Daily habits that compound

• Critique every image you take against the 10-point list, even the good ones. Skill grows at the review monitor, not the exposure button.
• Log your repeats and the reason. Patterns appear fast: most beginners have one dominant fault (usually rotation or clipped anatomy).
• Say your plan out loud before each exam: projection, SID, receptor, factors or AEC, breathing instruction. It forces the decision before the exposure.
• One physics question per day to a senior. “Why 125 kVp for chest but 55 for a hand?” beats a textbook chapter.

Terminology across countries

This page uses “radiographer” for brevity. SID is quoted in centimetres with US inch equivalents: 100 cm ≈ 40″, 180 cm ≈ 72″.

Dose quantities you must know

Typical effective doses (adult, approximate)

Values vary with patient size, technique and equipment; use them for patient conversations and perspective, not dosimetry.

Dose badges (personal dosimetry)

• Most departments issue OSL or TLD body dosimeters, exchanged monthly to quarterly.
• Wear position: trunk, usually chest height. If you wear a lead apron regularly, follow local policy — commonly the badge goes under the apron at waist/chest to estimate effective dose, with a second collar badge outside the apron in higher-dose roles (fluoroscopy).
• Never leave your badge in the X-ray room, on the apron rack, or take it through airport security in checked luggage; never wear someone else’s.
• Your readings are legally recorded. A general radiographer working well typically records at or near zero above background — a non-zero trend is a prompt to review your habits with the RPS/RSO.

Occupational dose limits (ICRP 103)

• Whole body: 20 mSv/year averaged over 5 years (no year >50 mSv). The US regulatory limit is 50 mSv/year (NRC), but facilities still optimise well below it.
• Public: 1 mSv/year — this is the number that governs carers and room design.
• Declared pregnancy (worker): dose to the embryo/fetus kept ~1 mSv for the remainder of pregnancy (UK/EU); US: 5 mSv gestation total. Declaring is your choice, but do it early — duties can be adjusted, and general radiography is almost always safe to continue.

Room design in 60 seconds

• Controlled area: the X-ray room itself while the tube can be energised — entry restricted, signage and warning light at the door (“X-ray on” illuminates during exposure).
• Structural shielding: walls, doors and the control-booth screen contain lead or equivalent (concrete/barium plaster), specified by a qualified physicist based on workload and what’s next door.
• Primary barriers face the beam (wall behind the chest stand, floor under the table); secondary barriers handle scatter and leakage only.
• Control booth: designed so you can see the patient and reach the exposure switch while fully behind the screen. If your body isn’t fully behind it, you’re doing it wrong.
• Tube leakage is limited by regulation (≤1 mGy/h at 1 m in typical standards) — the housing is part of the protection system; report damage.

Who may stay in the room?

• Default: nobody but the patient. Staff behind the screen; everyone else outside.
• Carers and comforters (a parent holding a child’s hand, a relative supporting a confused patient) may remain if the exposure can’t otherwise be achieved: they must consent, not be pregnant, wear a lead apron, and stand out of the primary beam. Their exposure is planned and documented under local “carers and comforters” dose constraints.
• Staff must not routinely hold patients or receptors. If holding is genuinely unavoidable, rotate the duty, wear protection, keep hands out of the primary beam — and treat every instance as a flag to find a positioning aid (sandbags, straps, cassette holders) next time.
• Mobiles/theatre: announce “X-ray!” before exposure, give staff time to step back 2 m or shield, and check the beam path — including what’s behind the patient.

kVp — quality (and quantity)

• Sets the maximum photon energy and shifts the whole spectrum up: higher kVp = more penetrating beam.
• Contrast lever: low kVp → more photoelectric absorption → high subject contrast (bone vs soft tissue “black and white”); high kVp → Compton dominates → longer grey scale. Chest at ~110–125 kVp deliberately flattens rib contrast so lungs and mediastinum are both readable.
• kVp also increases beam quantity (output rises roughly with kVp²), which is why the 15% rule works (Tab 4).
• Too low kVp cannot be fixed with mAs: photons that never reach the receptor make dose, not signal.

mAs — quantity only

• mA (tube current) × time (s). Controls how many photons — nothing about their energy.
• On digital receptors, mAs controls noise (quantum mottle), not brightness — the software rescales brightness whatever you do. Grainy image = not enough mAs (or too little reaching the receptor).
• Doubling mAs doubles patient dose, linearly. It is your most “expensive” dial.
• Reciprocity: 20 mA × 0.5 s = 100 mA × 0.1 s = 10 mAs. Choose high mA, short time whenever motion is possible — which is almost always.

Scatter — the image’s enemy

• Compton-scattered photons carry no positional information; they lay a grey fog over the image and reduce contrast — and they are the source of staff dose.
• Scatter increases with three things you partly control: field size (collimate!), part thickness (compress/position where appropriate; can’t change the patient), and kVp (higher kVp → relatively more Compton).
• Rules of thumb: the scatter problem becomes significant once the part is thicker than ~10 cm or you’re working above ~60–70 kVp — that’s the classic threshold for reaching for a grid.
• An air gap (increased object–receptor distance) also cuts scatter reaching the receptor — used deliberately in some lateral C-spine and chest techniques instead of a grid.

Grids — scatter cleanup, at a price

• Lead strips + interspace material between patient and receptor; lets primary through, absorbs oblique scatter.
• Grid ratio (strip height ÷ interspace width, e.g., 8:1, 10:1, 12:1): higher ratio = better cleanup, less positioning forgiveness, more dose. Wall buckys are commonly 10:1–12:1.
• Focused grids have angled strips converging at a focal distance — use them within their SID range and centred, or you get grid cutoff (Tab 8).
• The Bucky mechanism oscillates the grid to blur its lines. Stationary grids (mobile work) can show lines or moiré.
• Grids cost dose (Bucky factor typically 3–5× the mAs of a non-grid exposure). So: no grid for thin parts — extremities and paediatric work are usually gridless on the tabletop.

Manual technique — the rules of thumb

• 15% rule: increasing kVp by 15% has roughly the same effect on receptor exposure as doubling the mAs (and vice versa: −15% kVp ≈ halving mAs). Use it to trade contrast against dose/time: patient can’t hold still → +15% kVp, halve the mAs, halve the time.
• Part thickness: classic guide — every ±4–5 cm of tissue ≈ double/halve the mAs (or adjust ~2 kVp per cm for modest changes). Measure with calipers on bariatric or paediatric patients rather than guessing.
• Inverse square / exposure-maintenance: receptor exposure falls with distance². Moving SID from 100 cm to 180 cm needs ≈ (180/100)² = 3.2× the mAs for the same receptor exposure.
• Grid in/out: adding a typical grid needs roughly 3–5× mAs (Bucky factor). Tabletop extremity factors do not transfer to the bucky.
• Casts: dry plaster ≈ +8–10 kVp or ~2× mAs; fibreglass much less (often minimal change). Check local chart.

Exposure index — your feedback loop

• The exposure index (EI) estimates the dose that reached the receptor in the relevant anatomy, standardised by IEC 62494-1 and AAPM Task Group 116 (older vendor scales: Kodak/Carestream EI, Fuji/Konica “S number” — which runs inverse, Agfa lgM).
• The deviation index (DI) compares your EI to the department’s target for that exam: DI = 10 × log₁₀(EI / EI_target). DI 0 = on target; +1 ≈ 26% over; +3 ≈ double; −3 ≈ half the intended receptor exposure.
• Working band: keep DI roughly within −1 to +1; investigate anything beyond ±3. Persistent positive drift across the room = exposure creep.
• EI is not patient dose — a tightly collimated exposure can have a fine EI and lower patient dose than a sloppy one with the same EI. And EI is corrupted by wrong exam selection, unusual collimation, or metalwork in the field — read it with your eyes on the image, not instead of them.
• Never repeat on the number alone. If the image is diagnostic, a DI of +1.5 is a learning point, not a repeat.

Anatomical side markers

• A physical lead L or R marker in the primary beam, at the time of exposure, is the legal and safety standard. Digital annotation added afterwards is a last resort and must be flagged as such (many departments require a comment) — post-hoc labels have contributed to wrong-side surgery.
• Mark the side being examined for extremities; mark the patient’s right (or per local convention) for trunk and chest work.
• Place it in the collimated field but clear of anatomy — corner of the field, on the lateral aspect.
• For decubitus and horizontal-beam work, add the position (“L lateral decubitus”, arrow up) so the reader can interpret air–fluid levels.
• Buy your own set, engrave your initials (required in some countries for medicolegal traceability), and know where they are at all times. A radiographer without markers is a radiographer about to make an error.

Collimation

• Collimate to the area of clinical interest — no larger, but genuinely no smaller. Clipping the anatomy you were asked to image is the beginner’s most common repeat.
• Benefits stack: less irradiated tissue (dose), less scatter (contrast), cleaner AEC behaviour, better automatic processing.
• Visible collimation borders on all four sides of an extremity image is the classic evidence of good practice; on trunk imaging aim for borders within the receptor where anatomy allows.
• Do not crop electronically to fake tight collimation. Cropping hides the irradiated field from audit and hides anatomy from the reporter. Many departments prohibit masking beyond minimal tidy-up.
• Learn each exam’s field definition, e.g., chest = apices to costophrenic angles; abdomen = diaphragm to symphysis pubis; long bones = joint above and joint below for trauma.

Alignment: the three-point check

1. Part: positioned per the projection (rotation controlled, joints in true position), immobilised — sandbags, sponges, straps do the holding, not people.
2. Tube: centred to the defined centring point, correct angle, correct SID (read the tape/display, don’t eyeball).
3. Receptor: centred to the CR (detent positions help), correct orientation (crosswise vs lengthwise), correct bucky selected on the console.

Say it before every exposure: part–tube–receptor. Misalignment of any one gives cutoff, clipping or grid artifacts. With focused grids, tube–receptor alignment errors show up first.

Patient care around positioning

• Identify: three-point ID check (full name, date of birth, address/NHI/MRN per local policy) against the request — every patient, every time, even the one you just imaged an hour ago.
• Explain and consent: one sentence of what and why, then instructions in plain words. Cooperation you earn is immobilisation you don’t need.
• Remove artifacts before positioning: jewellery, ECG leads where permitted, bra fastenings, hair braids for C-spine/skull, dressings only per policy. Give a private space and a gown.
• Move the tube, not the patient, when the patient is in pain — adapt the projection to them (horizontal beam, cross-table) rather than forcing textbook positions.
• Dignity and safety: never leave a frail patient standing unsupported at the chest stand; use rails and chairs; watch for dizziness after lying-to-standing.
• Infection control: clean receptor and table between patients; cover portable panels for wound/ward work.

Knee (AP + lateral) — your first grid extremity

1. AP: supine/seated, leg extended, patella facing forward (leg usually rotated ~3–5° internally to correct natural external rotation). CR ~1 cm below the patellar apex; angle per body habitus (≈0° average, ~3–5° cephalad for wide pelvis, caudad for slim) to open the joint space.
2. Lateral: affected side down, knee flexed 20–30° (no more if patella injury suspected), CR 5–7° cephalad to superimpose the femoral condyles. Check: condyles superimposed, patellofemoral joint open.
3. Grid if the knee measures >10 cm (most adults); tabletop for slim knees and most paediatrics — and remember the mAs conversion (Tab 4).

Elbow (AP + lateral) — the fat-pad exam

1. AP: seated, arm fully extended, hand supinated, shoulder dropped to the table plane. CR to mid-elbow. Both epicondyles parallel to the receptor.
2. Lateral: elbow flexed exactly 90°, thumb up, humerus and forearm in the same plane. CR to the lateral epicondyle.
3. Why the fuss about 90°? The posterior fat pad is only reliable on a true 90° lateral — its visibility implies a joint effusion and, in trauma, an occult fracture (radial head in adults, supracondylar in children) even with no visible fracture line.
4. Can’t extend the injured elbow? Don’t force it — two AP-style projections (one for forearm, one for humerus) with the elbow as-is, per trauma protocol.

Repeats: the honest framework

• A repeat doubles the patient’s dose for that projection. The threshold is “does the fault defeat the clinical question?” — not “would I be proud of this image?”
• Before repeating, name the fault and the fix out loud. Repeating without changing anything is just irradiating hopefully.
• Log every repeat with a reason. Departmental reject/repeat analysis (typically targeting rates under ~5–8%) is a core QA activity — and your personal log is the fastest self-improvement tool in radiography.
• Deleting a suboptimal exposure without logging it is a serious professional breach in most jurisdictions: every exposure delivered a dose and must be accounted for.

Viewing conditions matter

• Critique on the acquisition console is a first pass only — its display is rarely diagnostic grade. Borderline calls (subtle mottle, faint pneumothorax) deserve a reporting-grade monitor.
• Use the window/level tools before declaring an exposure fault: digital images carry more information than any single rendering shows.
• Zoom to 1:1 to judge sharpness and noise; whole-image thumbnails hide both.
• Ambient light: bright rooms wash out low-contrast detail — dim the review area where you make repeat decisions.

Motion — still the #1 image killer

• Voluntary motion (breathing, moving): blur that follows the patient’s movement, often one region. Fix with communication, clear breathing instructions, immobilisation aids — and only then shorter times.
• Involuntary motion (peristalsis, tremor, cardiac): can’t be coached away; fix with high mA + shortest time (reciprocity, Tab 3) and the 15% rule if you need more headroom.
• Distinguish from geometric unsharpness (uniform softness everywhere): large focal spot, long OID or short SID — a setup fault, not a patient one.
• Sleeping ≠ still: children and sedated patients move on the stimulus of the exposure noise. Watch, time your exposure, expose mid-instruction-hold.

Double exposure (mainly CR)

• Two superimposed images on one plate — nothing in a CR cassette stops you re-exposing an unread plate.
• Signature: ghost anatomy from a different patient/projection underneath yours. Both exposures are lost → both must be repeated, and the incident documented (two patients received dose for nothing).
• Prevention is pure workflow discipline: read plates immediately, keep exposed and unexposed cassettes physically separated, and follow the local “exposed” flagging convention.
• Related CR faults: incomplete erasure/ghosting (faint prior image after a very high exposure — run an erase cycle) and fogged plates (stored near scatter or left days between erase and use).

Backscatter

• Radiation scattered from behind the receptor back into it — through the cassette/panel rear — imaging the hardware: a ghost of hinges, electronics, lead blockers or the cassette label superimposed on the anatomy.
• Risk rises with high kVp, large fields, thin patients/parts, and unprotected receptor backs (some cassettes carry lead backing for this reason).
• Classic scenario: mobile chest with the panel against a metal bed frame, generous field. Fix: collimate, and mind what’s behind the receptor.

Jewellery, clothing & external objects

• Necklaces over lung apices, bra wires/fasteners over lung bases, piercings everywhere, plaits/buns over C-spine and skull, ECG electrodes and lines on every ICU chest.
• Habit: gown-and-check before positioning, then a last visual sweep of the collimated field at the tube. It is faster than one repeat per week.
• What can’t come off (fresh dressings, monitoring you may not remove, fixed jewellery): position it out of the field where possible, and comment on the image so it isn’t mistaken for pathology.
• Artefacts inside the patient (pacemakers, surgical clips, foreign bodies) are findings, not faults — never try to “improve” them away.