Define-AI Directory

Our use cases are scenarios where use of artificial intelligence (AI) may help improve medical imaging care.

They were created to empower AI developers to produce algorithms that are clinically relevant, ethical, and effective. Each use case provides narrative descriptions and flow charts which specify the health care goal of the algorithm, the required clinical input, how it should integrate into the clinical workflow and how it will interface with users and tools.

Our use cases help radiologists and allied professionals by ensuring that AI tools provide needed information, can be efficiently implemented into daily workflow, and have the potential to improve the quality and efficiency of patient care.

Image Interpretive

Abdominal


Acute Appendicitis
Detect acute appendicitis

Cholecystitis on CT
Detect cholecystitis on CT

Cholecystitis on US
Detect cholecystitis on US

Colon Polyp Detection
Detect polyps greater than six millimeters
#C-RADS

Diverticulitis Detection
Detect colonic diverticulitis on CT

Kidney Stone Detection on CT
To detect kidney, ureteral, or bladder stones on CT

Kidney Stone Detection on Ultrasound
To detect kidney, ureteral, or bladder stones on US

Choledocholithiasis Detection on Ultrasound
Detection of choledocholithiasis on ultrasound

Choledocholithiasis Detection on CT
To detect choledocholithiasis on CT

Choledocholithiasis detection on MRCP
To detect choledocholithiasis on MRCP

Detection of Retained Surgical Foreign Body
If counts are off after a surgical procedure or multiple surgical teams were involved in a case, a radiograph is used to assess a patient for a retained foreign body (a missing needle or surgical sponge).

Placement of Enteric Tube
To assess the position of enteric tube placement. Enteric tubes refer to support devices placed for feeding patients who cannot swallow or for decompressing the GI tract. The tip of these tubes needs to be in the correct location to function, and a misplaced tube that is used can injure the patient. A frequent indication of chest and abdominal radiographs is to check position of the tube and an algorithm that could determine placement would automate this task.

No Prep CT Colonography
Find colon cancers and polyps on CT colonography without the need for bowel prep. Fecal tagging if necessary.

Identifying Small Bowel Obstruction
Detect small bowel obstruction

Breast Imaging


Breast Cancer Risk Assessment
To create reliable standard for breast cancer risk based on multivariable information

Breast Density
To automate the assessment of breast density on digital mammography and digital breast tomosynthesis by developing an AI tool that: (1) assigns a BI-RADS breast composition category, (2) assigns a composite numeric breast density score, and (3) provides regional breast density numeric sub-scores.
#BI-RADS

Breast Lesion Malignancy Evaluation
To automate classification of breast lesions into probability/likelihood  of malignancy at time of screening

Calcification morphology and follow-up
Detect morphology of calcifications on a mammogram and recommend follow-up by risk stratification

Classifying High-Risk Lesions (LN)
To automate classification of mammographic LN into categories based on level of suspicion of upgrade to malignancy at the time of radiology-pathology correlation incorporating pathology results

Classifying High-Risk Lesions (FEA)
To automate classification of mammographic FEA into categories based on level of suspicion of upgrade to malignancy at the time of radiology-pathology correlation incorporating pathology results

Classifying Suspicious Microcalcifications
Automate classification of breast microcalcifications into categories based on level of suspicion at time of diagnostic mammography

Malignancy of breast lesions on diffusion weighted images
Detect breast lesions and determine malignancy of breast lesions on diffusion weighted images using ADC values

Molecular Breast Imaging
Automation to reduce the dose and acquisition time for Molecular Breast Imaging ( MBI) scans

Similar Lesion Suggestion Engine
Provide a type of 'reverse image' search for a highlighted lesion to provide the radiologist with similar appearing lesions and their pathologic outcomes.

Technical Repeats in Mammography Practices
Reduction of Technical Recalls in Mammography Practices 

Cardiac


Acute Aortic Syndrome
Detect acute aortic syndrome

Aortic Diameter Measurements
Measurements relating to the aorta on 3D CT reconstruction

Aortic Valve Analysis
Automate aortic valve measurements  

Ascending Aortic Diameter
Automate ascending aortic diameter measurement
#MPA:AO

Automated Evluation of Pulmonary Trunk and Artery Dimensions
An automated algorithm, based on the reference values, should identify patients showing dilatation or reduction of pulmonary arteries caliber

Calcium Scoring of the aortic valve before TAVR
Identify the aortic valve and measure the Calcium score, and suggest follow up care(TAVR vs. SAVR) in routine TAVR CT reports.

Cardiac Output
Calculate cardiac output
#LV Functional Assessment

Cardiomegaly
Detect cardiomegaly on posterioranterior and anteroposterior radiographs
#Cardiac Size on Chest Radiograph

Cardiothoracic Ratio
Measure cardiothoracic ratio on posterioranterior and anteroposterior radiographs
#Cardiac Size on Chest Radiograph

Carina Angle Measurement
Measure angle of carina
#Cardiac Size on Chest Radiograph

Coronary Calcium Detection
Detection and quantification of coronary calcium on CT images

Coronary CTA
Quantification of stenosis of the coronary arteries 

Coronary Flow Reserve on PET
Automate measurement of myocardial blood flow and coronary flow reserve (CFR) on PET

Detection of pulmonary nodules in CCTA acquisition
Identify the pulmonary nodules in a common CCTA

Evaluations of Cardiac Amyloid
Detection and quantification of cardiac amyloid deposits as well as subtyping

Flow in Ascending Aorta
Automate measurement of flow in ascending aorta
#Qp:Qs

Flow in Pulmonary Artery
Automate measurement of flow in pulmonary artery
#Qp:Qs

Hematocrit estimation for myocardial ECV
MR image-based synthetic hematocrit estimation for myocardial ECV

Left Atrial Enlargement
Detect left atrial enlargement
#Cardiac Size on Chest Radiograph

Left Atrial Size
Automate measurement of left atrial size on CT

Left Ventricle Late Gadolinium Enhancement Assessment for MR
Automate late gadolinium enhancement assessment

Left Ventricle Myocardial Mass
Measure myocardial mass on left ventricle
#LV Functional Assessment

Left Ventricle T1 Mapping Quantification
Report quantitative left ventricle myocardial T1 relaxation time, before and after IV contrast. Use these values to compute extracellular volume.

Left Ventricle T2 Mapping Quantification
Quantify T2 relaxation time of the myocardium of the left ventricle.

Left Ventricle Volume

Automate standard left ventricle volumetric measurements
#LV Functional Assessment

Left Ventricle Wall Motion
Automate standard left ventricle wall motion assessment
#LV Functional Assessment

Left Ventricle Wall Thickening
Automate left ventricle wall thickening assessment
#LV Functional Assessment

Left Ventricle Wall Thickness Measurement
Automate standard left ventricle wall thickness measurements
#LV Functional Assessment

Left Ventricular Late Gadolinium Enhancement Assessment for MR
Automate late gadolinium enhancement assessment

Myocardial Iron Quantification
Myocardial iron quantification 

Myocardial Perfusion Quantification for CT
Automatic quantification and classification of myocardial perfusion

Pulmonary Artery Diameter
Automate pulmonary artery size measurements
#MPA:AO

Pulmonary Artery to Aortic Diameter Ratio
Calculate MPa:Ao and determined if enlarged
#MPA:AO

Pulmonary to Systemic Flow Ratio
Automate measurement of pulmonary to systemic flow ratio (Qp:Qs)
#Qp:Qs  

Pulmonary Veins Mapping Preablation Planning for CT
Document pulmonary veins and anatomy for ablation planning on CT

Quantification of Epicardial Adipose Tissue on CT
Automatically quantify and segment a patient’s epicardial adipose tissue on CT

TAVR Abdomen and Pelvis Access
Quantification of access within the abdomen and pelvis for TAVR 

TAVR Aortic Root Measurements
Automate aortic root and cardiac measurements for transcatheter aortic valve replacement (TAVR) necessary to determine the size and type of valve best for the patient.

Volumetric Quantification of Pericardial Effusion on CCTA
Pericardial effusion is difficult to assess and may show inter-observer variability, depending on what location the effusion is measured. 

Interventional


Identification and Characterization of Vena Cava Filters
Tool to automatically identify and characterize (type, location, potential complications) vena cava filters on imaging studies.

Musculoskeletal


Accessory Muscles Causing Neurovascular Compromise
Identify the presence of an accessory muscle

Automated Inall-Salvati ratio
Provide an automated calculation of the ISI to determine Patella Alta or Patella Baja

Automated leg length discrepancy assessment on CT
Quantify the lower extremity limb lengths on CT scanograms

Automated leg length discrepancy assessment on Radiograph
Quantify the lower extremity limb lengths on standing radiographs

Buckle fracture of the distal radius on pediatric wrist radiograph
Detection of buckle fracture of distal radius on pediatric wrist radiograph

Cervical Rib Detection on CT
Cervical rib detection on CT

Cervical Rib Detection on X-Ray
Cervical rib detection on x-ray

Cervical Rib Detection on MRI
Cervical rib detection on MRI

Chrondral Bone Lesion Characterization
Provide a probability of a chondral bone lesion having benign vs. malignant potential

Femoro-acetabular impingement (alpha angle)- CT
An automated method for alpha angle measurement to assist in the determination of femoroacetabular impingement
#alpha-angle

Femoro-acetabular impingement (alpha angle)- MRI
An automated method for alpha angle measurement to assist in the determination of femoroacetabular impingement
#alpha-angle

Genu varus and valgus Quantification
Assess presence and quantify degree of genu varus or valgus

Hip Osteolysis
Measure osteolysis in the hip following total hip arthroplasty
#THA

Hip Subsidence

Measure subsidence in the hip following total hip arthroplasty
#THA

Ischiofemoral impingement
To assist radiologists in the detection of an anatomical disturbance of the hip

Ligamentum Teres Injury Detection
Ligamentum teres injury detection

Lipohemarthrosis of the knee
To assist radiologists in the detection of an anatomical disturbance of the knee

Odontoid Fracture
Detect odontoid fracture on cervical spine trauma radiographs

Osteochrondritis Dissecans
Determine stability of osteochondritis dissecans

Patellar Tendon-Lateral Femoral Condyle Friction Syndrome Detection
Identify findings associated with patellar-tendon-lateral femoral condyle friction syndrome

Pediatric Elbow
Assist radiologists in the interpretation of pediatric elbow radiographs

Periprosthetic Hip Fracture
Detect periprosthetic hip fracture

Periprosthetic Hip Lucency
Measure periprosthetic lucency in the hip
#THA

Radial Head Dislocation
Detect radial head dislocation

Slipped Capital Femoral Epiphysis
Assist radiologists in the interpretation of pediatric hip radiographs

Snapping hip syndrome MRI
Iliopsoas tendinopathy detection

Soft-Tissue Tumor Bed Size Change
Detect and quantify volumetric changes of a soft-tissue tumor

Stener Lesion
Detect status of ulnal collateral ligament and presence of a Stener lesion

Tarsal Coalition

Detect, characterize, and classify tarsal coalitions

Tarsometatarsal Joint Status
Detect if Lisfranc joint is normal or abnormal

Trauma Fracture Detection
Detect all fractures on CT chest, abdomen, and pelvis performed for trauma

TT-TG distance
Provide automated measurement of common metric for assessment of patellofemoral instability

Automated Measurements of Foramina on CT
Create automated measures of foraminal stenosis on cross-sectional imaging studies of the lumbar spine

Automated Measurements of Lumbar Spinal Canal on CT 
Create automated measures of canal stenosis on cross-sectional imaging studies of the lumbar spine

Multiple Myeloma Evaluation on Total Body CT 
Detection and quantification of osteolytic lesions throughout the skeleton caused by Multiple Myeloma (Kahler's disease)

Nonaccidental Trauma and Child Abuse
Quantify the probability of nonaccidental trauma combining imaging (Neuroradiology and Skeletal Imaging) with clinical information to aid in diagnostic accuracy for radiologists and emergency room clinicians in the community setting without subspecialty training in pediatric imaging or pediatric emergency medicine.

Neurology


Aging Brain - Dementia
Detecting different patterns of brain atrophy which can be helpful for diagnosis the underlying etiology of dementia

ASPECT Score Calculation for Acute Infarction
Detect and define ASPECT Score in setting of acute infarction

Compression fracture detection
Describe spinal compression fracture

Cord Compression
Detect if spinal cord compression is present

Detection of acute infarct on DWI images
Detect acute infarcts on diffusion-weighted images (DWI) on brain MRI. Another goal is to simplify the language while reporting the lesions with diffusion restriction.

Epidural hematoma
Detect, classify and measure the size of an epidural hematoma

Focal Cortical Dysplasia
Detect focal cortical dysplasias on MRI scans of the brain in children

Glioma and other primary CNS neoplasms
Estimation the histopathological type, grade, gene mutation and outcome of gliomas and other primary CNS neoplasms from MRI

Lesion Detection on Diffusion Weighted Imaging
Early detection of DWI restricted lesions and notification

Midline Shift
Detect and measure midline shift

Motor Cortex Quantitative Susceptibility Mapping
Quantification of QSM of motor cortex and segmentation

Reduce MS surveillance MRI scan time
Patients with Multiple Sclerosis (MS) undergo multiple surveillance MRIs, many with contrast, throughout their lifetime. Real-time AI should be able to reduce scan time without compromising necessary information for treatment.

Subarachnoid hemorrhage
Detect, classify and measure the size of subarachnoid hemorrhage

Subdural hematoma on MRI
Detect, classify and measure the size of a subdural hematoma or collection on MRI

Subdural hematoma on CT
Detect, classify and measure the size of a subdural hematoma or collection on MRI

Oncology


Bony Metastatic Disease Detection
Detect changes in the bones during cancer staging  

Extranodal Extension
Detect/delineate lymph node involvement and extranodal extension on cross-sectional images

Pediatric


Craniosynostosis
Identify premature closure of cranial sutures

Detect injuries correlated with child abuse: metaphyseal fractures
Detect metaphyseal fractures of long bones which is an injury correlated with Nonaccidental Trauma (NAT)

Detect injuries correlated with child abuse: posterior rib fractures
Detect posterior rib fractures which are injuries correlated with Nonaccidental Trauma (NAT)

Pneumotosis plain film neonate
Identify findings of necrotizing enterocolitis

Scoliosis
Quantify the presence and change of scoliosis measurements

Shunt Detector
Detect if the ventricular shunt catheter is intact and does not contain kinks

Leg Length Discrepancy
Evaluate for discrepancy in leg length

Thoracic


COVID-19 Compatible Chest CT Pattern
Provide a likelihood of a diagnosis compatible with COVID-19

COVID-19 Compatible Chest Radiograph Pattern
Provide a likelihood of a diagnosis of COVID-19 compatible lung disease

Evaluating progression of lung masses
The increase, decrease, or stability of a known primary tumor is the starting point of all cancer staging exams.  The goal would be to develop an algorithm that given a cancer study could determine the dominant lesion.  Once that lesion is identified on the current study, It should be found on a comparison study and the volumes should be compared.

Fluid Status on Chest Xray
Determine if the more recent study has increased or decreased fluid in chest radiographs

Incidental Pulmonary Nodules on Chest Radiographs
Detection and characterization of incidental pulmonary nodules on chest radiographs (CXR). These are nodules that are detected on CXRs performed for other reasons than lung cancer screening. 

Incidental Pulmonary Nodules on CT
Detection and characterization of incidental pulmonary nodules on computed tomography (CT)

Pneumonia
Detection of pneumonia on chest radiograph in adults

Pneumothorax
Detection and quantification of pneumothorax on chest radiograph

Pulmonary Embolism
Detection and assessment of pulmonary embolism on CT imaging

Pulmonary parenchymal emphysema
Detect and quantitate the type and amount of emphysema in each of the five lobes of the lung on thoracic CT scans.

Rib Fracture
Identification and enumeration of rib fractures

Tuberculosis Screening
Screen detection of tuberculosis on chest radiograph, in asymptomatic adults

Tuberculosis Triage
Detection and characterization of tuberculosis on chest radiograph, in symptomatic adults

Placement of Endotracheal (ET) Tube
When patients are intubated, it is critical for successful ventilation that the tube tip is within the mid trachea.  ET Tubes that are too low or too high can lead to serious complications such as under ventilation, cuff leak, or tracheal injury and therefore need to be repositioned and reimaged.

Non-Interpretive

Business Facing


Automated Follow-up Program
Detect and extract follow-up recommendations from radiology reports, identify and automatically close the loop when an appropriate follow-up examination has been performed, and notify assigned user(s) when the required follow-up examination has not been done within a pre-specified time frame.

Computerized auto-coding of reports with real-time dictation feedback
Develop a system that can automatically add appropriate codes for billing (both CPT & ICD-10) and provide real-time notifications for any missing key elements required for billing.

Intelligent routing of exams for optimal performance
Increase the number of sub-specialized interpretations,reduce turnaround time, and improve the work balance of radiologists by intelligently routing studies based on optimized staffing levels as well as radiologist training and expertise. The algorithm may also suggest when to request back-up assistance based on predicted workload and staffing.

Predict Patient No-Shows For Radiology Appointments
Prediction of no-shows for radiology appointments

Predicting Volume to Optimize Staffing
By predicting the number and types of exams that are likely to be encountered in a given time period, the system allows for appropriate staffing in terms of number and types of radiologists. This should reduce turnaround time and improve work balance of radiologists. Algorithm may also suggest when to request back-up assistance based on predicted workload and staffing. 
#Worklist-Management

Reconciling discrepancies between payments from insurance to contracted rates
Reconciling discrepancies between payments from insurance companies and previously contracted rates

Patient Facing


Chatbot to answer radiology based procedure patient questions (breast imaging)
Facilitate patient education and patient - radiologist interaction by providing interactive patient education

Computer aided translation of radiology reports (thyroid ultrasound) to lay person
Encourage development & implementation of patient friendly radiology reports for thyroid ultrasound

Information Delivery on Incidental Findings (Pulmonary Nodules)
Deliver accessible information to the patient and care team on incidental findings, such as a pulmonary nodules, and subsequent recommendations

Produce multi-media reports that are easier to understand
To automatically create a radiology report with embedded images and videos based on a radiologist’s plain text report.

Update Patient on Wait Times
Automated updates of exam start times to be communicated to waiting patients

Population Health Facing


Decreasing Variability In Follow-Up Recommendations (Incidental Thyroid Nodules)
To develop a system that facilitates the use of the lexicon and follow-up guidelines for incidental thyroid nodules, to drive consistency in radiology reporting and adherence to national guidelines. 
#TI-RADS #Structured-Reporting

Reading Room Facing


Automated Cross-Sectional Co-Registration
Accurate co-registration of imaging data sets both within and between patient studies.

Prioritization of Exams on the Worklist
Intelligently prioritize studies based on the acuity of findings, patient condition, and turn-around time standards 

Radiology and Pathology Report Correlation
Correlation of Radiology and Pathology reports 

Virtual Transcriptionist/Dictation Assistant
Provide a more intelligent speech recognition through AI-assisted virtual transcriptionist/dictation assistant that can address common errors that occur within radiology reports, including age, sex/gender, laterality, speech-recognition errors, and affirmative/negative correlation between findings and impression

Technologist Facing


Detecting image quality in medical imaging
Detect image quality in real time, it can warn technologists of suboptimal study before patient leaves and study can be re-done in time.
 

COVID-19


Actionable Finding Follow-Up 
Identify overdue recommendations for follow-up from radiology reports
#Actionable finding #critical result #incidental finding #follow-up

Breast Imaging Patient Triage
Risk stratify patients in a breast imaging screening workflow for triaged scheduling based on their risk of breast malignancy particularly in a time of limited clinical availability due to COVID-19 limitations.
#Breast cancer #screening #scheduling #risk assessment #mammography

CT Lung Screening Patient Triage
Risk stratify individuals in a CT lung screening program with delayed screening and follow-up due to COVID-19.
#Lung screening #patient triage

COVID-19 Surge Planning
Predict imaging volume based upon community data regarding COVID-19.

COVID-19 Risk Profile Assessment
Predict the risk of COVID-19 infection in patients during the imaging scheduling process through the analysis of clinical symptoms and environmental and occupational risk factors. Staff would then be alerted of patients at high-risk for COVID-19 infection who are being scheduled for upcoming imaging. This risk information would then be used to ensure that high-risk patients are scheduled at specific times, at specific facilities, and on specific reserved equipment and that any staff handling these patients takes appropriate protective precautions.