AI powered whole body MRI screening
Systems and methods that leverage the power of artificial intelligence (AI) to enhance the process of whole-body MRI scanning. AI models optimize the acquisition protocol, resulting in shorter and more patient-friendly scan durations. Furthermore, AI models aid in the automatic interpretation of the imaging data, highlighting potential areas of concern and streamlining the diagnostic process.
1 . A method for Whole Body MRI Screening using AI powered models, the method comprising:
determining, by a first AI model, a personalized set of MR sequences for a single-session whole body MRI screening spanning from at least a patient's head to pelvis, wherein the personalized set of MR sequences is tailored for the patient based on at least one of the patient's age or prior medical history in addition to initial images and available clinical data;
performing the single-session whole body MRI screening for the patient comprising the personalized set of MR sequences; and
identifying, using one or more second AI models, one or more abnormal regions for the patient based on results of the single-session whole body MRI screening.
2 . The method of claim 1 , further comprising:
adjusting the whole body MRI screening based on real-time analysis, by the one or more second AI models, of real-time results of one or more of the personalized set of MR sequences.
3 . The method of claim 1 , further comprising:
identifying the one or more abnormal regions for the patient based on results from an initial sequence of the personalized set of MR sequences; and
selecting a new sequence not included in the personalized set of MR sequences based on the identification.
4 . The method of claim 1 , further comprising:
comparing results of the single-session whole body MRI screening with prior MRIs, CTs, or other images from the patient's history; and
identifying any changes or developments.
5 . The method of claim 4 , wherein comparing is performed using an autoencoder-based anomaly detection method.
6 . The method of claim 1 , further comprising:
generating a report based on the interpreted results.
7 . The method of claim 1 , wherein the first AI model and/or the one or more second AI models comprise machine trained neural networks.
8 . The method of claim 1 , wherein the one or more second AI models comprise at least one machine trained network trained for segmenting image data for a specific region of the patient.
9 . The method of claim 8 , wherein the at least one machine trained network comprises a machine trained segmentation network.