Waveform visualization tool for facilitating medical diagnosis
A system includes a processing circuit configured to receive signals corresponding ultrasound data, extract a blood flow waveform from the signals, the blood flow waveform corresponds to a single pulse of the signals, determine a curvature characteristic of the blood flow waveform based on a plurality of local curvature parameters, and identify a medical condition for the blood flow waveform using the blood flow waveform. Each of the plurality of local curvature parameters indicates a degree to which the blood flow waveform deviates from a straight line at a location on the blood flow waveform.
1. A system, comprising:
a processing circuit configured to:
receive, from a headset device, signals corresponding to ultrasound data, the signals acquired by a transducer of a probe of the headset device;
extract a blood flow waveform from the signals, the blood flow waveform corresponds to a single pulse of the signals;
enhance at least one sharp upslope defining a start of the blood flow waveform;
identify one or more window locations within the enhanced blood flow waveform, the one or more window locations associated with pulse onset and determined by:
applying a threshold to the enhanced blood flow waveform;
determining at least one threshold crossing point of the enhanced blood flow waveform; and
enforcing a refractory period after determining the threshold crossing point, wherein the refractory period disallows determination of further threshold crossing points within the refractory period;
determine, within the one or more window locations, a curvature characteristic of the enhanced blood flow waveform based on a plurality of local curvature parameters, each of the plurality of local curvature parameters indicating a degree to which the enhanced blood flow waveform deviates from a straight line at a location on the enhanced blood flow waveform;
identify, based on the curvature characteristic of the enhanced blood flow waveform, a medical condition for the blood flow waveform; and
control an output device to display the enhanced blood flow waveform comprising a graphical indication indicating at least one of the one or more window locations associated with the identification of the medical condition.
2. The system of claim 1 , further comprising:
an ultrasound device configured to output the signals corresponding to the ultrasound data, wherein the signals corresponding to the ultrasound data is received by the processing circuit from the ultrasound device, wherein the processing circuit is further configured to:
determine morphological attributes of the blood flow waveform; and
determine whether misalignment of the ultrasound device with respect to a subject has occurred based on the morphological attributes by using a classifier to classify the morphological attributes as one of a plurality of categories, the one of the plurality of categories corresponds to a misalignment issue type.
3. The system of claim 1 , wherein the processing circuit is configured to extract the blood flow waveform by:
dividing the signals into a plurality of blood flow waveforms, each of which corresponds to a pulse of a subject; and
deriving the blood flow waveform from the plurality of blood flow waveforms.
4. The system of claim 1 , wherein the curvature characteristic is a single curvature metric determined by summing the plurality of local curvature parameters.
5. The system of claim 1 , wherein each of the plurality of local curvature parameters is determined based on a second derivative of the blood flow waveform at the location.
6. The system of claim 1 , wherein:
each the plurality of local curvature parameters is above a beat canopy threshold of the blood flow waveform; and
the beat canopy threshold comprises a percentage of a total minimum-maximum range of the blood flow waveform.
7. The system of claim 6 , wherein the total minimum-maximum range of the blood flow waveform corresponds to a diastolic-systolic range of the blood flow waveform.
8. The system of claim 1 , wherein identifying the medical condition comprises:
selecting one of a plurality of curvature templates from a database, the one of the plurality of curvature templates corresponds to the medical condition; and
comparing the curvature characteristic with the selected one of the plurality of curvature templates.
9. The system of claim 8 , wherein the one of the curvature templates is selected using at least one attribute of a subject.
10. The system of claim 8 , wherein identifying the medical condition for the blood flow waveform comprises determining a likelihood of the medical condition based on comparing the curvature characteristic with the selected one of the plurality of curvature templates.
11. The system of claim 8 , wherein the medical condition associated with the selected one of the plurality of curvature templates comprises large vessel occlusion.
12. The system of claim 1 , wherein identifying the medical condition comprises using a classifier to classify the curvature characteristic as one of a plurality of categories, the one of the plurality of categories corresponds to the medical condition.
13. The system of claim 12 , wherein the plurality of categories are determined using learning examples, each of the learning examples comprising curvature and locations of the curvature as correlated with a respective medical condition.
14. The system of claim 12 , wherein the classifier comprises a support vector machine (SVM) or a spectral regression kernel discriminant analysis (SR-KDA).
15. The system of claim 1 , wherein the processing circuit is further configured to:
send a command to a medical device to automatically administer a drug to treat the medical condition in response to identifying the medical condition.
16. The system of claim 1 , further comprising the output device.
17. The system of claim 1 , wherein the probe is configured to automatically locate a middle cerebral artery of a subject wearing the headset device.
18. The system of claim 1 , wherein the processing circuit is further configured to:
apply a bandpass filter to the signals.
19. A method, comprising:
receiving, from a headset device, signals corresponding to ultrasound data, the signals acquired by a transducer of a probe of the headset device;
extracting a blood flow waveform from the signals, the blood flow waveform corresponds to a single pulse of the signals;
enhancing at least one sharp upslope defining a start of the blood flow waveform;
identifying one or more window locations within the enhanced blood flow waveform, the one or more window locations associated with pulse onset and determined by:
applying a threshold to the enhanced blood flow waveform;
determining at least one threshold crossing point of the enhanced blood flow waveform; and
enforcing a refractory period after determining the threshold crossing point, wherein the refractory period disallows determination of further threshold crossing points within the refractory period;
determining, within the one or more window locations, a curvature characteristic of the enhanced blood flow waveform based on a plurality of local curvature parameters, each of the plurality of local curvature parameters indicating a degree to which the enhanced blood flow waveform deviates from a straight line at a location on the enhanced blood flow waveform;
identifying, based on the curvature characteristic of the enhanced blood flow waveform, a medical condition for the blood flow waveform; and
controlling an output device to display the enhanced blood flow waveform comprising a graphical indication indicating at least one of the one or more window locations associated with the identification of the medical condition.
20. A non-transitory processor-readable medium storing processor readable instructions such that, when executed, causes a processor to:
receiving, from a headset device, signals corresponding to ultrasound data, the signals acquired by a transducer of a probe of the headset device;
extract a blood flow waveform from the signals, the blood flow waveform corresponds to a single pulse of the signals;
enhancing at least one sharp upslope defining a start of the blood flow waveform;
identifying one or more window locations within the enhanced blood flow waveform, the one or more window locations associated with pulse onset and determined by:
applying a threshold to the enhanced blood flow waveform;
determining at least one threshold crossing point of the enhanced blood flow waveform; and
enforcing a refractory period after determining the threshold crossing point, wherein the refractory period disallows determination of further threshold crossing points within the refractory period;
determining, within the one or more window locations, a curvature characteristic of the enhanced blood flow waveform based on a plurality of local curvature parameters, each of the plurality of local curvature parameters indicating a degree to which the enhanced blood flow waveform deviates from a straight line at a location on the enhanced blood flow waveform;
identify, based on the curvature characteristic of the enhanced blood flow waveform, a medical condition for the blood flow waveform; and
control an output device to display the enhanced blood flow waveform comprising a graphical indication indicating at least one of the one or more window locations associated with the identification of the medical condition.