Facility

ELEKTA BEIJING MEDICAL SYSTEMS CO., LTD

110 recall(s) on record — most severe on record: Class II.

Key facts

110recalls · worst Class II

—inspections in our data

—warning letters in our data

Recalls are listed for the parent company: openFDA recall records do not name a facility. How we attribute records

FDA Establishment Identifier (FEI)	3009894282
Establishment type	Manufacture Medical Device for Another Party (Contract Manufacturer)
Location	No. 21, Chuang Xin Road, Science and Technology Park Beijing, CN 10220
Registration status	Active · expires 2026
Parent company	ELEKTA AB (owner-operator 8030893)
Brand names	EMLA (Elekta Evo); EMLA (VersaHD); EMLA (Elekta Harmony Pro); EMLA (Elekta Infintiy); EMLA (Elekta Harmony); EMLA (Elek

Activity by year

Recalls (110)

Devices registered here (1)

Accelerator, Linear, Medical
Product code IYE (opens in a new tab) · Class II · Radiology · 21 CFR 892.5050

Recalls & enforcement

20264 records

Class IIOngoing2026-05-21
Software may fail to load adapted baseline shift plan during treatment delivery when a communication error occurs.
Class IIOngoing2026-04-03
Due to safety concerns regarding the detachment of the tip of the intrauterine tube (IU tube) that may remain within a patient.
Class IIOngoing2026-03-19
Locking mechanisms may not properly secure to the Arc system, resulting in shifting of coordinates during procedures.
Class IIOngoing2026-02-18
Due to adapter not being properly aligned or locked in place.

20255 records

Class IIOngoing2025-12-14
Due to a manufacturing issue and electrical grounding of systems.
Class IIOngoing2025-12-12
Due to an error when defining a new stereotactic reference that users fail to first change an already defined obsolete reference to a pre-plan reference. This may result in a treatment plan that potentially irradiates the incorrect patient location if the error is not detected b…
Class IIOngoing2025-12-10
Using oncology information system that manages workflow may result in overtreatment if:Plan delivered with Particle Therapy IHE-RO TDW-II interface,With one field per fraction or last field partial treatment,Partial treatment occurs followed by treatment field interruptions with …
Class IIOngoing2025-09-30
Due to increased risk that paint flakes from the engravings may separate from the device and there is the possibility that the flake may inadvertently enter the wound
Class IIOngoing2025-01-15
Image offsets calculated by oncology Information System software may be incorrect for Linear Accelerator (LINACs) not characterized at IEC61217 scale.

20234 records

Class IIOngoing2023-10-25
Elekta has identified that if the microswitch within the middle arm is incorrectly setup, there is a potential for an uncontrolled extension of iViewGT / XVI detector arms. When the detector arm is fully deployed there is no risk of uncontrolled extension.
Class IIOngoing2023-05-05
A component within a limited number of Elekta Linacs has been realized to be non-compliant to IEC 60601
Class IIOngoing2023-04-06
There is a low possibility that an electrical connector in the MR gradient coil will overheat on Elekta Unity systems.
Class IIOngoing2023-02-28
Re-optimization, after adding contours without forced density outside the external structure, may result in inaccurate dose presentation.

20221 record

Class IIOngoing2022-12-05
A reconstruction error may occur when using the Catheter Bending functionality in Applicator Modeling or Implant Modeling

20216 records

Class IIOngoing2021-10-13
Under certain circumstances, there can be a difference between the planned and delivered dose of the two ovoid channels.
Class IITerminated2021-09-17
There is a risk that the Diode D1 in the Modulator can fail in a destructive manner when removed from the machine, and that the protective tube surrounding the diode may fail to contain the ejected matter.
Class IIOngoing2021-09-14
Contour changes can be saved on an unintended image set. In addition, these contour edits do not cause the dose to be frozen on the plans associated with that image set.
Class IIOngoing2021-09-02
A drug strength in MOSAIQ can be changed during the ordering process and approved, but the saved drug strength reverts to the original value.
Class IIOngoing2021-07-26
If a couch is removed while a plan is loaded, and then added back to the Studyset when the plan is unloaded, it is possible to calculate dose with the couch unassigned to the beams.
Class IITerminated2021-04-20
Malware attack

20204 records

Class IITerminated2020-05-21
In TSM (Treatment Session Manager) Motion Monitoring workflows, under certain conditions there can be a mismatch between the contour data overlay with respect to the motion monitoring images of the monitored structure.
Class IITerminated2020-04-30
potential risk that Legionella may be present in the Unity machine room, specifically the heat exchanger condensation collection tray.
Class IIOngoing2020-04-16
The Monaco RTP Radiation Treatment Planning System may change the shape and volume of the contour potentially resulting in the device delivering an inaccurate dose.
Class IITerminated2020-03-30
A faulty locking piece of the Instrument Carrier that does not fulfil the requirement of locking force to the Arc has been found during an internal check of a Leksell Vantage Stereotactic Arc System.

201911 records

Class IITerminated2019-12-18
There have been reports of excessive heating in the MRL Gradient Connection Module.
Class IIOngoing2019-11-22
Monaco is using the incorrect energy when optimizing and calculating dose.
Class IIOngoing2019-09-04
It is possible that the forced electron density settings will be changed for some structures unintentionally and this can result in incorrect dose calculation.
Class IITerminated2019-08-23
Monaco is displaying the Anatomy and Beam shift direction incorrectly on the Monaco Scan and Setup Reference Report and is DICOM exporting incorrect (Anatomy/Beam) shift directions.
Class IIOngoing2019-07-03
When creating 3D plans using either MU or Dose weighting modes, if the user changes the number of fractions, rescales the plan and then changes the wedge angle, the Monitor Units are scaled incorrectly.
Class IITerminated2019-06-24
The QA software solution to perform the MR to MV alignment check, does not display the stored MR to MV offset values. The user is unable to independently inspect the values during their QA.
Class IITerminated2019-06-19
Potential for iGUIDE to incorrectly monitor the 3D position.
Class IITerminated2019-04-03
There is no warning in the Elekta Unity manual for the administration of gadolinium-based or other contrast agents for imaging or image guidance purposes has not been validated.
Class IITerminated2019-03-12
Users need to be aware when using these protocols for daily on line plan adaptation that: 1) The images acquired using these protocols do not represent the average position of the anatomy during the respiratory motion cycle. The images are based on data acquired around full expi…
Class IITerminated2019-02-08
If Improve Target Dose was chosen as an optimization model in a previous treatment session, Monaco® will automatically use this optimization model again when proceeding with the online plan adaptation of a completion plan when it should not.
Class IICompleted2019-01-04
In rare circumstances, due to reasons such as inadequate hospital facility maintenance, water leakage, and weakening of speaker fixings during the lifetime of the system, the speakers can become dislodged and be pulled into the bore.

20184 records

Class IIOngoing2018-12-13
There is a potential that Wedge IDs were not included in the DICOM RT PLAN sent from the TPS, which will cause the field in MOSAIQ to be created with no wedge. Treatment of a field without the planned wedge will receive more dose than the treatment plan indicates.
Class IITerminated2018-12-04
If a Setup field (CT, kV, MV or MVCT) is created by COPYING a treatment field or another setup field, then the copied field will have the same isocenter as the parent and that isocenter may not be the correct isocenter for imaging when it is sent to the machine.
Class IITerminated2018-08-03
Order Status Remains Approved When Should Indicate Complete.
Class IITerminated2018-05-16
Upon completion of a 6D workflow, the HexaPOD evo Module may be in a tilted state if it is moved to the position used for 3D (non-iGUIDE) treatments resulting in incorrect patient positioning.

20178 records

Class IITerminated2017-12-01
There is a potential for an uncontrolled extension of iViewGT / XVI detector arms.
Class IITerminated2017-11-20
Potential for positional errors following automatic table movement (ATM).
Class IITerminated2017-09-01
Incorrect DICOM mapping of the exported collimator or couch angles from ERGO, which would lead to incorrect rotation of the collimator or couch when using a MLC device for planning.
Class IITerminated2017-06-30
The table shift in the vertical direction was inverted during the use of CMA (Couch Move Assistant). Therapy may be directed to an incorrect area.
Class IITerminated2017-06-21
Retention of Outdated Information in an Order Set. In the Care Plan Library, double clicking to add or delete a Component/Procedure to an Order Set does not trigger the update of the associated Care Plans.
Class IITerminated2017-06-19
Incorrect PEC values.
Class IITerminated2017-05-23
If the gantry module is replaced after original installation the configuration settings may be missing. The identified risk for this issue is electrical safety for technicians doing maintenance on the X-ray generator.
Class IITerminated2017-01-10
Incorrect dose after editing beam number an wedge angle.

201614 records

Class IITerminated2016-12-23
Cross profile for Varian 60 degree wedge shows "horns."
Class IITerminated2016-12-21
Incorrect Enhanced Dynamic Wedge (EDW) or Virtual Wedge (VW) Calculations.
Class IITerminated2016-12-02
Edits to Particle field definition parameters may not be saved when the field definition window is saved.
Class IITerminated2016-11-17
Contours are not handled correctly if the CT dataset contains slices with DICOM Z coordinates of xx.x50mm.
Class IITerminated2016-11-03
Incorrect Dose when using the reset function.
Class IITerminated2016-09-29
When creating 3D plans using either MU or Dose weighting modes, if the user changes the Physician's Intent Rx Dose and/or the number of fractions, and then modifies the wedge angle, the MU value is scaled incorrectly.
Class IITerminated2016-08-18
Potentially unrecognized incorrect position of the treatment couch in 3D workflow, i.e. the HexaPOD has not moved fully to the 3D position.
Class IITerminated2016-08-18
In a specific workflow where contours are edited (enlarged, moved, copied, etc.) and then the contour is deleted on some slices and saved in the same session, it is possible that the deleted contours are still present which would not be the intent of the planner.
Class IIOngoing2016-07-15
It is possible that a change to an Order Set will not be saved in the current open Care Plan. This can result in a wrong chemo dose in the order.
Class IITerminated2016-07-08
There have been instances in hospitals where the joint between the drive gear and the gantry base drive wheel has become loose, leading to fatigue failure of some of the securing bolts.
Class IITerminated2016-06-17
It is possible that that the selected Plan-of-the Day could be accidently changed by scrolling the mouse wheel up and down and inadvertently selecting another plan.
Class IITerminated2016-05-18
When DICOM exporting a 3D Monaco plan and the "Composite Field Sequencing" (CFS) checkbox is selected, the Y jaws will snap to the port shape when they should remain where they were defined.
Class IITerminated2016-05-12
If by mistake the initial pre-treatment imaging was performed BEFORE the HexaPOD was moved to the DRIVE (*START) position, iGUIDE offers the possibility to branch into a specific catch-up workflow, which was implemented to use the initial scan data.
Class IITerminated2016-04-19
The latches of the frame adapter can be locked even if the locating pins of the frame adapter is not inserted into the corresponding holes in the coordinate frame.

201511 records

Class IITerminated2015-12-18
When using the option "Tumor Overlap Fraction" in VMAT planning it has been observed that in rare cases the system does include an organ at risk as target volume. This could result in open MLC, and open jaws in areas away from the target volume.
Class IITerminated2015-12-09
Incorrect drug dosage due to "Age Limit" and patient weight data item issue.
Class IITerminated2015-11-24
First Databank (FDB) is a third party product incorporated into MOSAIQ. FDB routinely modifies drugs in their database to reflect market and formulary changes. In FDB Framework v2.0, those drugs are not being included in drug-drug, drug-allergy and duplicate drug therapy interact…
Class IITerminated2015-10-16
Unintended update of Dose and MU and Incorrect Assignment of Bolus.
Class IITerminated2015-09-15
Dose and MU are incorrect when CT images are viewed from the head, and, when using multiple prescriptions with forced densities.
Class IITerminated2015-08-31
Incorrect Treatment Delivery Using Third Party Fixed Wedges.
Class IITerminated2015-07-01
A problem exists in MOSAIQ resulting in the incorrect field size being sent to the treatment machine for stereotactic plans using cones.
Class IITerminated2015-06-01
Memory can become corrupted when creating a fused study via drag and drop in Leksell GammaPlan 10.2.
Class IITerminated2015-05-21
Incorrect dose calculation for Regions of Interest (ROIs) defined on a secondary image series.
Class IITerminated2015-01-29
It is possible to position the Treatment Table with errors greater than 5mm and no inhibits are displayed, if the positional sensors are not correctly installed during corrective maintenance. Clinical mistreatment may occur if a position error is not detected.
Class IITerminated2015-01-07
Elekta has identified a batch of securing bolts that fix the Gantry counter frame weight stack may fail.

201412 records

Class IITerminated2014-12-19
Incorrect CT to ED when using the Monaco Image Statistics Tool.
Class IITerminated2014-12-19
Incorrect CT to ED when using the Monaco Image Statistics Tool.
Class IITerminated2014-11-03
A problem can exist in MOSAIQ resulting in the display of incorrect numeric data due to a dose rounding error on printed reports.
Class IITerminated2014-10-06
Elekta has become aware that the latches may be locked before they have been fully turned resulting in a poorly locked frame.
Class IITerminated2014-08-18
Incorrect Delivery of Composite VMAT Fields - When DICOM is exporting an arc plan (VMAT, Dynamic Conformal or 3D Static Arcs) and using the Composite Field Sequencing (CFS) feature, the DICOM exported plan will not match the approved treatment plan.
Class IITerminated2014-08-18
When DICOM is exporting an arc plan (VMAT, Dynamic Conformal of 3D Static Arcs) and using the Composite Field Sequencing (CFS) feature, the DICOM exported plan will not match the approved treatment plan.
Class IITerminated2014-05-29
A bug in the software prevents the iGUIDE software from logging off.
Class IITerminated2014-05-29
MOSAIQ does not re-calculate the agent volume when the user changes the ordering dose (to other than what was calculated) prior to saving the order.
Class IITerminated2014-05-23
When using CMA, if the user hovers the mouse pointer exactly above the direction dropdowns and then uses the mouse thumbwheel or touchpad scroll bar, the direction selection might change when the intention was only to move the mouse to a different area of the screen.
Class IITerminated2014-04-23
XVI can incorrectly calculate the target position of the treatment table.
Class IITerminated2014-04-17
If screws are not properly tightened with a torque wrench they may become loose over time.
Class IITerminated2014-02-17
Dose and Monitor Unit values are not computed correctly when Elekta Motorized Wedges are used with Dynamic Conformal arcs.

201319 records

Class IITerminated2013-11-25
During the treatment table, gantry, and collimator calibration procedures, it is possible to type incorrect values or invert a value.
Class IITerminated2013-11-25
During the treatment table, gantry, and collimator calibration procedures, it is possible to type incorrect values or invert a value.
Class IITerminated2013-11-25
When the number of fractions is changed for a proton spot beam, the MU are recomputed and rounded if necessary, but the dose is not recalculated as it should be even after saving and re-opening the plan.
Class IITerminated2013-11-25
During the treatment table, gantry, and collimator calibration procedures, it is possible to type incorrect values or invert a value.
Class IITerminated2013-11-25
When bolus is present, the Effective Depth (with bolus) to the weight point returned on the Source Data Report is incorrect.
Class IITerminated2013-11-25
Potential collision risk when using XVI and external beam shaping devices.
Class IITerminated2013-11-25
XVI can incorrectly calculate the target position of the treatment table.
Class IITerminated2013-08-15
Clinical mistreatment may occur when there are errors in the clinical workflow, or settings are incorrect.
Class IITerminated2013-07-23
Potential for clinical errors.
Class IITerminated2013-07-02
Upgrade installation script is designed to overwrite the same pouch files in the Treatment Couch Library with the original default values.
Class IITerminated2013-05-28
Customers are resetting the default values outside recommended factory settings.
Class IITerminated2013-05-28
Customers are resetting the default values outside recommended factory settings.
Class IITerminated2013-05-22
Patient needs to be positioned as indicated by the treatment plan.
Class IITerminated2013-05-01
It is possible for an intermittent error to occur in the gantry sensor of the DMLC IV Systems
Class IITerminated2013-04-30
Unexpected movement during beam delivery may not be detected if there are incorrect static tolerances for Linac Geometric Axes loaded in Integrity R3.0.
Class IITerminated2013-03-28
The problem is that the "static tolerances" from the calibration files (loaded database) are used for beam delivery instead of the machine calculated values.
Class IITerminated2013-03-14
The Column rotation can accidentally be moved to a different position which in the locked position.
Class IITerminated2013-02-11
The beam is displayed at the "gantry" angle, but the dose is calculated at the "rotation_start" angle.
Class IITerminated2013-01-18
If the Apex collimator is incorrectly installed on the radiation head, a touchguard inhibit activates the view restrictions dialog box only shows the top 12 restrictions.

20127 records

Class IITerminated2012-12-19
The DICOM tags do not contain sufficient information for XVI to identify the different volumes that have the same name.
Class IITerminated2012-09-18
A patient was seriously mistreated after a user made and saved an inadvertent change to the definition of a treatment filed in MOSAIQ.
Class IITerminated2012-08-21
During the planning process the beam weights can be changed by the customers. When the beam weights are changed and renumbered and this happens during the dose calculation it is possible to save and reopen the plan without any error messages.
Class IITerminated2012-08-06
There have been a small number of instances in hospitals where the joint between the drive gear and the gantry base drive wheel has become loose, leading to fatigue failure of some of the securing bolts.
Class IITerminated2012-07-27
Information released with the product Agility did not originally contain information on compatibility with the add-on Dynamic Multileaf Collimator Apex.
Class IITerminated2012-06-21
Incorrect patient shift directions when the Setup reference dialog is printed out when the DICOM coordinates option is enabled for reports.
Class IITerminated2012-06-21
Incorrect patient shift directions when the Setup reference dialog is printed out when the DICOM coordinates option is enabled for reports.

Source: U.S. FDA via openFDA (opens in a new tab) (device registration & enforcement). Data as of 2026-10-06. Verify this facility record at FDA → (opens in a new tab)

Cite this record

RegWarden. “ELEKTA BEIJING MEDICAL SYSTEMS CO., LTD (FEI 3009894282) — FDA facility record.” Compiled from U.S. FDA public records (openFDA; FDA Data Dashboard). Data as of 2026-10-06. https://regwarden.com/facility/3009894282-elekta-beijing-medical-systems-co-ltd/

Not open yet

Alerts for ELEKTA BEIJING MEDICAL SYSTEMS CO., LTD

Watchlist alerts will email you when FDA publishes a new recall, inspection classification, or warning letter for a facility you follow. They aren't open to the public yet — leave your email and we'll tell you when they are.

Planned

Want a one-page facility brief?

A single-page summary of this facility's registration and recall history for diligence. Not selling it yet — register interest and we'll follow up.