Medical imaging device fault handling
Proposed are concepts pertaining to aiding fault diagnosis through the consideration of the relevance of different log files to a fault. In particular, embodiments of the invention propose generating predicted relevance scores for different log files, and/or content contained therein. The most relevant log file(s), or parts thereof, may then be prioritized for provision to an engineer, for example.
1 . A method for diagnosing a fault of a medical imaging device, the method comprising:
determining a relevance of log files to the fault of the medical imaging device, the medical imaging device comprising a plurality of subsystems, and each log file being associated with a respective subsystem of the medical imaging device, the determining a relevance of log files to the fault of the medical imaging device comprising:
obtaining initial problem input data relating to the fault of the medical imaging device;
performing a first natural language processing analysis on the initial problem input data to identify one or more features of the medical imaging device fault;
providing one or more features of the medical imaging device fault to a machine learning algorithm, the machine learning algorithm being trained to predict, for each of the plurality of subsystems of the medical imaging device, a fault probability indicating a likelihood that the fault of the medical imaging device originated from the subsystem;
obtaining a prediction result from the machine learning algorithm, the prediction result comprising a fault probability for each of the plurality of subsystems of the medical imaging device; and
determining, for content of each log file, a relevance score based on the obtained prediction result; and
wherein the method for diagnosing the fault of the medical imaging device further comprises:
generating a prioritized list of the log files based on the determined relevance of the content of a plurality of log files.
2 . The method of claim 1 , wherein the method for diagnosing the fault of the medical imaging device further comprises:
transmitting the log files in the prioritized list over a data communication channel to a remote processing device, and in a priority order defined by the prioritized list, the remote processing device being configured for analysing the transmitted log files and for diagnosing the fault of the medical imaging device based on a result of the analysing.
3 . The method of claim 1 , wherein obtaining initial problem input data comprises:
receiving, via in input interface, initial problem input data provided by a respondent in response to a request for data.
4 . The method of claim 3 , wherein the request for data comprises a fault analysis questionnaire.
5 . The method of claim 1 , wherein the machine learning algorithm is trained using a training algorithm configured to receive an array of training inputs and respective known outputs, wherein a training input comprises one or more features of a fault of a medical imaging device and respective known output comprises, for each of the plurality of subsystems of the medical imaging device, a fault probability indicating a likelihood that the fault of medical imaging device originated from the subsystem.
6 . The method of claim 1 , wherein the machine learning algorithm comprises a binary classifier trained to make a decision about the fault of the medical imaging device originating in a software subsystem or a hardware subsystem.
7 . The method of any of claim 1 , wherein the step of determining, for content of each log file, a relevance score comprises:
determining, for content of each log file, system relevance scores indicating a relevance of the log file content to each of the subsystems;
determining, for each subsystem, an error probability indicating a probability that the fault of the medical imaging device relates to the subsystem; and
processing the system relevance scores and error probabilities to obtain, for content of each log file, a relevance score.
8 . The method of claim 7 wherein the processing comprises:
processing the system relevance scores and error probabilities based on the following equation:
R
(
e
,
m
,
f
)
=
R
comp
(
f
)
P
comp
(
e
,
m
)
+
∑
i
=
1
n
R
sub
i
(
f
)
P
sub
i
(
e
,
m
)
wherein: R(e,m,f) is the relevance score of a log file f for the medical imaging device m with the fault e; R comp (f) indicates how relevant the log file f is for a component related fault; R sub i (f) indicates how relevant the log file f is for a fault in a subsystem; P comp (e,m) is the probability of the fault e being component-related; and P sub i (e,m) is the probability of the fault e being related to a subsystem.
9 . A method for determining severity of a fault of a medical imaging device comprising a plurality of subsystems, the method comprising:
obtaining initial input data from a user relating to the fault of the medical imaging device;
performing a first natural language processing analysis on the initial input data to determine a first severity score relating to a description of the fault;
analysing the initial input data and the plurality of subsystems to determine a second severity score relating to the subsystems of the medical imaging device; and
determining an overall severity score indicating a severity of the fault based on the first severity score and the second severity score;
wherein the method further comprises for a plurality of log files associated with respective subsystems of the medical imaging device, generating a prioritized list of the plurality of log files based on the determined severity of the fault.
10 . The method of claim 9 , further comprising:
determining a third severity score relating to a priority of a user of the medical imaging device,
and wherein determining an overall severity score is further based on the third severity score.
11 . The method of claim 9 , further comprising:
applying a weighting to the overall severity score based on at least one of: the first severity score; the second severity score; and a parameter value of a log file identified for the fault.
12 . A method for diagnosing a fault of a medical imaging device, the method comprising:
determining the relevance of log file content to a fault of a medical imaging device according to claim 1 ;
determining a severity of the fault of the medical imaging device; and
generating a prioritized list of the log files based on the determined relevance of the plurality of log files and the determined severity of the fault of the medical imaging device, and preferably further based on a parameter value of each log file, such as log file size.
13 . A computer program comprising code configured to implement the method of claim 1 when executed on a processing system.
14 . A system for determining a relevance of log files to a fault of a medical imaging device, the medical imaging device comprising a plurality of subsystems, and each log file being associated with a respective subsystem of the medical imaging device, the system comprising:
a data interface configured to obtain initial problem input data relating to the fault of the medical imaging device;
a natural language processing component configured to perform a first natural language processing analysis on the initial problem input data to identify one or more features of the medical imaging device fault; and
a machine learning algorithm configured to receive one or more features of the medical imaging device fault, the machine learning algorithm being trained to output a prediction result comprising a fault probability for each of the plurality of subsystems of the medical imaging device, a fault probability indicating a likelihood that the fault of the medical imaging device originated from the subsystem; and
a processor configured to determine, for each log file, a relevance score based on the obtained prediction result; and
wherein the processor is further configured to:
generate a prioritized list of the log files based on the determined relevance of a plurality of log files; and
transmit the log files in the prioritized list over a data communication channel to a remote processing device, and in a priority order defined by the prioritized list, the remote processing device being configured for analysing the transmitted log files and for diagnosing the fault of the medical imaging device based on a result of the analysing.