The Level 2 method for C-arms and angiography: logs, subsystem isolation, your limits and radiation protection
Read the logs of a mobile C-arm or a fixed angiography system for the first and causal event, isolate the fault to a subsystem with the cheapest safe test, stay inside your authorised level, and protect yourself and others when a test needs radiation.
What is assumed and what is new
Angiography and C-arm Service — Level 1 gave you the architecture of mobile C-arms and fixed angiography rooms, the safety rules, the imaging chain and a first fault-finding method. X-ray Service — Level 2 gave you the generator, the tube and fluoroscopy at fault-finding depth. This course does not repeat them. It applies them to two families of systems that fail in their own way.
A mobile C-arm is two machines joined by one cable: the C-arm stand with the generator, tube, collimator and receptor, and the monitor cart or workstation with the computer, the image processing and the storage. Many faults live in that join: the interconnect cable, its connectors and the safety loop that runs through it. A mobile C-arm is also pushed between theatres, cleaned with liquids and bumped every day, so mechanical and connector faults are far more common than on any fixed room.
A fixed angiography system is a network of controllers: generator, stand, table, collimator, detector, image system, collision system, control consoles in the room and in the control room, often two planes. It is used during long interventional procedures in which a patient may depend on it. That changes the priorities. Before you think about the cause, ask whether a procedure is running, whether the clinicians have a fallback, and whether the fault is a safety fault that must stop the system now.
Read the logs before you touch anything
Every modern system keeps some kind of event log, and most separate the events by weight. The Ziehm Solo, for example, sorts its messages into errors, alerts and plain messages: after an alert the user may continue with reduced function, after an error the system needs service. The GE Brivo OEC 850 shows a code in the kV and mA field of its control panel and a text on the workstation; its manual tells the user to restart once and, if the message returns, to switch off, unplug and call service. Fixed systems keep separate logs for each controller, often with separate levels for errors, warnings and information.
Three rules make a log useful. First, save it before any restart, because some systems overwrite or shorten the list at power-up; export it, or photograph the screen with the time visible. Second, find the first event of the sequence, not the loudest one: one lost supply can produce a page of communication errors from every module that depended on it. Third, read the warnings as well as the errors. A warning such as reduced speed, a sensor out of range or a retried message often appears days before the error that stops the system.
Learn what is normal noise on the system you look after. Some systems record harmless events as alarming messages, such as a hardware change every time a mouse or keyboard is moved to another port. Others record user actions: the patient data audit log of a Ziehm Solo, kept for data protection, records logins and changes to patient records, not faults. It is not yours to delete; national data rules and the hospital decide how long it is kept. Check the system clock too. A flat backup battery or a changed time zone puts events in the wrong order, and a sequence that looks impossible may simply have wrong time stamps.
The power-up and exposure sequence is your map
Place the fault in a sequence before you choose a suspect. On a mobile C-arm the power-up sequence runs roughly like this: mains and the power distribution board, the safety loop, the low-voltage supplies, the computer boot, the start of the application, then each module reporting to the host over its bus, and only then ready. A system that does not start at all points at the supply, the emergency stop, the key switch or the safety loop. On the Brivo OEC 850 the safety loop runs through the interconnect cable and several boards in the stand, so a missing cable or a board that has lifted out of its slot after a knock stops the power-up. A system that boots but reports missing modules points at the bus, the module's supply or the module.
The exposure sequence is the second map: request from a hand or foot switch, checks of preconditions such as key switch, collimator, detector ready and temperature, generator start, tube current, image acquisition, processing, display and storage. A message such as 'switch pressed during power-up' or 'short circuit in foot switch' places the fault at the first step. A kV or mA message places it at the generator and tube. A system that exposes, shows kV and mA on the console and still gives no image places it after the beam: receptor, acquisition or display.
Isolate: compare, swap, separate
C-arms and angiography systems give you several built-in comparisons. Use them before any part is ordered. Fluoroscopy against a single exposure, manual against automatic dose rate control, large focus against small focus, one field size against another, the hand switch against the foot switch, the room console against the control room console. On a biplane system the two planes are often built from the same modules, and a fault that moves with a module when you swap it by procedure has been found. A fault that stays where it was has not.
Separation is the second tool. With the system switched off, disconnect a suspect pedal, a remote console or a peripheral and see whether a false request or a communication error disappears. On a mobile C-arm, see whether the workstation starts without the stand. Remember that some workstations are designed to shut down or stop when the stand is missing, so a workstation that switches off alone is not proof of a fault; the Level 1 course showed such a case.
Intermittent communication faults need patience. Reseat the connectors of the link concerned, follow its cable through the arm or the cable chain, look for a bent pin or corrosion, and only then suspect the interface chips at each end. A fault that disappears after reseating is a strong clue, not a repair: note the connector, tell the hospital and plan a check at the next visit.
What the field engineer may change
Most manufacturers design C-arms for repair at the level of the field replaceable unit; the Brivo OEC 850 manual states this openly. Component repairs in the field are possible, but they move responsibility for the result to you and your company, and they belong to the workshop and the Electronics course, not to a theatre at night.
Service software is usually layered by authorisation. The Philips BV family service tool, for example, gives different functions to an administrator and to service users at several levels, protected by a hardware key. You work only with the functions your level and your training allow. Before you change any configuration or calibration, save what was there. Never copy calibration or configuration data from another unit, never install software that is not the manufacturer's release for that system and licence, and never bypass, bridge or disable an interlock, a key switch, a collision sensor or an emergency stop to keep a procedure going. If the system cannot run safely, it does not run.
Radiation protection when you are the operator
During fault-finding you press the switch, often many times, and often in a room built for clinical work, not for testing. Prefer tests without radiation: the log, the service displays, power-up messages, movement tests and measurements on the de-energised system. When radiation is needed, keep each exposure as short as the test allows, use the lowest technique that answers the question, collimate, and make the exposure from the longest distance the cable or the control room allows.
No patient on the table. Nobody stays in the room unless the test needs them there, and anyone who stays is protected. Only staff authorised under national radiation rules make exposures. Doors closed, warning lamps working. Wear your apron and thyroid shield, and eye protection if you will be close to the field; wear your personal dosimeter. Use a test object or attenuator in the beam where the manufacturer's procedure asks for one. Some procedures also ask you to protect an image intensifier or detector from the direct beam during dose checks. Mobile C-arms add their own risk: in an operating theatre other people may walk in, so tell the theatre staff what you are doing and when. Keep a note of the tests and their duration; your own dose is part of your record.
| Part | What it does | How it fails | |
|---|---|---|---|
| 01 | Mains supply and power distribution | Feeds and switches every unit, with soft start | Phase or socket fault, failed relay or soft-start resistor |
| 02 | Safety loop, key switch and emergency stops | Allow power-up, movement and radiation | Pressed stop after transport, open loop through a lifted board |
| 03 | Interconnect cable (mobile C-arm) | Joins stand and workstation: power, signals, safety loop | Crushed cable, bent pin, broken contact |
| 04 | Host computer and application | Boots, loads configuration, controls the session | Flat backup battery, disk fault, corrupted files |
| 05 | Module bus (CAN, serial, Ethernet) | Carries commands and status between controllers | Missing module, timeouts, loose connector, interface chip |
| 06 | Hand and foot switches | Request fluoroscopy and exposures | Stuck contact, short circuit, broken wire in the cable |
| 07 | Generator, tube and collimator | Produce and shape the beam | kV and mA faults, arcs, filament and heat errors |
| 08 | Receptor and image system | Convert, process, display and store the image | No initialisation, noise, uniformity faults, storage full |
| 09 | Motion and collision control | Move stand and table safely | Position sensor drift, brake, motor, false collision signal |
An error that a restart always cured, for a while
A GE OEC FluoroStar 7900 C-arm kept dropping out with error 10018. Sometimes it worked for a whole list, sometimes it failed within minutes; each time a restart brought it back for a while. The engineer had no description of the code and asked colleagues. One of them had met it before and pointed him at a serial link between the C-arm side and the image processing computer: a connector on that RS-232 line, under one of the touch screens, was to be unplugged and pushed home again, then the connector at the computer end, and only if the fault returned should the interface chips of that link be checked. He was also advised to reseat the other connectors on the same path while he had the covers open.
An intermittent communication error that a restart clears points at the link, not at the modules at either end. Reseat and inspect the link first, record which connector it was, and treat a disappearing fault as a clue until it stays away.
A hardware failure every time the mouse moved
On a Siemens angiography system the engineers kept seeing a hardware failure message in the image system log. It appeared after work on the desk: a keyboard or mouse unplugged, moved to another USB port or exchanged. Colleagues who knew the system explained that it reacts to any change in the connected USB devices, including simply moving the same mouse to another port, and records it as a hardware failure. The message did not stop clinical work and was not a sign of a failing computer. Knowing this saved a search for a faulty board.
Learn the normal noise of the log on each system you look after, and connect messages with what was done just before them.
Nothing under Error, something under Warning
On a Siemens Artis zee biplane the table-side console sometimes lost the patient data, so the staff could not mirror the image or select the patient position. Restarts helped only sometimes. The engineer opened the log of the console controller and found no errors. A colleague asked whether he had looked at the warnings as well. He had not; under the warnings there were entries at the times of the complaints. A firmware reload of the console was proposed as the next step, by the manufacturer's procedure.
An empty error list is not an empty log. Read the warnings at the time of the complaint before you conclude there is nothing recorded.
On-site checklist
- Ask whether a procedure is running and whether there is a fallback before you start.
- Save or photograph the full log with time stamps before any restart.
- Note the first event, the warnings before it, and the system time.
- Place the fault in the power-up or exposure sequence.
- Use the built-in comparisons: fluoro and exposure, manual and auto, focus, switch, plane.
- Save configuration before you change anything; work only at your authorised level.
- For any test with radiation: no patient, short exposures, collimate, distance, apron, dosimeter.
- Record what you tested, how long you exposed and what you found.