and green spikes used to support the specimens within the scanner; in order to guarantee the best resolution, the model quality (H) is always set up as 'high' and with a polygon count (i) as 2 million units; (j) texture resolution is set to the maximum of 8192 × 8192 pixels; final file output is set to (k).png for texture map and (l).obj for mesh." /> </p> <p></p> <hd id="AN0186371264-10">Structured light 3D scanning</hd> <p>Finally, the specimen was scanned using a high‐resolution structured light 3D scanner based on blue light technology: Artec® 3D Space Spider scanner (Artec, 4 Rue Lou Hemmer, L‐1748 Senningerberg, Luxembourg). This structured light 3D scanner projects light from a single light source onto a surface of interest. After the light reflects off the object's surface, it is captured by three additional cameras on the scanner, which compares the initial emission angle of light to the final captured angle of light, allowing for determination of an object's surface configuration. In addition, the scanner incorporates a "texture camera," which captures the colored information of an object's surface, allowing for the rendering of a colored 3D model. The Artec® 3D Space Spider represents one of the best structured light 3D scanners currently available, allowing for 0.05 mm of point accuracy and 0.1 mm of resolution of the final 3D model. The handheld nature of this scanner allows the user to manually control the angles and rate of scanning. Additionally, real‐time visual feedback of the scanning process is provided, which allows the user to ensure that all surfaces are captured to the maximal extent before rendering a final 3D model.</p> <p>For the scanning process, the specimen was placed on a rotatable tray with the scanner held between 17 and 30 cm from the specimen's surface. Scanning was conducted in a dark room with artificial ambient lighting set at a color temperature of 52,000 K. The 3D data obtained during the scanning process was then rendered into a final 3D model using Artec Studio 17 (Artec, 4 Rue Lou Hemmer, L‐1748 Senningerberg, Luxembourg) (Figure 3).</p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/8Z8K/01jul25/ase70062-fig-0003.jpg?ephost1=dGJyMNXb4kSepq84yOvqOLCmsE6epq5Srqa4SK6WxWXS" alt="ase70062-fig-0003.jpg" title="3 Structured light 3D scanning configuration. (A) The main component of this technique includes the Artec® 3D Space Spider scanner, which is connected to a computer system containing Artec Studio 17 software. Optionally, the specimen can be placed on a rotatable tray in order to aid with scanning from 360°. (B) The side of the structured light scanner that faces the specimen measures reflected light by three 3D cameras (a and c) and emitted by a single light source (d) to determine the structural components of the specimen. Additionally, the scanner includes a texture camera surrounded by 6‐LED lights (b) in order to capture colored information about the specimen. (C) During the scanning process, the scanner is held between 17 and 30 cm from the specimen's surface. The scanner can be moved superior–inferior and turned clockwise‐counterclockwise to capture all aspects of the specimen, while the rotatable tray is turned to capture 360° of the specimen. (D) A visual representation of the captured surfaces of the specimen are displayed in the 3D scanning software during the scanning process. The area actively being captured is depicted in green. The vertical histogram (red box) displays the ideal scanning distance of the scanner for a particular surface, with the goal of having the majority of the scanning data being within the middle green three boxes. As the scanning process takes place, the user should follow this graph and continue to make adjustments to positioning and distance of the 3D scanner to ensure that data falls within this area of the graph." /> </p> <p></p> <hd id="AN0186371264-12">Covariates of interest for comparison of 3D scanning techniques</hd> <p>Alongside describing the resources required to produce 3D models (i.e., human time required, total time required, financial cost, technological knowledge required), all three models were compared based on parameters that encompass the structural accuracy (i.e., mesh) and resolution of the models themselves (i.e., texture map).</p> <p>Human time was defined as the portion of time in the 3D model building process that required human involvement, including specimen position, manual use of any equipment for scanning, and required user inputs for the 3D model rendering software. Alternatively, total time was considered human time in addition to time solely attributed to software processing. Technological knowledge refers to the relative understanding and proficiency required for users to manipulate raw data within 3D model rendering software, resulting in the creation of a final 3D model. Cost was calculated in United States dollars (USD) based on the current market value of all required devices and instruments used during the scanning process.</p> <p>The mesh of a 3D model represents the structural surface representation of a scanned object. During the 3D scanning process, structural aspects of the object are recorded as vertices (3D points in space) that define the shape and positions of points in the mesh. During model construction, the vertices become connected together to neighboring vertices to form a complex network of polygons (often triangular or quadrangular but can have more complex shapes such as pentagons or hexagons) that approximate the shape and structure of the subject being modeled. An increased number of vertices and subsequent polygons allows for the creation of complex curves and surfaces with greater precision, resulting in increased overall structural detail and intricacy of the final model. This ultimately translates into a more accurate representation of structural relationships within the model, thereby allowing for the ability to use the model to perform structural measurements. To evaluate the mesh among the three 3D models, we compared the number of polygons, file size of the mesh, and accuracy of performing measurements on the 3D model. Evaluation of the polygon number was performed for each model using Blender (Version 3.1.0, https://<ulink href="http://www.blender.org">www.blender.org</ulink>).</p> <p>Texture maps are two‐dimensional images applied to the surface of 3D models to simulate color and pattern details. Following the scanning of an object, all visual information of the colored surface details is compiled into a single file (often.png or.jpg). The final resolution and visual clarity of individual visual components within the texture map correspond to the resolution of the 3D model, which is then overlaid onto the mesh to create the final 3D model.</p> <hd id="AN0186371264-13">Survey evaluation of 3D models</hd> <p>To evaluate the visual quality and resolution of the 3D models, we surveyed residents within our institution's Department of Neurologic Surgery using an independent, unsponsored 6‐question survey in Qualtrics, in which models were ranked in order and blinded to the technique utilized—first, second, and third—based on which technique captured the best visual quality as specified by multiple parameters including best lighting, fine details, deep details, coloration, etc. Within the survey, residents compared each of the three models to photographs captured from identical perspectives, enabling standardized visual comparison across the three models (Figure 4). No incentives were provided for completion of the survey. Data from the survey was collected at a single timepoint.</p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/8Z8K/01jul25/ase70062-fig-0004.jpg?ephost1=dGJyMNXb4kSepq84yOvqOLCmsE6epq5Srqa4SK6WxWXS" alt="ase70062-fig-0004.jpg" title="4 Images included within the survey to assess the visual quality of 3D models. Sets of images include those obtained by 2D photography (A–C) and 3D models captured using "manual" photogrammetry (D–F), "automatic" photogrammetry (G–I), and structured light 3D scanning (J–L). Note that all 2D photos and 3D models are in their unedited form. Additionally, photos of 3D models have been acquired from the SketchFab® website, with all models uploaded to the website using the same settings." /> </p> <p></p> <hd id="AN0186371264-15">3D model processing</hd> <p>While there are multiple options for model processing and editing to further enhance the quality of the mesh and texture map (a process that can markedly vary based on user skillset and experience with graphic design), we sought to compare solely the baseline characteristics of 3D models without any processing beyond what is minimally required to create the model. As such, following the rendering of the 3D models using each of the techniques described above, all models remained unedited with no further adjustments to color, contrast, or artifact removal. Additionally, model meshes remained the same as their original rendering without the decimation (reduction) of polygons. While 3D scanners automatically capture information used to generate models with real, scaled dimensions (for the purposes of performing measurements), photogrammetric models necessitate calibration using reference marker(s)—a process which is optional in the creation of 3D models via photogrammetry, and therefore was not utilized in this study. Finally, to ensure that our three models could be uniformly compared through visual inspection, the models were all uploaded to the SketchFab® website (Epic, Cary, NC) and utilized the same display settings.</p> <hd id="AN0186371264-16">RESULTS</hd> <p>The results of our 3D model comparison are summarized in Table 1. Regarding the human time and total time required for the entire 3D model generation process, the "manual" photogrammetry approach necessitated the longest duration while "automatic" photogrammetry was the least time‐intensive 3D scanning technique (180 vs. 30 min and 240 vs. 140 min, respectively). Similarly, "manual" photogrammetry required the most amount of technological experience and expertise in order to generate a 3D model from individual photos, while the 3D rendering process is automatically built into the "automatic" photogrammetry, therefore requiring little to no human effort for rendering a 3D model. For using structured light 3D scanning, some basic knowledge of the Artec Studio 17 software is required to combine data from multiple scans; however, the remainder of the 3D model process is relatively automated. The costs of equipment required for each technique varied greatly, ranging from 80,000 USD for automatic photogrammetry to 26,000 USD for the structured light 3D scanning, and manual photogrammetry requiring the smallest financial investment of 5000 USD.</p> <p>1 TABLE Summary of techniques comparison.</p> <p> <ephtml> <table><thead valign="bottom"><tr><th align="left">Parameter</th><th align="left">Manual photogrammetry<xref ref-type="fn" rid="tfn1" /></th><th align="left">Automatic photogrammetry<xref ref-type="fn" rid="tfn2" /></th><th align="left">Structured light 3D scanning<xref ref-type="fn" rid="tfn3" /></th></tr></thead><tbody valign="top"><tr><td align="left">Human time required (min)</td><td align="left">180</td><td align="left">30</td><td align="left">120</td></tr><tr><td align="left">Total time required (min)</td><td align="left">240</td><td align="left">140</td><td align="left">180</td></tr><tr><td align="left">Technological knowledge required</td><td align="left">+++</td><td align="left">+</td><td align="left">++</td></tr><tr><td align="left">Financial cost (USD)</td><td align="left">$5000</td><td align="left">$80,000</td><td align="left">$26,000</td></tr><tr><td align="left">Polygons (#)</td><td align="left">250,000</td><td align="left">4,883,284</td><td align="left">3,735,318</td></tr><tr><td align="left">Mesh file size (Mb)</td><td align="left">14</td><td align="left">530</td><td align="left">309</td></tr><tr><td align="left">Ability to perform measurements</td><td align="left">−</td><td align="left">−</td><td align="left">+</td></tr><tr><td align="left">Texture resolution (pixels)</td><td align="left">8192 × 8192</td><td align="left">8192 × 8192</td><td align="left">8192 × 8192</td></tr></tbody></table> </ephtml> </p> <p>1 a Sketchfab® link: https://sketchfab.com/3d‐models/optic‐pathway‐manual‐photogrammetry‐650a17efe2c64c03ba0f8862711e0c80.</p> <ulist> <item>2 b Sketchfab® link: https://sketchfab.com/3d‐models/optic‐pathway‐automatic‐photogrammetry‐08b83e4614e340edb30ad1ab67ed32d0.</item> <item>3 c Sketchfab® link: https://sketchfab.com/3d‐models/optic‐pathway‐structured‐light‐scan‐d6f45f0ca0184c6a9cd7b391fa5d73cd.</item> </ulist> <p>"Automatic" photogrammetry resulted in the highest polygon count and subsequently the largest file size when compared to "manual" photogrammetry and structured light 3D scanning. As previously described, photogrammetric models necessitate a reference marker for accurate scaling to real dimensions, whereas 3D scanners used in this article are explicitly designed for precise measurements, with an accuracy of approximately 0.05 mm for the structured scans. Therefore, in using baseline 3D models without any additional scaling of the models or external validation, only structured light 3D scanning is capable of performing accurate and precise measurements.</p> <p>The same texture resolution (8192 × 8192 pixels) was achieved using all three 3D scanning techniques. However, results from the survey with 20 respondents (Table 2) demonstrated the 3D model generated with "manual" photogrammetry possessed the best fine details, while the 3D model acquired using structured light 3D scanning had the poorest visual quality. Specifically, 100% of respondents considered "manual" photogrammetry to possess the best publication quality. Differences between the models were noted in the representation of surface details and very deep structures, as shown by 89.5% and 84.2% of the participants preferring the "manual" photogrammetry 3D model for these respective features. In addition, light and coloration of the models was best captured in the "manual" photogrammetry according to 100% and 89.5% of respondents, respectively. Overall, "automatic" photogrammetry ranked second for all parameters and structured light 3D scanning ranked third for all parameters, except for 'most similar to professional 2D photographs' in which the two techniques were equivalent.</p> <p>2 TABLE Survey results (n  = 20 responses) comparing the visual quality of 3D models as ranked by their first, second, and third choices.</p> <p> <ephtml> <table><thead valign="bottom"><tr><th align="left">Survey question</th><th align="left">Most common answer (percentage)</th></tr></thead><tbody valign="top"><tr><td align="left">Best visibility of fine details</td><td align="left">First choice: Manual photogrammetry (89.5%)Second choice: Automatic photogrammetry (94.7%)Third choice: Structured light 3D scanning (84.2%)</td></tr><tr><td align="left">Best lighting</td><td align="left">First choice: Manual photogrammetry (100%)Second choice: Automatic photogrammetry (88.9%)Third choice: Structured light 3D scanning (88.9%)</td></tr><tr><td align="left">Best coloration</td><td align="left">First choice: Manual photogrammetry (89.5%)Second choice: Automatic photogrammetry (52.6%)Third choice: Structured light 3D scanning (52.6%)</td></tr><tr><td align="left">Best visualization of deep structures</td><td align="left">First choice: Manual photogrammetry (84.2%)Second choice: Automatic photogrammetry (57.9%)Third choice: Structured light 3D scanning (63.2%)</td></tr><tr><td align="left">Best "publication quality"</td><td align="left">First choice: Manual photogrammetry (100%)Second choice: Automatic photogrammetry (63.2%)Third choice: Structured light 3D scanning (63.2%)</td></tr><tr><td align="left">Most similar to professional 2D photographs</td><td align="left">First choice: Manual photogrammetry (73.7%)Second choice: Automatic photogrammetry and structured light 3D scanning (52.6%)</td></tr></tbody></table> </ephtml> </p> <p>Given our experience with these three 3D scanning techniques, we have summarized the overall benefits, limitations, and suggested uses for each scanning technique in Table 3.</p> <p>3 TABLE Summary of overall benefits and limitations for each 3D scanning technique.</p> <p> <ephtml> <table><thead valign="bottom"><tr><th align="left">Parameter</th><th align="left">Manual photogrammetry</th><th align="left">Automatic photogrammetry</th><th align="left">Structured light 3D scanning</th></tr></thead><tbody valign="top"><tr><td align="left">Benefit(s)</td><td align="left"><list list-type="Bullet"><list-item><p>Allows for creation of high‐quality 3D models</p></list-item><list-item><p>Able to capture fine details and structures presents in a given dissection</p></list-item><list-item><p>Most cost‐effective strategy</p></list-item></list></td><td align="left"><list list-type="Bullet"><list-item><p>Fastest process that requires the least amount of human effort</p></list-item><list-item><p>Does not require as much technological knowledge given automation of most processes</p></list-item></list></td><td align="left"><list list-type="Bullet"><list-item><p>Models are automatically scaled to size during the scanning process without the need for any additional procedures</p></list-item><list-item><p>Measurements can be taken with the models and have consistency/accuracy given fixed scale</p></list-item></list></td></tr><tr><td align="left">Limitation(s)</td><td align="left"><list list-type="Bullet"><list-item><p>Time‐intensive process</p></list-item><list-item><p>Requires the most knowledge of 3D model rendering software</p></list-item></list></td><td align="left"><list list-type="Bullet"><list-item><p>Most expensive option</p></list-item><list-item><p>Automated process may result in artifacts within the model</p></list-item></list></td><td align="left"><list list-type="Bullet"><list-item><p>Has significant cost associated</p></list-item><list-item><p>Poor quality of texture map and model surface does not allow for accurate representation of colored texture</p></list-item></list></td></tr><tr><td align="left">Optimal use of scanning technique</td><td align="left"><list list-type="Bullet"><list-item><p>Production of high‐quality 3D models for educational or publication purposes</p></list-item><list-item><p>Suited for neuroanatomical dissections that contain fine details</p></list-item></list></td><td align="left"><list list-type="Bullet"><list-item><p>High‐volume production of 3D models</p></list-item><list-item><p>Suited for 3D models that contain basic structures without fine details</p></list-item></list></td><td align="left"><list list-type="Bullet"><list-item><p>3D models to be used for measurements of structures</p></list-item><list-item><p>Suited for 3D models that do not need high‐fidelity reproduction of colored surface texture</p></list-item></list></td></tr></tbody></table> </ephtml> </p> <hd id="AN0186371264-17">DISCUSSION</hd> <p>As technology advances, new methodologies emerge to bridge the gap between 2D and 3D object representation. With the increasing accessibility and ease of production of 3D models,[<reflink idref="bib20" id="ref18">20</reflink>] their application in neuroanatomical education and neurosurgery is expanding. In educational contexts, 3D models serve as potent tools, enabling learners to grasp complex 3D relationships among neuroanatomical structures—a feat unattainable through traditional 2D atlases. These models facilitate a deeper understanding of fundamental neuroanatomical relationships and have been further documented as step‐by‐step guides of anatomical dissection.[[<reflink idref="bib21" id="ref19">21</reflink>], [<reflink idref="bib23" id="ref20">23</reflink>]] Moreover, with the availability of free 3D model platforms like Sketchfab®, the sole prerequisite for utilizing these models is access to a phone, tablet, or computer, thereby democratizing neuroanatomical education and broadening its reach to resource‐constrained regions worldwide.[<reflink idref="bib25" id="ref21">25</reflink>] 3D models have also been integrated into virtual reality and augmented reality environments to aid in planning intricate procedures and approaches, along with learner education.[[<reflink idref="bib3" id="ref22">3</reflink>], [<reflink idref="bib25" id="ref23">25</reflink>], [<reflink idref="bib27" id="ref24">27</reflink>], [<reflink idref="bib29" id="ref25">29</reflink>]] Furthermore, their utility extends to the realm of clinical applications through use in spine deformities[[<reflink idref="bib12" id="ref26">12</reflink>], [<reflink idref="bib30" id="ref27">30</reflink>]] and the treatment of pediatric conditions such as craniosynostosis and congenital craniofacial abnormalities.[[<reflink idref="bib11" id="ref28">11</reflink>], [<reflink idref="bib31" id="ref29">31</reflink>]]</p> <p>Through characterizing and comparing three different 3D scanning techniques—"manual" photogrammetry with one camera, "automatic" photogrammetry with a 5‐camera machine, and structured light 3D scanning—we have found certain benefits and limitations of each technique that have important implications for their ideal application within anatomical education and neurosurgery. Firstly, although we found similar pixel resolution between each technique, the visual quality of the resultant models was markedly different, with "manual" photogrammetry displaying the best visibility of fine and deep structures, lighting, and coloration, as determined by our survey results. Importantly, this technique was unanimously considered to possess the best "publication quality" by our survey respondents. Particularly, substantial differences between these models were noted in the representation of surface details and very deep structures, some of which are blurred or missing in "automatic" photogrammetry and structured light 3D scanning models compared to the model generated with "manual" photogrammetry.</p> <p>However, it is worth noting that this 3D scanning technique also resulted in the lowest polygon count. While the number of polygons is important for creating the most accurate depiction of model shape and structure, we found that for the purpose of anatomical representation, the number of polygons does not necessarily affect the visual minutia of a model. Therefore, one can potentially minimize the number of polygons (within reason), yet still produce a visually appealing model.[<reflink idref="bib33" id="ref30">33</reflink>] This can ultimately reduce the time necessary for model creation and upload onto 3D model viewing platforms (such as Sketchfab®), thereby further enhancing the user experience on such platforms.</p> <p>Importantly, this principle underscores that prioritizing optimal image quality for the texture map outweighs the significance of polygon count when aiming for the most precise visual representation of fine details and structures in a 3D model. However, this raises the dilemma that, despite both our "manual" and "automatic" photogrammetry techniques utilizing high‐quality DSLR cameras, these methods yielded notably distinct models upon visual inspection. While the 3D model produced with "automatic" photogrammetry exhibits superior image quality compared to structured light 3D scanning, it still falls short of the standard set by the "manual" photogrammetry model, despite the use of five cameras versus one. It is crucial to note that "manual" photogrammetry offers the flexibility to selectively photograph specific structures and adjust distances, whereas "automatic" photogrammetry treats all object structures uniformly, capturing images at fixed intervals and distances. Consequently, the image quality of smaller and deeper structures obtained through "automatic" photogrammetry may deteriorate upon zooming into the model, whereas "manual" photogrammetry maintains integrity. Nonetheless, the quality of the DSLR cameras allowed for relatively accurate and detailed reproduction of the specimen structure and exceeded the quality attained in structured light 3D scanning.</p> <p>While "automatic" photogrammetry may not be as adept at producing highly detailed 3D models as "manual" photogrammetry, it still holds the advantage of automation, requiring minimal user time and input, thus enabling efficient scanning and 3D model creation. This is particularly advantageous for scenarios in which there is a need for high‐volume production of 3D models—that is, capturing pathological processes in autopsy specimens—that may not require as much fine detail as those used in an educational setting.</p> <p>Within clinical applications, the structural integrity of the mesh holds significant implications, particularly when obtaining measurements based on 3D models. In clinical practice, the utilization of 3D models to assess postoperative outcomes in spine and congenital craniofacial surgeries has predominantly relied on photogrammetry techniques.[[<reflink idref="bib11" id="ref31">11</reflink>], [<reflink idref="bib30" id="ref32">30</reflink>], [<reflink idref="bib32" id="ref33">32</reflink>]] Conversely, structured light 3D scanning has demonstrated its value in 3D model applications outside of neurosurgery—for example, prosthesis reconstruction in orthopedic surgery.[[<reflink idref="bib34" id="ref34">34</reflink>]] Apart from cost considerations, structured light 3D scanning presents a notable advantage in performing measurements, as distances are automatically generated during the scanning process with a set accuracy of 0.05 mm, unlike photogrammetry, which necessitates fiducial markers. This technique has also been demonstrated to capture more accurate geometries of complex specimens than photogrammetry.[<reflink idref="bib36" id="ref35">36</reflink>] Further, this technique offers the advantage of being handheld and portable, thereby having less size limitations than other photogrammetry techniques. However, limitations may arise in capturing deep structures of interest, posing challenges for structured light 3D scanning. Therefore, for practices prioritizing measurement accuracy and structural reconstruction, the cost of structured light 3D scanning may be justified by the precision and consistency it offers.</p> <p>Alongside the technique outlined in our study, we acknowledge that 3D scanning has become increasingly accessible with the integration of 3D technologies into smartphones,[<reflink idref="bib37" id="ref36">37</reflink>] eliminating the need for additional camera equipment. Moreover, the portable nature of smartphones allows for photogrammetry to be conducted in virtually any setting, as operators can maneuver around the object of interest without requiring a turntable or stand. Similarly, photogrammetric techniques have been adapted for use in operative settings, where overlapping pictures from surgical footage have been utilized to generate 3D models, particularly in skull base procedures.[<reflink idref="bib38" id="ref37">38</reflink>] This marks a significant advancement in 3D model acquisition, as operative footage can serve educational and documentation purposes without the need for additional resources in the operating room to perform 3D capture. While we still prefer to use "manual" photogrammetry in our practice to capture the finest detail of a specimen, the ongoing advancements in smartphone cameras and other photography systems with each subsequent iteration have the potential to lead to increased accessibility of 3D models, though additional studies are needed to assess the quality of such techniques.</p> <hd id="AN0186371264-18">Limitations and future directions</hd> <p>While our study aimed to compare 3D models using three distinct scanning techniques, it is crucial to acknowledge that the final 3D models represent only 'baseline' models without any additional manipulation or editing. Several commercially available 3D model software packages facilitate the editing of such models, enabling tasks such as color changes, addition or removal of structures, shaping, and scaling models, among other functions. However, as these tasks become increasingly complex, requiring sophisticated manipulation of 3D models, the technological expertise needed to operate the appropriate software grows exponentially. Eventually, it may become necessary to collaborate with dedicated medical illustrators and artists to achieve desired outcomes. Nevertheless, if available to an institution, such individuals are invaluable assets for maximizing the potential and quality of 3D models produced with any scanning technique. For example, artifacts such as green pins accidentally incorporated into "automatic" photogrammetry can be removed from the final model using editing software. Moreover, issues such as deep shadows resulting from inadequate lighting in "automatic photogrammetry" or lack of colored contrast in structured light 3D scanning can be adjusted to get closer to a more accurate representation of the original object. Additionally, 3D models acquired through both "manual" and "automatic" photogrammetry can be scaled to size for measuring discrete structures or distances. Therefore, there exists significant potential for enhancing the quality of the 'baseline' 3D models represented in our study through additional technological manipulation and, potentially, the employment of personnel with expertise in 3D editing techniques.</p> <p>Regarding the survey assessment of our 3D models, we aimed to directly compare the models to capture which 3D modeling technique has the best visual quality and resolution. However, our survey does not capture the degree of quality for each technique in comparison to other techniques. Therefore, we recommend further studies using more robust, Likert scale format surveys—with tested validity and reliability—to provide more detailed insights about the degree of differences between 3D modeling techniques.</p> <hd id="AN0186371264-19">CONCLUSION</hd> <p>Multiple methods exist for obtaining neuroanatomical 3D models based on the desired quality of the end product and intended use of the 3D model. In our experience, "manual" photogrammetry allows for the most detailed 3D models and is the most cost‐effective strategy, though there is a considerable time component to using this technique. While fine textured details may not be ideally captured with structured light 3D scanning, this technique is most suitable for obtaining clinically relevant measurements given the high degree of accuracy in capturing the model shape (i.e., mesh). Finally, "automatic" photogrammetry represents a hybrid between obtaining relatively high‐quality models in a time‐effective and user‐friendly manner, though this method does require a significant financial investment. Regardless of the technique employed, the flexibility and detailed quality of neuroanatomical 3D models make them a powerful tool in both the educational and clinical settings.</p> <hd id="AN0186371264-20">AUTHOR CONTRIBUTIONS</hd> <p> <bold>Megan M. J. Bauman:</bold> Conceptualization; methodology; data curation; investigation; formal analysis; visualization; writing – original draft; writing – review and editing. <bold>Amedeo Piazza:</bold> Conceptualization; methodology; data curation; investigation; formal analysis; visualization; writing – review and editing; writing – original draft. <bold>Fabio Torregrossa:</bold> Methodology; visualization; writing – review and editing; investigation. <bold>Charles Wes Price:</bold> Methodology; investigation; visualization; writing – review and editing. <bold>Jonathan M. Morris:</bold> Supervision; writing – review and editing. <bold>Luciano C. P. C. Leonel:</bold> Resources; supervision; writing – review and editing. <bold>Maria Peris‐Celda:</bold> Conceptualization; writing – review and editing; supervision; investigation; visualization.</p> <hd id="AN0186371264-21">ACKNOWLEDGMENTS</hd> <p>The authors wish to thank the generosity of the families and the body donors who generously donated their bodies to the Mayo Clinic Body Donation Program in the Department of Clinical Anatomy, MN, USA. They were essential for carrying out this research.</p> <hd id="AN0186371264-22">FUNDING INFORMATION</hd> <p>The Joseph I. and Barbara Ashkins Endowed Professorship in Neurosurgery and the Charles B. and Ann L. Johnson Endowed Professorship in Neurosurgery, Department of Neurosurgery, Mayo Clinic, Rochester, Minnesota.</p> <hd id="AN0186371264-23">ETHICS STATEMENT</hd> <p>The study was conducted in accordance with the Declaration of Helsinki, and the protocol was approved by our Institutional Review Board.</p> <ref id="AN0186371264-24"> <title> Footnotes </title> <blist> <bibl id="bib1" idref="ref1" type="bt">1</bibl> <bibtext> Megan M. J. Bauman and Amedeo Piazza contributed equally to this study and are both 1st co‐authors.</bibtext> </blist> </ref> <ref id="AN0186371264-25"> <title> REFERENCES </title> <blist> <bibtext> Sotgiu MA, Mazzarello V, Bandiera P, Madeddu R, Montella A, Moxham B. Neuroanatomy, the Achille's heel of medical students. A systematic analysis of educational strategies for the teaching of neuroanatomy. Anat Sci Educ. 2020 ; 13 (1): 107 – 116.</bibtext> </blist> <blist> <bibl id="bib2" idref="ref2" type="bt">2</bibl> <bibtext> Graffeo CS, Peris‐Celda M, Perry A, Carlstrom LP, Driscoll CLW, Link MJ. Anatomical step‐by‐step dissection of complex skull base approaches for trainees: surgical anatomy of the retrosigmoid approach. J Neurol Surg B Skull Base. 2021 ; 82 (3): 321 – 332.</bibtext> </blist> <blist> <bibl id="bib3" idref="ref22" type="bt">3</bibl> <bibtext> Oliveira ASB, Leonel L, LaHood ER, Nguyen BT, Ehtemami A, Graepel SP, et al. Projection of realistic three‐dimensional photogrammetry models using stereoscopic display: a technical note. Anat Sci Educ. 2024 ; 17 (1): 39 – 46.</bibtext> </blist> <blist> <bibl id="bib4" idref="ref3" type="bt">4</bibl> <bibtext> Vigo V, Asmaro KP, Nunez MA, Fernandez‐Miranda JC. Fronto‐orbitozygomatic approach for cavernous sinus hemangioma: 3‐dimensional operative video. Oper Neurosurg (Hagerstown). 2023 ; 25 (2): e80 – e81.</bibtext> </blist> <blist> <bibl id="bib5" type="bt">5</bibl> <bibtext> Vigo V, Asmaro KP, Nunez MA, Bobrow A, Dodd RL, Desai A, et al. Extreme far‐lateral approach for recurrent chordoma: 3‐dimensional operative video. Oper Neurosurg (Hagerstown). 2023 ; 24 (5): e365.</bibtext> </blist> <blist> <bibl id="bib6" idref="ref4" type="bt">6</bibl> <bibtext> Graffeo CS, Perry A, Carlstrom LP, Peris‐Celda M, Alexander A, Dickens HJ, et al. 3D printing for complex cranial surgery education: technical overview and preliminary validation study. J Neurol Surg B Skull Base. 2022 ; 83 (Suppl 2): e105 – e112.</bibtext> </blist> <blist> <bibl id="bib7" type="bt">7</bibl> <bibtext> Piazza A, Petrella G, Corvino S, Campione A, Campeggi A, Serioli S, et al. 3‐dimensionally printed affordable nose model: a reliable start in endoscopic training for young neurosurgeons. World Neurosurg. 2023 ; 180 : 17 – 21.</bibtext> </blist> <blist> <bibl id="bib8" idref="ref5" type="bt">8</bibl> <bibtext> de Oliveira ASB, Leonel L, LaHood ER, Hallak H, Link MJ, Maleszewski JJ, et al. Foundations and guidelines for high‐quality three‐dimensional models using photogrammetry: a technical note on the future of neuroanatomy education. Anat Sci Educ. 2023 ; 16 (5): 870 – 883.</bibtext> </blist> <blist> <bibl id="bib9" idref="ref9" type="bt">9</bibl> <bibtext> Xu Y, Vigo V, Klein J, Nunez MA, Fernandez‐Miranda JC, Cohen‐Gadol AA, et al. Pursuing perfect 2D and 3D photography in neuroanatomy: a new paradigm for staying up to date with digital technology. J Neurosurg. 2022 ; 138 : 1766 – 1772.</bibtext> </blist> <blist> <bibtext> Tse R, Booth L, Keys K, Saltzman B, Stuhaug E, Kapadia H, et al. Reliability of nasolabial anthropometric measures using three‐dimensional stereophotogrammetry in infants with unrepaired unilateral cleft lip. Plast Reconstr Surg. 2014 ; 133 (4): 530e – 542e.</bibtext> </blist> <blist> <bibtext> Elkhill C, Liu J, Linguraru MG, LeBeau S, Khechoyan D, French B, et al. Geometric learning and statistical modeling for surgical outcomes evaluation in craniosynostosis using 3D photogrammetry. Comput Methods Prog Biomed. 2023 ; 240 : 107689.</bibtext> </blist> <blist> <bibtext> Janjua MB, Zhou PL, Vasquez‐Montes D, Moskovich R. Photogrammetric analysis: an objective measure to assess the craniocervical range of motion after cervical laminoplasty surgeries. J Clin Neurosci. 2020 ; 71 : 76 – 83.</bibtext> </blist> <blist> <bibtext> Shimizu F, Uehara M, Oatari M, Kusatsu M. Three‐dimensional visualization of the human face using DICOM data and its application to facial contouring surgery using free anterolateral thigh flap transfer. J Plast Reconstr Aesthet Surg. 2016 ; 69 (1): e1 – e4.</bibtext> </blist> <blist> <bibtext> Gaweł D, Danielewicz K, Nowak M. The method of geometrical comparison of 3‐dimensional objects created from DICOM images. Stud Health Technol Inform. 2012 ; 176 : 273 – 276.</bibtext> </blist> <blist> <bibtext> Piazza A, Spiriev T, Corvino S, Corrivetti F, Laleva L, Iaconetta G, et al. The course of the trochlear nerve presented via a 3‐dimensional photorealistic anatomic model. World Neurosurg. 2024 ; 186 : e156 – e160.</bibtext> </blist> <blist> <bibtext> Seminati E, Canepa Talamas D, Young M, Twiste M, Dhokia V, Bilzon JLJ. Validity and reliability of a novel 3D scanner for assessment of the shape and volume of amputees' residual limb models. PLoS One. 2017 ; 12 (9): e0184498.</bibtext> </blist> <blist> <bibtext> Franco de Sá Gomes C, Libdy MR, Normando D. Scan time, reliability and accuracy of craniofacial measurements using a 3D light scanner. J Oral Biol Craniofac Res. 2019 ; 9 (4): 331 – 335.</bibtext> </blist> <blist> <bibtext> Leonel LCP, Carlstrom LP, Graffeo CS, Perry A, Pinheiro‐Neto CD, Sorenson J, et al. Foundations of advanced neuroanatomy: technical guidelines for specimen preparation, dissection, and 3D‐photodocumentation in a surgical anatomy laboratory. J Neurol Surg B Skull Base. 2021 ; 82 (Suppl 3): e248 – e258.</bibtext> </blist> <blist> <bibtext> Oliveira A, Leonel L, Bauman M, Oliveira ASB, Leonel LCPC, Bauman MMJ, et al. Photogrammetry scans for neuroanatomy education—a new multi‐camera system: technical note. Neuroinformatics. 2024 ; 22 : 317 – 327.</bibtext> </blist> <blist> <bibtext> Corvino S, Piazza A, Spiriev T, Tafuto R, Corrivetti F, Solari D, et al. The sellar region as seen from transcranial and endonasal perspectives: exploring bony landmarks through new surface photorealistic three‐dimensional model reconstruction for neurosurgical anatomy training. World Neurosurg. 2024 ; 185 : e367 – e375.</bibtext> </blist> <blist> <bibtext> Krogager ME, Fugleholm K, Mathiesen TI, Spiriev T. Simplified easy‐accessible smartphone‐based photogrammetry for 3‐dimensional anatomy presentation exemplified with a photorealistic cadaver‐based model of the intracranial and Extracranial course of the facial nerve. Oper Neurosurg (Hagerstown). 2023 ; 25 (2): e71 – e77.</bibtext> </blist> <blist> <bibtext> Spiriev T, Nakov V, Cornelius JF. Photorealistic 3‐dimensional models of the anatomy and neurosurgical approaches to the V2, V3, and V4 segments of the vertebral artery. Oper Neurosurg (Hagerstown). 2023 ; 25 (1): e15 – e21.</bibtext> </blist> <blist> <bibtext> De Benedictis A, Nocerino E, Menna F, Remondino F, Barbareschi M, Rozzanigo U, et al. Photogrammetry of the human brain: a novel method for three‐dimensional quantitative exploration of the structural connectivity in neurosurgery and neurosciences. World Neurosurg. 2018 ; 115 : e279 – e291.</bibtext> </blist> <blist> <bibtext> Hanalioglu S, Romo NG, Mignucci‐Jiménez G, Tunc O, Gurses ME, Abramov I, et al. Development and validation of a novel methodological pipeline to integrate neuroimaging and photogrammetry for immersive 3D cadaveric neurosurgical simulation. Front Surg. 2022 ; 9 : 878378.</bibtext> </blist> <blist> <bibtext> Aydin SO, Barut O, Yilmaz MO, Sahin B, Akyoldas G, Akgun MY, et al. Use of 3‐dimensional modeling and augmented/virtual reality applications in microsurgical neuroanatomy training. Oper Neurosurg (Hagerstown). 2023 ; 24 (3): 318 – 323.</bibtext> </blist> <blist> <bibtext> Gurses ME, Gonzalez‐Romo NI, Xu Y, Mignucci‐Jiménez G, Hanalioglu S, Chang JE, et al. Interactive microsurgical anatomy education using photogrammetry 3D models and an augmented reality cube. J Neurosurg. 2024 ; 141 : 1 – 10.</bibtext> </blist> <blist> <bibtext> Gurses ME, Hanalioglu S, Mignucci‐Jiménez G, Gökalp E, Gonzalez‐Romo NI, Gungor A, et al. Three‐dimensional modeling and extended reality simulations of the cross‐sectional anatomy of the cerebrum, cerebellum, and brainstem. Oper Neurosurg (Hagerstown). 2023 ; 25 (1): 3 – 10.</bibtext> </blist> <blist> <bibtext> Gurses ME, Gungor A, Rahmanov S, Gökalp E, Hanalioglu S, Berker M, et al. Three‐dimensional modeling and augmented reality and virtual reality simulation of fiber dissection of the cerebellum and brainstem. Oper Neurosurg (Hagerstown). 2022 ; 23 (5): 345 – 354.</bibtext> </blist> <blist> <bibtext> Gurses ME, Gungor A, Gökalp E, Hanalioglu S, Karatas Okumus SY, Tatar I, et al. Three‐dimensional modeling and augmented and virtual reality simulations of the white matter anatomy of the cerebrum. Oper Neurosurg (Hagerstown). 2022 ; 23 (5): 355 – 366.</bibtext> </blist> <blist> <bibtext> Pedersen PH, Vergari C, Alzakri A, Vialle R, Skalli W. A reduced micro‐dose protocol for 3D reconstruction of the spine in children with scoliosis: results of a phantom‐based and clinically validated study using stereo‐radiography. Eur Radiol. 2019 ; 29 (4): 1874 – 1881.</bibtext> </blist> <blist> <bibtext> Choi TM, Liu X, Abdel‐Alim T, van Veelen M, Mathijssen IMJ, Wolvius EB, et al. Automated three‐dimensional analysis of facial asymmetry in patients with syndromic coronal synostosis: a retrospective study. J Craniomaxillofac Surg. 2024 ; 52 (1): 48 – 54.</bibtext> </blist> <blist> <bibtext> Abdel‐Alim T, Kurniawan M, Mathijssen I, Dremmen M, Dirven C, Niessen W, et al. Sagittal craniosynostosis: comparing surgical techniques using 3D photogrammetry. Plast Reconstr Surg. 2023 ; 152 (4): 675e – 688e.</bibtext> </blist> <blist> <bibtext> Vandenbossche V, Van de Velde J, Avet S, Willaert W, Soltvedt S, Smit N, et al. Digital body preservation: technique and applications. Anat Sci Educ. 2022 ; 15 (4): 731 – 744.</bibtext> </blist> <blist> <bibtext> Yu F, Zeng L, Pan D, Sui X, Tang J. Evaluating the accuracy of hand models obtained from two 3D scanning techniques. Sci Rep. 2020 ; 10 (1): 11875.</bibtext> </blist> <blist> <bibtext> Paoli A, Neri P, Razionale AV, Tamburrino F, Barone S. Sensor architectures and technologies for upper limb 3D surface reconstruction: a review. Sensors (Basel). 2020 ; 20 (22): 6584.</bibtext> </blist> <blist> <bibtext> Dixit I, Kennedy S, Piemontesi J, Kennedy B, Krebs C. Which tool is best: 3D scanning or photogrammetry—it depends on the task. Adv Exp Med Biol. 2019 ; 1120 : 107 – 119.</bibtext> </blist> <blist> <bibtext> Quispe‐Enriquez OC, Valero‐Lanzuela JJ, Lerma JL. Smartphone photogrammetric assessment for head measurements. Sensors (Basel). 2023 ; 23 (21): 9008.</bibtext> </blist> <blist> <bibtext> Gonzalez‐Romo NI, Hanalioglu S, Mignucci‐Jiménez G, Abramov I, Xu Y, Preul MC. Anatomic depth estimation and 3‐dimensional reconstruction of microsurgical anatomy using monoscopic high‐definition photogrammetry and machine learning. Oper Neurosurg (Hagerstown). 2023 ; 24 (4): 432 – 444.</bibtext> </blist> </ref> <aug> <p>By Megan M. J. Bauman; Amedeo Piazza; Fabio Torregrossa; Charles Wes Price; Jonathan M. Morris; Luciano C. P. C. Leonel and Maria Peris‐Celda</p> <p>Reported by Author; Author; Author; Author; Author; Author; Author</p> </aug> <nolink nlid="nl1" bibid="bib10" firstref="ref6"></nolink> <nolink nlid="nl2" bibid="bib11" firstref="ref7"></nolink> <nolink nlid="nl3" bibid="bib13" firstref="ref8"></nolink> <nolink nlid="nl4" bibid="bib15" firstref="ref10"></nolink> <nolink nlid="nl5" bibid="bib16" firstref="ref11"></nolink> <nolink nlid="nl6" bibid="bib18" firstref="ref14"></nolink> <nolink nlid="nl7" bibid="bib19" firstref="ref17"></nolink> <nolink nlid="nl8" bibid="bib20" firstref="ref18"></nolink> <nolink nlid="nl9" bibid="bib21" firstref="ref19"></nolink> <nolink nlid="nl10" bibid="bib23" firstref="ref20"></nolink> <nolink nlid="nl11" bibid="bib25" firstref="ref21"></nolink> <nolink nlid="nl12" bibid="bib27" firstref="ref24"></nolink> <nolink nlid="nl13" bibid="bib29" firstref="ref25"></nolink> <nolink nlid="nl14" bibid="bib12" firstref="ref26"></nolink> <nolink nlid="nl15" bibid="bib30" firstref="ref27"></nolink> <nolink nlid="nl16" bibid="bib31" firstref="ref29"></nolink> <nolink nlid="nl17" bibid="bib33" firstref="ref30"></nolink> <nolink nlid="nl18" bibid="bib32" firstref="ref33"></nolink> <nolink nlid="nl19" bibid="bib34" firstref="ref34"></nolink> <nolink nlid="nl20" bibid="bib36" firstref="ref35"></nolink> <nolink nlid="nl21" bibid="bib37" firstref="ref36"></nolink> <nolink nlid="nl22" bibid="bib38" firstref="ref37"></nolink>
Header DbId: eric
DbLabel: ERIC
An: EJ1476155
AccessLevel: 3
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: A Comprehensive Guide to High-Fidelity 3D Neuroanatomical Modeling Techniques: A Quantitative Comparison between Photogrammetry and Structured Light Scanning
– Name: Language
  Label: Language
  Group: Lang
  Data: English
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Megan+M%2E+J%2E+Bauman%22">Megan M. J. Bauman</searchLink> (ORCID <externalLink term="https://orcid.org/0000-0003-1910-6617">0000-0003-1910-6617</externalLink>)<br /><searchLink fieldCode="AR" term="%22Amedeo+Piazza%22">Amedeo Piazza</searchLink><br /><searchLink fieldCode="AR" term="%22Fabio+Torregrossa%22">Fabio Torregrossa</searchLink><br /><searchLink fieldCode="AR" term="%22Charles+Wes+Price%22">Charles Wes Price</searchLink><br /><searchLink fieldCode="AR" term="%22Jonathan+M%2E+Morris%22">Jonathan M. Morris</searchLink><br /><searchLink fieldCode="AR" term="%22Luciano+C%2E+P%2E+C%2E+Leonel%22">Luciano C. P. C. Leonel</searchLink><br /><searchLink fieldCode="AR" term="%22Maria+Peris-Celda%22">Maria Peris-Celda</searchLink>
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="SO" term="%22Anatomical+Sciences+Education%22"><i>Anatomical Sciences Education</i></searchLink>. 2025 18(7):697-708.
– Name: Avail
  Label: Availability
  Group: Avail
  Data: Wiley. Available from: John Wiley & Sons, Inc. 111 River Street, Hoboken, NJ 07030. Tel: 800-835-6770; e-mail: cs-journals@wiley.com; Web site: https://www.wiley.com/en-us
– Name: PeerReviewed
  Label: Peer Reviewed
  Group: SrcInfo
  Data: Y
– Name: Pages
  Label: Page Count
  Group: Src
  Data: 12
– Name: DatePubCY
  Label: Publication Date
  Group: Date
  Data: 2025
– Name: TypeDocument
  Label: Document Type
  Group: TypDoc
  Data: Journal Articles<br />Reports - Research
– Name: Audience
  Label: Education Level
  Group: Audnce
  Data: <searchLink fieldCode="EL" term="%22Higher+Education%22">Higher Education</searchLink><br /><searchLink fieldCode="EL" term="%22Postsecondary+Education%22">Postsecondary Education</searchLink>
– Name: Subject
  Label: Descriptors
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Anatomy%22">Anatomy</searchLink><br /><searchLink fieldCode="DE" term="%22Neurology%22">Neurology</searchLink><br /><searchLink fieldCode="DE" term="%22Visual+Aids%22">Visual Aids</searchLink><br /><searchLink fieldCode="DE" term="%22Fidelity%22">Fidelity</searchLink><br /><searchLink fieldCode="DE" term="%22Photography%22">Photography</searchLink><br /><searchLink fieldCode="DE" term="%22Electronic+Equipment%22">Electronic Equipment</searchLink><br /><searchLink fieldCode="DE" term="%22Medical+Students%22">Medical Students</searchLink><br /><searchLink fieldCode="DE" term="%22Graduate+Students%22">Graduate Students</searchLink>
– Name: DOI
  Label: DOI
  Group: ID
  Data: 10.1002/ase.70062
– Name: ISSN
  Label: ISSN
  Group: ISSN
  Data: 1935-9772<br />1935-9780
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: Cadaveric dissections, which are considered the most realistic model to study neuroanatomy, are expensive and not readily available in all centers. Given the surge of technological advances, incorporation of three-dimensional (3D) scanning technologies and 3D models has gained popularity, both in the educational and clinical settings. We present our institutional experience in creating high-fidelity neuroanatomical 3D models using three 3D scanning techniques: structured light 3D scanning, "manual" photogrammetry with a single DSLR camera, and "automatic" photogrammetry using a scanner equipped with five vertically arranged DSLR cameras and an automatic turntable within a square box. A survey study was conducted with 20 neurosurgical residents to assess the quality of the three resulting 3D models. In the study, "manual" photogrammetry was determined to be the most cost-effective technique, while "automatic" photogrammetry was the most time-effective and user-friendly technique. The best visual quality was obtained using "manual" photogrammetry, as determined from survey results of 20 neurosurgical residents. While structured light 3D scanning had the lowest quality of resolution of the texture map, this technique was the most accurate to use for determining measurements, with a fixed accuracy of 0.05 mm. Overall, "manual" photogrammetry can allow for the most detailed 3D models and is the most cost-effective strategy, while structured light 3D scanning is most suitable for obtaining clinically relevant measurements given the high degree of structural accuracy. Alternatively, "automatic" photogrammetry can serve as a hybrid between obtaining relatively high-quality models in a time-effective and user-friendly manner.
– Name: AbstractInfo
  Label: Abstractor
  Group: Ab
  Data: As Provided
– Name: DateEntry
  Label: Entry Date
  Group: Date
  Data: 2025
– Name: AN
  Label: Accession Number
  Group: ID
  Data: EJ1476155
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=eric&AN=EJ1476155
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1002/ase.70062
    Languages:
      – Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 12
        StartPage: 697
    Subjects:
      – SubjectFull: Anatomy
        Type: general
      – SubjectFull: Neurology
        Type: general
      – SubjectFull: Visual Aids
        Type: general
      – SubjectFull: Fidelity
        Type: general
      – SubjectFull: Photography
        Type: general
      – SubjectFull: Electronic Equipment
        Type: general
      – SubjectFull: Medical Students
        Type: general
      – SubjectFull: Graduate Students
        Type: general
    Titles:
      – TitleFull: A Comprehensive Guide to High-Fidelity 3D Neuroanatomical Modeling Techniques: A Quantitative Comparison between Photogrammetry and Structured Light Scanning
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Megan M. J. Bauman
      – PersonEntity:
          Name:
            NameFull: Amedeo Piazza
      – PersonEntity:
          Name:
            NameFull: Fabio Torregrossa
      – PersonEntity:
          Name:
            NameFull: Charles Wes Price
      – PersonEntity:
          Name:
            NameFull: Jonathan M. Morris
      – PersonEntity:
          Name:
            NameFull: Luciano C. P. C. Leonel
      – PersonEntity:
          Name:
            NameFull: Maria Peris-Celda
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 07
              Type: published
              Y: 2025
          Identifiers:
            – Type: issn-print
              Value: 1935-9772
            – Type: issn-electronic
              Value: 1935-9780
          Numbering:
            – Type: volume
              Value: 18
            – Type: issue
              Value: 7
          Titles:
            – TitleFull: Anatomical Sciences Education
              Type: main
ResultId 1