Slide 1 — Title TrajectAR: An AR-Guided Trajectory Planning Trainer for HoloLens 2 [Name] | [Roll no.] | Computer Graphics and Gaming | [Teacher] | [Date] Slide 2 — Motivation / Problem Statement Planning a safe instrument path (biopsy needle, probe) through anatomy is a real, actively-researched clinical problem — trajectory must avoid bone, vessels, and other risk structures Traditionally planned on 2D monitor slices, requiring mental 3D reconstruction under time pressure Real published research (2024-2026) already explores AR-guided trajectory planning for biopsy/needle procedures, including HoloLens-based systems for prostate and splenic biopsy This project builds an education-scoped trainer based on that same real-world problem Image: a 2D slice-based path-planning screenshot vs. your rendered 3D volume with a trajectory line, side by side Slide 3 — Concept, Scope & Objectives What it is: HoloLens 2 app where a user places entry/target points in a real, world-anchored CT/MRI volume; the app samples tissue density along that path and classifies trajectory risk Objectives: (1) accurate volumetric rendering, (2) real-time trajectory definition and risk evaluation, (3) intuitive hands-free interaction, (4) generalize across datasets Target users: radiology/surgical trainees, anatomy students Scope boundary: education/training tool, not a certified clinical planning system; no live patient registration No image required Slide 4 — Requirement Analysis Functional: load volume data; ray-march render; interactive transfer function; hand grab/rotate; clipping-plane slicing; place entry/target points; sample density along trajectory; classify and visualize risk Non-functional: real-time frame rate on mobile GPU; stable world-locking; fully hands-free Layout: two-column table Slide 5 — Benchmarking Medivis AnatomyX/SurgicalAR | Novarad OpenSight | 3D Slicer/OsiriX (desktop) | published AR biopsy-navigation research (HoloLens-based prostate/splenic trajectory systems) Table: System | Platform | AR/Desktop | Clinical/Educational | Trajectory Planning? Takeaway: "Positioned as an accessible, education-scoped version of an active clinical AR research direction — same core problem, without certification requirements." Image: comparison table, built in-slide Slide 6 — Application Elements & CG Algorithm Plan Volume data: 3D voxel grid from CT/MRI Ray marching + volume rendering equation: I(D) = ∫₀ᴰ C(t)·e^(−∫₀ᵗτ(s)ds) dt approximated per-step, front-to-back: C_out = C_in + (1−A_in)·C_sample·A_sample A_out = A_in + (1−A_in)·A_sample Trilinear interpolation (density at any sample point, from its 8 surrounding voxels): value = lerp(lerp(lerp(v000,v100,fx),lerp(v010,v110,fx),fy), lerp(lerp(v001,v101,fx),lerp(v011,v111,fx),fy), fz) Transfer function: (R,G,B,A) = f(density), piecewise-linear, custom-generated in C# Trajectory sampling (new algorithm, reuses the above): march along the entry→target line segment at fixed intervals, sample density via the same triline
Slide 1 — Title TrajectAR: An AR-Guided Trajectory Planning Trainer for HoloLens 2 [Name] | [Roll no.] | Computer Graphics and Gaming | [Teacher] | [Date] Slide 2 — Motivation / Problem Statement Planning a safe instrument path (biopsy needle, probe) through anatomy is a real, actively-researched clinical problem — trajectory must avoid bone, vessels, and other risk structures Traditionally planned on 2D monitor slices, requiring mental 3D reconstruction under time pressure Real published research (2024-2026) already explores AR-guided trajectory planning for biopsy/needle procedures, including HoloLens-based systems for prostate and splenic biopsy This project builds an education-scoped trainer based on that same real-world problem Image: a 2D slice-based path-planning screenshot vs. your rendered 3D volume with a trajectory line, side by side Slide 3 — Concept, Scope & Objectives What it is: HoloLens 2 app where a user places entry/target points in a real, world-anchored CT/MRI volume; the app samples tissue density along that path and classifies trajectory risk Objectives: (1) accurate volumetric rendering, (2) real-time trajectory definition and risk evaluation, (3) intuitive hands-free interaction, (4) generalize across datasets Target users: radiology/surgical trainees, anatomy students Scope boundary: education/training tool, not a certified clinical planning system; no live patient registration No image required Slide 4 — Requirement Analysis Functional: load volume data; ray-march render; interactive transfer function; hand grab/rotate; clipping-plane slicing; place entry/target points; sample density along trajectory; classify and visualize risk Non-functional: real-time frame rate on mobile GPU; stable world-locking; fully hands-free Layout: two-column table Slide 5 — Benchmarking Medivis AnatomyX/SurgicalAR | Novarad OpenSight | 3D Slicer/OsiriX (desktop) | published AR biopsy-navigation research (HoloLens-based prostate/splenic trajectory systems) Table: System | Platform | AR/Desktop | Clinical/Educational | Trajectory Planning? Takeaway: "Positioned as an accessible, education-scoped version of an active clinical AR research direction — same core problem, without certification requirements." Image: comparison table, built in-slide Slide 6 — Application Elements & CG Algorithm Plan Volume data: 3D voxel grid from CT/MRI Ray marching + volume rendering equation: I(D) = ∫₀ᴰ C(t)·e^(−∫₀ᵗτ(s)ds) dt approximated per-step, front-to-back: C_out = C_in + (1−A_in)·C_sample·A_sample A_out = A_in + (1−A_in)·A_sample Trilinear interpolation (density at any sample point, from its 8 surrounding voxels): value = lerp(lerp(lerp(v000,v100,fx),lerp(v010,v110,fx),fy), lerp(lerp(v001,v101,fx),lerp(v011,v111,fx),fy), fz) Transfer function: (R,G,B,A) = f(density), piecewise-linear, custom-generated in C# Trajectory sampling (new algorithm, reuses the above): march along the entry→target line segment at fixed intervals, sample density via the same triline
Created using ChatSlide
TrajectAR is an innovative AR-guided trajectory training tool designed to address the limitations of 2D planning in 3D reasoning. It focuses on defining user objectives and safety boundaries while mapping functional and mobile-performance requirements. By benchmarking against AR, desktop, and clinical systems, it positions itself as an accessible, education-focused training solution. The rendering pipeline employs advanced techniques like ray marching and trilinear sampling to visualize...