The Aquilion platform: family, assemblies and the path from tube to image
Identify which Aquilion model and software generation is in front of you, name its main gantry, couch and console assemblies as the Aquilion documentation names them, and follow power, X-rays and data from the power distributor to the image on the display console.
One family, many badges
In CT Service Level 1 you learned the generic CT: what stands, what rotates, and how a slice is made. This course assumes all of that and goes into one family, the Toshiba Aquilion, sold since the company became Canon Medical Systems under the Canon name. In the field you will meet the older multislice machines, Aquilion 16, 32 and 64, the wide-detector Aquilion ONE, the Aquilion Prime and Prime SP, and the large-bore Aquilion LB, with a scattering of CXL, RXL and 128 units. The badge on the gantry cover tells you the marketing name. It does not tell you enough to order a part.
What tells you more is the system type number on the data plate, in the TSX-xxx form, with a suffix after it. The on-site modification manual we use in this course, for system software V4.40, applies to two type numbers: TSX-301A for Aquilion ONE and TSX-301B for the other Aquilion systems of that generation, and its example configuration screen shows a 160-slice Aquilion under the B type. FDA recall notices list the Aquilion 16, 32 and 64 under one type, TSX-101A, told apart by the suffix (for example /6 or /G for a 16, /D or /I for a 32, /E or /H for a 64), the Aquilion LB as TSX-201A and the Aquilion Prime as TSX-303A. The lesson is that the type number groups machines by platform generation, not by slice count, and that two machines with the same badge can be different generations inside.
So before you plan any work, read and write down four things: the data plate (type number with suffix, serial number, date), the system software version shown at start-up (for example V3.35 or V4.40 with its release suffix), the tube and generator type set in the system configuration, and the console arrangement. These decide which service manual volume, which software media, which spare parts and which procedures apply. Field reports in this course name model and software version for the same reason: a fix that worked on an Aquilion 64 with V3.35 is a clue, not a rule, for a Prime SP.
What differs between the models
The architecture is shared: a rotating frame carrying the tube, high-voltage generator and detector, a slip ring for power and an optical link for data, a couch with servo drives, and a scan console with separate reconstruction hardware and a raw data store. The differences that matter in service are mostly in detector width, generator power, bore size and computer generation. Treat the published figures below as identification aids; confirm the actual configuration on the machine and in its documentation.
- Aquilion 16, 32, 64: the older multislice generation, still the most common in independent service. Many run Windows-based console software of the V3.x generation and older disk technology. Most field cases in this course come from them.
- Aquilion ONE: the wide-area detector; the manufacturer's launch material describes 320 rows of 0.5 mm covering 16 cm in one rotation. Wider detector means more DAS hardware, more data per rotation and heavier reconstruction and raw data demands.
- Aquilion Prime and Prime SP: the manufacturer lists the Prime SP with 80 rows of 0.5 mm, about 4 cm coverage, 0.35 s rotation and a 78 cm bore. Prime units were subject of a 2016 field correction for oil leaking from the tube heat exchanger.
- Aquilion LB: large bore for radiotherapy planning and bariatric use; published data give a 90 cm aperture and a 70 cm scan field with 16 rows. Expect a different couch top and laser arrangement from the diagnostic models.
- Aquilion CXL, RXL, 128: variants of the 64-row-class platform with different consoles and options; their type numbers are not confirmed here, so read them from the data plate. Treat each as its own configuration and check the parts list for the serial number.
The gantry: what rides on the rotor
The Aquilion service manual for the scanner groups the hardware into the X-ray system, the data acquisition section, the data transmission unit, the console and the power distributor. That grouping is worth adopting, because the error logs and the troubleshooting chapters follow it.
On the rotor sits the X-ray system: the tube in its oil-filled housing with its own heat exchanger, the high-voltage generator with its inverter, transformer and voltage multiplier, the filament supply, and the starter that spins the anode up. The generator is controlled by the X-ray controller board, which the logs and the engineers call XC. Field reports name small parts of this chain that cause big symptoms: a 24 V supply mounted on the high-voltage multiplier on Aquilion 64, a temperature sensor on the tube, the high-voltage cable ends sealed with silicone compound in their receptacles. Opposite the tube sit the collimator, the detector and the DAS, which converts each detector channel into numbers. The data transmission unit takes those numbers off the rotor through an optical link; on the stationary side they continue by fibre to the reconstruction hardware.
Power reaches the rotor through the slip ring and its brush blocks. Aquilion engineers report two types of brush in use on the Aquilion 64, and they are not interchangeable at will: brushes must suit the ring material, because the wrong brush wears the ring and a worn ring is a far bigger job than a set of brushes. The rotor is driven by a motor and belt and has a mechanical lock for service. All of this rotates at high speed with high voltage on board, so everything in CT Service Level 1 lessons 2 and 3 about lock-out, discharge and radiation applies before the covers come off.
The couch
The Aquilion couch moves vertically and horizontally. The horizontal movement of the tabletop is the one that matters for the scan, because helical acquisition and reconstruction both depend on knowing exactly where the table is. On the models in our field cases the horizontal drive is an AC servo motor with a servo amplifier, a gearbox and an electromagnetic brake between motor and gearbox. As engineers describe it, position is read twice: by an encoder built into the motor and by a second encoder on the gearbox output. The servo amplifier compares what it commanded with what it measured, and the couch control board, named KGSTM in field reports on the Aquilion 16, coordinates the couch with the scan sequence.
This double measurement explains a whole class of couch faults on this family. If the brake slips, a belt stretches, a bearing tightens or an encoder gets dirty, the two readings disagree or the motor needs more current than expected, and the servo amplifier stops the couch with an error code. The operator sees a couch that jams at random, often only at high speed or only in one direction, and that works again after a power cycle. Lesson 4 follows that fault in detail.
The console side: S-con, D-con, reconstruction and raw data
The Aquilion console is not one computer. The software documentation describes a scan console (S-con) that runs the scan, controls the gantry and couch and hosts the scan-side settings, and a display console (D-con) for image processing and review. Both run the same base software and differ in configuration: host type, computer name and IP address are set during installation. They talk through a hub on a private network. The manual for V4.40 is explicit that during software loading each console must be cabled to the hub; a broken network at that moment can leave the system configuration disabled.
Behind the S-con sits the reconstruction box (REC BOX), which contains the reconstruction hardware and a server PC that the manual calls the RTM, with a hardware inventory file it needs in order to start. Raw data from the gantry are written to the raw data disk (RDD), a hardware RAID connected by Fibre Channel, which is why a Fibre Channel driver is reinstalled with the software. Reconstructed images go to image disks: on Aquilion ONE they sit in a RAID inside the PC BOX. On older Aquilion 64 systems, field reports describe a separate image server with its own disk array. A gantry-side touch display unit, called i Station in the software, has its own resource and configuration files.
Knowing these boxes by name saves hours. A message that says the file system cannot run, a study stuck with HOLD in the transfer list, or a scan that never reaches reconstruction each points at one of them: image disk and file system, the link between reconstruction and console, or the RDD and its parameters.
Follow the signal, and what is specific to this family
Trace one scan through the machine. The S-con prepares the scan and sends parameters to the gantry. The XC brings up the filament and the starter, then the generator applies kV and mA. X-rays pass the collimator and the patient and reach the detector. The DAS digitises every channel; the data transmission unit sends the stream off the rotor; fibre carries it to the REC BOX, where raw data go to the RDD and into reconstruction. Images are written to the image disk and shown on the consoles, then sent by DICOM to PACS. Power runs the other way: from the power distributor through the slip ring to the rotor supplies.
Three things make this family different from the generic CT. First, its error reporting is organised by subsystem and by board, so learn the names XC, DAS, RDD, REC, S-con and D-con before you read a log. Second, its calibration data, scan protocols (eXam Plans) and site settings are stored as named data sets that a backup tool can save and restore; losing them is the most expensive mistake you can make on this machine, and lesson 2 shows how to avoid it. Third, many of its disks, power supplies and memory modules are specific, and generic replacements are often rejected or only work after compromise; use qualified parts.
A last point about history. Several Aquilion problems were addressed by the manufacturer through field corrections: a heat exchanger oil leak on the Prime, an optical conversion connector that broke communication between units on Aquilion ONE type TSX-301A/B systems, a SureStart scan delay in early Aquilion 32/64 software. Before you diagnose a known symptom, check whether the machine received the corrections that apply to its type and serial number.
| Part | What it does | How it fails | |
|---|---|---|---|
| 01 | X-ray tube with housing and heat exchanger | Produces the beam; oil carries heat to the exchanger | Arcing, filament or rotor faults, oil seep, overheating |
| 02 | High-voltage generator and multiplier | Turns rotor power into tube kV and mA | Over kV or over mA trips, failed small supply on the multiplier |
| 03 | X-ray controller (XC) and starter | Controls filament, anode rotation and exposure | XC errors, filament timeout, starter communication error |
| 04 | Slip ring and brush blocks | Carry power to the rotor | Worn or wrong brushes, carbon dust, noise, intermittent power |
| 05 | Detector, DAS and collimator | Measure the beam and digitise every channel | Rings, bands, noise; supply ripple; board failure |
| 06 | Data transmission unit and fibre | Send raw data from rotor to reconstruction | Damaged fibre or connector: noise mid-scan, lost data |
| 07 | Couch servo drive, brake and encoders | Move and position the tabletop | Servo errors, encoder mismatch, brake slip, tight bearings |
| 08 | S-con and D-con | Run scans, processing, review and settings | Start-up failure, disk faults, lost configuration |
| 09 | REC BOX with RTM server | Reconstructs images from raw data | Does not start without its inventory file; board faults |
| 10 | Raw data disk (RDD) and image disks | Store raw data and reconstructed images in RAID | Degraded array, incompatible replacement disk |
Aquilion 64: the scan that breaks up halfway
An Aquilion 64 produced normal scanograms, but every helical scan went wrong in the same way: about half the images were good, then the rest were covered in noise and streaks. The air calibration made no difference and no tube or generator error was logged. A colleague who had seen the same pattern on another unit pointed at the data path rather than the detector: the fibre carrying raw data from the gantry to the reconstruction side. The engineer inspected the optical cables and connectors along their route and found the damage. The detector and DAS were never touched.
When images degrade partway through an acquisition while the scanogram is fine and no X-ray error appears, think of the data link before the detector. Know the route of the fibre and its connectors.
Aquilion 32: the transfer that holds
On an Aquilion 32, contrast studies stopped transferring between the computers. The remaining series counter showed HOLD and nothing moved, while plain studies went through. The engineer first framed it as a problem between two consoles. Colleagues who knew the platform corrected the picture: on this system the images do not really move from one console to the other; they come from the reconstruction unit to the scan console. Large, multi-phase contrast studies loaded the reconstruction unit hardest, and that was where the fault had to be looked for. The discussion moved from network settings to the health of the reconstruction hardware.
Draw the data path before you troubleshoot. If you misplace the box where the data come from, you will test the wrong link.
On-site checklist
- Data plate read: type number with suffix, serial number, date
- System software version and release suffix noted from start-up
- Tube and generator type in the system configuration noted
- Console arrangement noted: S-con, D-con, REC BOX, RDD, image disks or image server
- Applicable service manual volumes and software media identified for this type number
- Field corrections and recalls checked for this type and serial number