Qure.ai's portfolio of head CT scan tools* includes a triage aid to prioritize and notify critical head CT scans, a TBI progress monitoring tool, and a reporting assistance mode that that pre-populates radiologist templates.
Radiologists and emergency care clinicians are alerted by a 'Telegram' notification as soon as a head CT scan with a critical abnormality is detected. The notification, coupled with non-diagnostic preview images sent to treating physicians’ mobile phones helps ensure that scans are reviewed as soon as possible.Learn More
Radiology worklists can prove complex to manage, especially in tele-radiology settings or imaging centers with high scan volume. ‘STAT’ designations coded at the time of ordering the CT scan don’t always reflect the degree of severity of the abnormality on the image.
The qER prioritization tool helps meet stroke and trauma reporting standards and maintain consistent turn-around times for critical abnormalities.Learn More
A full text description on the abnormalities is generated and used to pre-populate the radiology reporting template, saving dictation time. The text includes the name and nature of at the abnormality detected, its anatomical location within the brain, as well as its severity and extent.Learn More
Qure.ai's deep learning algorithms quantify the volume of intracranial bleeds, infarcts and brain ventricles at a level of precision that surpasses radiologists.
This capability is used by clinicians to track the progress of patients with traumatic brain injury, infarcts and hydrocephalus and by researchers to develop new quantitative outcome measures.Learn More
Qure.ai’s deep learning algorithms detect, localise and quantify a growing list of brain pathologies including intra-cerebral bleeds and their subtypes, infarcts, mass effect, midline shift, and cranial fractures.
In October 2018, a study validating Qure.ai's head CT scan algorithms was published in The Lancet, evaluating performance on detecting intracranial bleeds, fractures, mass effect and midline shift.
The study measured algorithm accuracy versus a 3-radiologist majority on 500 scans and an additional 25,000-scan validation dataset, showing that qER is able to detect these critical abnormalities with near-radiologist accuracy.
We have made the CQ500 dataset publicly available so that others can test their algorithms and build upon our results. We provide anonymized dicoms and the corresponding radiologist reads for the published validation set.
The list of abnormalities that the deep learning algorithms can detect has grown since the results were first published in The Lancet, and algorithm accuracy is now even higher.
The accuracy of each algorithm is reported individually below, using radiologist opinion as ground truth. By altering the detection threshold, each algorithm can be operated at either a high-sensitivity or high-specificity operating point depending on the clinical setting.
|Abnormal finding||AUC (Confidence interval)||Operating point 1 (sensitive)||Operating point 2 (specific)|
|Intracranial haemorrhage (all 5 types)||0.95 (0.95 - 0.96)||0.9||0.89||0.85||0.96|
|Intraparenchymal haemorrhage||0.95 (0.94 - 0.96)||0.9||0.86||0.85||0.93|
|Subarachnoid hemorrhage||0.95 (0.94 - 0.96)||0.9||0.89||0.85||0.93|
|Subdural hemorrhage||0.96 (0.95 - 0.97)||0.9||0.89||0.84||0.95|
|Extradural hemorrhage||0.97 (0.96 - 0.98)||0.95||0.87||0.9||0.93|
|Intraventricular haemorrhage||0.98 (0.97 - 0.99)||0.95||0.91||0.9||0.97|
|Infarct||0.95 (0.94 - 0.95)||0.9||0.85||0.8||0.93|
|Cranial Fracture||0.92 (0.91 - 0.93)||0.9||0.75||0.8||0.92|
|Midline Shift||0.96 (0.95 - 0.97)||0.9||0.94||0.9||0.94|
|Mass Effect||0.9419 (0.93 - 0.95)||0.9||0.86||0.8||0.95|
|Atrophy||0.92 (0.91 - 0.92)||0.9||0.79||0.8||0.86|
“It’s one of the best radiology–AI efforts to date, because it widens the deep learning interpretation task to urgent referral of many different types of head CT findings,” says Eric Topol, referring to Qure.ai's algorithms.Read Nature Medicine article
Two million cells a minute. That’s how quickly brain damage happen when the cells get no oxygen in a stroke or in some brain injuries. Both can have tragic consequences — paralysis, memory loss, speech difficulties and even death.
But doctors can’t start treatment without an initial diagnosis, and that requires reading a CT scan as soon as the test’s completed...Read article on NVIDIA blog
Write to us at firstname.lastname@example.org to request a demo.