AP · 0° · tilt 0°
table 0 / 0 mm
—
Menu. Press Esc (or ☰) at any time to choose the mode, region and degeneration, start a new case, restart or quit the current one, or change settings (screening, image, teaching overlays, theme).
Practice (the fourth mode): choose any procedure, level and side. A coach beside the needle controls walks you through the steps — target view, entry down the beam, depth check, target, contrast — with live C-arm and needle corrections, hazard warnings and, if you like, the target and ideal path drawn on the image. Finishing gives a detailed debrief.
Eponymous conditions (Esc menu, off by default) add Baastrup's disease, Bertolotti's syndrome and Maigne's syndrome to the cases, with the changed anatomy and the injections that treat them. Costotransverse joint injections and facet cyst rupture are always available.
Repositioning. Each needle in the patient has a number (tagged on the image too). Tap its number in the Needles card — or press 1–9 — to pick it up and reposition it; + starts a new needle.
The procedure screen shows only the fluoroscope and the needle. 📋 above the image brings back the case summary or prescription; the needle set-up pops up before each needle, and the feedback pops up when you finish.
Multi-needle cases. Every case is one patient with as many needles as you want: place a needle, press Leave it in & add another needle (it stays in the patient and on the image), choose the next needle and scrub in again. Finish & assess identifies every needle and scores the case on placement (technique), risk (complications and unneeded needles) and efficacy (how well the needed targets were treated). Guided gives a multi-level prescription, Sandbox sets none, and Clinical case gives a referral for you to treat as you see fit. A major complication, or a procedure that should not have been done, fails the case.
The small solid figure in the image's lower-left corner is the patient seen from the beam's angle (orange = the back, with its spinous processes). An X-ray is transparent, so a moving image can seem to turn inside-out; the figure shows which side is toward you. Spine and joints in the case card adds degenerative change to each new case (mild, moderate or severe); severe cases may also carry previous surgery: fusions, laminectomies, disc or joint replacements, cement augmentation or a stimulator. Landmarks draws the radiological targets and structures to avoid; Nerves draws the course of the nerve being targeted; Measure is a caliper; Mag and Invert work as on a C-arm. Assessment mode (in the case card) turns the aids off and scores a series of cases.
Hold SCREEN on the image with one finger while you work the other controls with another. The strip under the image moves the C-arm and table; hold a button to keep moving, faster after a moment. Tap the image to insert the needle. Touch and hold any control to see what it does. To use FluoroSim like an app, open it in Safari and choose Share, then Add to Home Screen.
Space screen · ← → oblique · ↑ ↓ tilt (Shift: 5°) · W S advance / withdraw · A aspirate · C contrast · 1 2 electrode (bipolar RF) · Esc close a dialog · ? this page
This page opens by itself on a first visit; the Quick start link in the header brings it back. The full notes are under About & anatomy notes.
By Dr Joseph Kluver, Specialist Pain Medicine Physician — Herston Pain Pathways, Brisbane.
Getting started. Quick start in the header has the five steps of a case and the keyboard shortcuts (press ? at any time). The patient is built in your browser the first time the page opens, which takes 15–20 seconds; the result is kept in the browser's own storage, so later visits with the same version open in a few seconds. Progress (the running tally, the session log and saved assessment reports) lives in this browser too: Export beside the tally saves it to a file, and Import loads it on another computer.
A browser-based trainer for image-guided pain procedures of the spine, pelvis, head and neck, shoulder, hip and knee. Each case is randomly generated (procedure, level, side) from the selected region — head, cervical, thoracic, lumbar, sacral/pelvic, shoulder, hip, knee, or mixed. You position a free-moving C-arm, choose a needle, click the image to set a trajectory down the beam, then advance, withdraw or redirect while checking from any view. The needle is drawn at its full length with its hub on the outer end: down the beam the hub is a ring around the needle (hub over tip), and off the coaxial view it marks which end is outside the patient. (The hub is drawn denser than real moulded plastic, which is usually burnt out, so that it always shows.) Before scrubbing in you choose the needle: straight or curved tip, sharp (Quincke) or blunt, and the gauge, starting from the procedure's standard needle. A blunt tip goes in through a short introducer and pushes vessels, nerve roots, peripheral nerves and the dura aside on first contact — advance again to push through; a 25G needle is comfortable but bends at depth and often fails to draw back blood from small vessels; and an RF cannula's gauge sets the size of its lesion, and with it how accurately it has to be placed. The choice is fed back with the result (good, acceptable or not ideal for that procedure, and why) and recorded in assessment reports. The final 3D tip position is scored against the procedure's anatomical target, and every structure the needle touched or traversed is reported with its likely clinical consequence.
One patient, the whole spine — with the head, shoulders, hips and knees. The anatomy is a single continuous model from the vertex to the knees — the skull and face, occiput, atlas and axis, every vertebra from C2 to L5 on one smoothly curved column (cervical lordosis, thoracic kyphosis, lumbar lordosis, sacral slope), the sacrum, coccyx and pelvis, the femora, knees and upper legs, the ribs, sternum, clavicles, scapulae and humeri, the airway, lungs, heart and great vessels, the abdominal and pelvic viscera, the cord, thecal sac and roots, and one soft-tissue envelope from the head to the lower legs. T12 is transitional, with thoracic-type facets above and lumbar-type facets below. The region buttons only choose which procedures are drawn; each case starts with the table centred on its region (the sella or C1, C5, T7 or L3 under the isocentre; the sacral procedures start on the segment they concern — S1 for the S1 foramen, S2 for the sacroiliac joint, S3 for the hiatus, S4 for the ganglion impar, L4–L5 for the hypogastric plexus and the cluneal nerves — and the shoulder, hip or knee cases on the joint of the case's side) and the table pans ±400 mm along the spine, so from a lumbar case you can drive up to the neck or down to the coccyx and see the levels run continuously, exactly as on a real table. The HUD names the vertebra under the isocentre. The model is built once per session (a progress notice shows the stages); the exposure follows the isocentre along the spine as automatic brightness control would.
The image. Each frame is a cone-beam projection: every solid in the model is integrated along the rays (density × path), and the summed attenuation is displayed through the chain of a modern C-arm — detector and focal-spot blur, a scatter veil, the image processor's local contrast enhancement, a display curve and quantum mottle, inside a 150 mm circular field. Cancellous bone carries a trabecular texture (the path through each bone's cancellous core is integrated separately and modulated by a pattern that stays fixed to the patient as the table moves), so vertebral bodies, the sacrum, the pelvis and the diploë read as bone rather than flat grey. An automatic brightness control samples the central third of the field after every move and sets the exposure so its mean brightness stays constant — a lateral through the shoulders or the skull base settles to the same grey as an AP of the lumbar spine, as it does on the real machine — and the image is rendered at the panel's own pixel size (up to 1152 px) on large or high-density screens.
Each case starts in the position the procedure is normally done in: prone for the spine and pelvis (tube under the table, needle from the back), supine for most shoulder, hip and knee injections and the anterior head and neck procedures. The Patient toggle switches a case to supine until a needle is in — the tube then sits behind the patient and the needle enters from the front. Only the stellate ganglion block and the anterior cranial procedures (Gasserian, sphenopalatine, supraorbital, infraorbital and mental) are fixed supine, because their approach is from the front. If the patient is not in the position the procedure is normally done in, an amber notice says so under the task, and Reset orientation (in the case card and on the image toolbar) puts things back: the standard position for that procedure and the C-arm at AP over a centred table. If the patient has to be turned with a needle already in, that needle comes out and a fresh one is placed; nothing else about the case changes.
Pedal mode, free mode and saving images. The simulator starts in pedal mode, like a real C-arm: the image is live only while the Space bar or the SCREEN button is held, and otherwise the last image is held while you move the C-arm, the table or the needle — so you learn to move first and screen briefly, and the screening time is counted and reported with each case (it is not scored). Each new case starts with a blank screen until you screen. Free mode on the image toolbar keeps the image live all the time instead; the choice is remembered on this computer, and the pill on the C-arm card shows which mode you are in. Save image downloads the current frame as a PNG named after the case and the view, and Measure is a caliper: two clicks on the image give the distance between them in the isocentre plane, as the machine's own ruler does (it clears when the view changes).
Running score. The header keeps a tally of the cases done on this computer (cases, average, best, streak) with the session log beneath the C-arm card; Reset score (in the header or the log) clears both. Assessment reports are kept separately.
Assessment mode. Assessment — choose procedures… in the case card opens a checklist of every procedure by region. Tick the ones you want to be examined on and the simulator serves one case per ticked procedure (×1, ×2 or ×3) at random levels and sides in a shuffled order, with the region buttons locked. Each case is scored and shown as usual after you submit, but it cannot be retried or restarted and the hidden anatomy is not revealed. Identifying the level is part of the examination, so the level readout beside the table position is hidden and bone contacts are reported without naming the bone; the Landmarks and Nerves aids are off and come back afterwards as you left them. After the last case a report gives the average score and grade, a pass/fail verdict (pass: average 75 or more, no case below 50, no critical complication), the average, best and lowest score for each procedure, and a line for every case (score, tip-to-target distance, accuracy / orientation / safety marks, penalised structures, contrast use, time). The report can be downloaded as a self-contained page, and the last twelve reports are kept on this computer and can be reopened from the assessment panel. Abandon stops an assessment without saving it.
RF lesions: monopolar or bipolar. For the radiofrequency neurotomies (lumbar, thoracic and cervical medial branches, third occipital nerve, sacral lateral branches, the hip's articular branches and the genicular nerves) the case card offers a Bipolar option: two cannulae of the same type are placed one after the other (a click on the image places whichever electrode is not yet in; the Electrode 1 / 2 switch or keys 1 2 choose which one the advance, redirect, aspirate and contrast controls act on) and the HUD reports their shaft-to-shaft spacing, angle and tip alignment. In review the estimated lesion is drawn in red — an ellipsoid about each active tip, bridged into a strip when the pair is close enough — and the green ideal is a pair of parallel electrodes straddling the nerve.
Viewpoint. The image is always shown from where the operator stands: behind a prone patient, in front of a supine one. The perspective centre of the projection is on that same side, so the structures nearest you are slightly larger and sweep furthest when you oblique the C-arm or pan the table — a prone patient's spinous processes swing across the vertebral bodies the way they do when you walk round a patient's back, and the needle comes toward you. An X-ray is transparent, so in live movement those cues are what tell your eye which side you are looking from. (A real tube-under-table C-arm has its perspective centre at the tube on the far side. A still image is otherwise the same, but in live movement that makes the patient look as if seen from the other side — prone from the front, supine from behind — so the simulator keeps the centre on your side; the attenuation physics is unchanged.) The rule is fixed and there is no mirrored display option: the image is exactly what you see from where you stand. Prone — viewed from behind, spinous processes toward you, the patient's left on the left of the screen. Supine — viewed from the front, the patient's left on the right. A tag at the top of the image states the position at all times, and the needle always enters from your side of the patient.
Nerve courses. The Nerves toggle on the toolbar (beside Landmarks, and usable with or without it, live screening included) draws the course of the nerve the procedure is aimed at over the image, in yellow, with the segment you are trying to reach in green and labels that follow the C-arm. A medial branch block or RF shows the dorsal ramus leaving the foramen, the medial branch crossing the root of the transverse process at its junction with the SAP, passing under the mamillo-accessory ligament and over the lamina, with its articular branches to the two joints it supplies (and the lateral branch running off over the TP); in the neck the branch wraps round the waist of the pillar, and the third occipital nerve crosses the C2–3 joint. A facet joint injection shows the two medial branches that supply that joint. A transforaminal or DRG procedure shows the exiting root and its ganglion in the foramen; sympathetic blocks show the chain with its ganglia (a ramus communicans block its grey ramus; a coeliac block the greater splanchnic nerve passing through the crus to the ganglion); sinuvertebral RF the nerve turning back into the foramen to the back of the disc; the SI joint shows the lateral branches leaving the posterior sacral foramina; the cluneal, pudendal, impar and hypogastric procedures their own nerves or plexus. In the head and limbs it shows the nerve blocked or the joint's supply — the trigeminal ganglion and its divisions, the sphenopalatine ganglion, the supraorbital, infraorbital, mental and greater occipital nerves, the C1 nerve and the C2 ganglion beside the atlanto-occipital and atlantoaxial joints, the suprascapular nerve (with the axillary nerve for the glenohumeral joint), the genicular nerves running with their arteries, the obturator and femoral articular branches to the hip, and the sciatic nerve leaving below piriformis. Epidurals and bursal or AC joint injections have no target nerve, so nothing is drawn and the status message says so. The courses are schematic — placed from typical anatomy relative to the bony model, not traced from imaging — and real nerves vary. It is switched off in assessment mode.
The patient is a schematic 3D model: the bones are smooth signed-distance solids and the soft tissues, vessels and nerves simpler shapes. The fluoroscopic image is a true cone-beam projection: every structure contributes attenuation proportional to the path length the ray travels through it times its density, so overlapping bone appears denser, cortices appear as rims, and signs such as the Scotty dog emerge from the geometry rather than being drawn. The perspective centre is 600 mm from the isocentre on the operator's side (see Viewpoint), with a 150 mm field at the isocentre at normal magnification (what a 9-inch image intensifier covers at the patient), 100 mm and 75 mm in the two magnification modes — so, as in theatre, you pan the table to the level you want rather than seeing the whole spine at once. Soft tissue attenuates at about 3% of cortical bone per millimetre and air at the negative of that, and the exposure setting follows the isocentre along the spine (as automatic brightness control would give), so the thin neck is displayed with more gain than the pelvis and the airway, lungs and bowel gas read as lucencies. The detector image is then processed the way a flat-panel fluoroscopy system processes it: a little unsharpness, a scattered-radiation veil that lowers contrast behind thick parts, mild edge enhancement, quantum mottle that changes from frame to frame, and a display curve that puts soft tissue in the mid-greys and bone toward white. Nerves, vessels and viscera are invisible on fluoroscopy, as in life — they can be revealed after each attempt. The head is always at the top; the side markers show which side of the patient is which (L on the left of a prone image, on the right of a supine one), changing to ANT/POST in near-lateral views. The orbit is limited to ±90° so the tube always stays on the table side of the patient and the needle always enters from the operator's side (the back of a prone patient, the front of a supine one). In the review overlay, nearer structures overlie deeper ones from the same viewpoint, so on a prone patient the dura and posterior elements lie in front of the aorta and viscera. Its colour legend folds to a title bar (the – / + button, or click the bar) so it does not cover the image; the choice is remembered. The Landmarks toggle on the toolbar is a separate learning aid that works in every phase, live screening included: it draws the radiological landmarks of the procedure in hand over the image, with labels that follow the C-arm — the target as a ⊕, the bony landmarks that define it in green and blue, and the structures to avoid in orange. A medial branch block shows the pedicle, SAP and TP of the level below (the Scotty dog's eye, ear and nose) and their junction; a transforaminal shows the pedicle's 6 o'clock, the subpedicular safe triangle, Kambin's triangle, the foramen on the lateral and the radicular artery in its antero-superior part; an interlaminar epidural the window, the midline and the spinolaminar line; a sympathetic block the chain on the anterolateral bodies, the lateral body margin and the anterior third; a cervical transforaminal the foramen, the mid-pillar limit on the AP and the vertebral artery; lateral branch RF the posterior foramina with the 2:30–5:30 / 6:30–9:30 lesion arcs; a caudal the hiatus, the canal and the S2 end of the thecal sac; a Gasserian the foramen ovale, the clival line and the V3 / V2 / V1 depths; and so on for every procedure. Marks that belong to one view appear only near it — lateral-view lines vanish on the AP, pedicle rings on the lateral. It is switched off in assessment mode.
Light theme. The Light theme button in the header switches the page to dark text on light panels for bright rooms; Dark theme switches back, and the choice is remembered on this computer. The X-ray image itself is the same in both — for a white-background image in strong light use Invert on the image toolbar, which also turns the side markers dark.
Accuracy (70 points, or 55 when orientation is scored): distance from the tip to the procedure's ideal target zone — full marks inside the zone radius, falling to zero at the miss radius. Orientation (15 points, RF neurotomy only): the distal electrode should lie parallel to the nerve course. For the spinal medial branches it is judged as on the lateral view — the electrode's craniocaudal slope against the nerve's — because the electrode is laid along the SAP from a slightly ipsilateral oblique and cannot follow the nerve's small mediolateral drift; for the other RF targets the full 3D angle counts. In bipolar mode the 15 points go instead to the strip lesion: full marks need the shafts 3–10 mm apart with an 18G 10 mm tip (3–7 mm with a 20G 5 mm tip), within 15° of parallel and with the active tips side by side (staggered by less than a third of their length), and the strip stays continuous to about 1.5 mm beyond that spacing, 35° off parallel or a stagger of about 60% of the tip length; wider spacing leaves two separate lesions, and accuracy then becomes the nerve's distance from the strip axis (or from the nearer lesion when no strip forms), so the electrodes need not be parallel to the nerve — it only has to lie within the gap. Safety (30 points): graded penalties per structure, weighted by severity, with a heavier penalty when the tip ends inside the structure (injection into it). A severe event (dura, radicular artery, aorta, bowel, pleura) caps the case at 55 points and a critical event (conus) at 40, whatever the accuracy. Fluoro shots and time are deliberately not scored. All weights live in SEVERITY_CONFIG in the page source.
The engine is region-agnostic: a procedure is a target function, an approach and a list of expected contacts on the shared whole-patient model (spine, pelvis, head, shoulders, hips and knees), so a new procedure only needs those added to PROCEDURE_LIBRARY.
Tick the procedures you want to be examined on. You get one case per ticked procedure (times the repeat count), at random levels and sides in a shuffled order. Each case is scored as usual and shown after you submit, but there are no retries or restarts and the hidden anatomy stays hidden. A report follows the last case: overall grade, pass/fail, per-procedure averages and every case, with a download. Reports are kept on this computer.