Last month, we received our third FDA clearance within just 9 months for our AI solution for triaging cervical spine fractures. While I’m personally very proud of our team- this is truly a win for patients, radiologists and other physicians across the nation. At this moment, there are dozens of AI algorithms being built for healthcare. Yet, the critical question we have to ask ourselves is whether these algorithms have a true clinical impact?
I wanted to share a recent clinical story in which Aidoc’s C-Spine solution assisted in triaging a patient, ensuring them a faster course of treatment. This feedback is what makes us tick, and articulates the value one can get by using an AI triage solution.
The clinical workup
We deployed our cervical spine solution at a hospital in the Northeast on June 13th, 2019. They had been using our ICH solution for over 6 months already.
A day after the first go-live, we had the first positive C-spine case:
09:32: An unresponsive 85-year-old man with a prior medical history of diabetes, atrial fibrillation and hip fracture one year ago, is brought to the ED. He was last known well and conscious at 10:00 PM and was found unresponsive by his caretaker in the morning.
At the scene, EMS measured finger glucose of 35. There were also bruises over his left face, left chest and left thigh, as well as a urine pool next to him. The working hypothesis was a seizure due to hypoglycemia. EMS gave him glucose IV and brought him to the ER where his state improved.
09:41: On exam, the patient moves all limbs anti-gravity, but is still unresponsive.
09:46: The physician orders a trauma-protocol full-body CT scan (head, neck, chest abdomen pelvis)
10:01: Within four minutes of the scan leaving the scanner, Aidoc flags a subtle non-displaced c2 left pedicle fracture and the radiologist prioritizes the scan in their workflow, focusing their attention to dramatically reduce turnaround time.
10:09: The radiologist immediately reviews and reports the finding to the ED physician in less than 30 minutes after the scan is done. The physician initiates spinal precautions, repeats the neurologic exam, and consults neurosurgery for further management.
Subtle fracture of the left pedicle of C2
Unprecedented value in cervical spine fracture triage
This is a story of innovative technology and physicians coming together to improve the quality of treatment for a patient who needed immediate medical attention.
In this case, the AI helped one of the many members in this patient’s care team to identify a possible pathology and rapidly initiate a protocol to the patient’s benefit. Prioritization of this case allowed the prompt application of a cervical collar, preventing cervical movements which could expose the patient to worsening of this fracture, and even frank spine instability, which could lead to serious nerve damage.
The hospital where this took place now regularly uses our solution to help prevent similar problems. But this patient is not the only one who benefits in this story. I envision several downstream benefits which we have only begun to discover and quantify. These include improved potentially better outcomes due to timely treatment which could influence the length of stay and the treatment plan.
Receiving such feedback from our customers, like the story above, is not just our raison d’être, but it’s what keeps us moving forward, towards building technological solutions to some of the most painful bottlenecks in medicine today.