AI Program Could Aid Decision-Making in Medical Imaging
|
By HospiMedica International staff writers Posted on 24 Sep 2018 |

Image: The Explainable Artificial Intelligence (XAI) program can help with decision-making in various medical fields (Photo courtesy of Raytheon BBN Technologies).
Researchers are developing a first of its kind neural network that explains itself and could help with decision-making in the medical field, among others. Raytheon BBN Technologies (Cambridge, MA, USA) is developing the neural network under the Defense Research Project Agency's (DARPA) Explainable Artificial Intelligence program (XAI). The aim of the XAI program is to create a suite of machine learning techniques, which produce more explainable models while maintaining a high level of performance. It also aims to help human users understand, appropriately trust and effectively manage the emerging generation of artificially intelligent partners.
The Explainable Question Answering System (EQUAS) by Raytheon BBN will allow Artificial Intelligence (AI) programs to 'show their work,' increasing the human user's confidence in the machine's suggestions. EQUAS will show users which data mattered most in the AI decision-making process. Using a graphical interface, users can explore the system's recommendations and see why it chose one answer over another. Although the technology is still in its early phases of development, it has the potential to be used for a wide-range of applications. As the system is enhanced, EQUAS will be able to monitor itself and share factors that limit its ability to make reliable recommendations. This self-monitoring capability will help developers refine AI systems, allowing them to inject additional data or change how data is processed.
"A fully developed system like EQUAS could help with decision-making not only in DoD operations, but in a range of other applications like campus security, industrial operations and the medical field," said Bill Ferguson, lead scientist and EQUAS principal investigator at Raytheon BBN. "Say a doctor has an x-ray image of a lung and her AI system says that its cancer. She asks why and the system highlights what it thinks are suspicious shadows, which she had previously disregarded as artifacts of the X-ray process. Now the doctor can make the call – to diagnose, investigate further, or, if she still thinks the system is in error, to let it go."
Related Links:
Raytheon BBN Technologies
The Explainable Question Answering System (EQUAS) by Raytheon BBN will allow Artificial Intelligence (AI) programs to 'show their work,' increasing the human user's confidence in the machine's suggestions. EQUAS will show users which data mattered most in the AI decision-making process. Using a graphical interface, users can explore the system's recommendations and see why it chose one answer over another. Although the technology is still in its early phases of development, it has the potential to be used for a wide-range of applications. As the system is enhanced, EQUAS will be able to monitor itself and share factors that limit its ability to make reliable recommendations. This self-monitoring capability will help developers refine AI systems, allowing them to inject additional data or change how data is processed.
"A fully developed system like EQUAS could help with decision-making not only in DoD operations, but in a range of other applications like campus security, industrial operations and the medical field," said Bill Ferguson, lead scientist and EQUAS principal investigator at Raytheon BBN. "Say a doctor has an x-ray image of a lung and her AI system says that its cancer. She asks why and the system highlights what it thinks are suspicious shadows, which she had previously disregarded as artifacts of the X-ray process. Now the doctor can make the call – to diagnose, investigate further, or, if she still thinks the system is in error, to let it go."
Related Links:
Raytheon BBN Technologies
Latest AI News
- Explainable AI Tool Improves Brain Disorder Screening from 4D fMRI
- New Benchmark Highlights Evidence Gaps in AI Responses to Clinical Questions
- AI-Powered ECG Model Gains FDA Authorization for Emergency Cardiac Assessment
- AI Improves Non-Contrast CT Interpretation for Time-Sensitive Stroke Assessment
- Explainable AI Tool Identifies High-Risk Patients Before Heart Attack Reperfusion
- AI Tools Show Potential to Detect Subtle Signs of Interval Breast Cancer
- AI Framework Aims to Turn Medical Care from Reactive to Preventive
- AI Model Predicts Indwelling Catheter Use After Spinal Cord Injury
- AI Tool Uses Mammograms to Detect Common Cardiovascular Diseases
- AI Tool Predicts Post-Heart Attack Trajectories to Personalize Follow-Up
Channels
Artificial Intelligence
view channel
New Benchmark Highlights Evidence Gaps in AI Responses to Clinical Questions
Many clinical decisions still rely on limited evidence, while most artificial intelligence benchmarks overlook the patient context that shapes real-world care. General large language models can also struggle... Read more
Explainable AI Tool Improves Brain Disorder Screening from 4D fMRI
Functional magnetic resonance imaging (fMRI) captures changes in brain activity over time, but the resulting image sequences can be difficult to analyze. Screening for brain disorders requires assessing... Read moreCritical Care
view channel
Early Stem Cell Transplantation May Improve Survival in Rare Inherited Bone Marrow Disorder
Shwachman-Diamond syndrome (SDS) is a rare inherited bone marrow disorder that can progress to severe hematologic disease. More than 90% of affected children survive beyond age 20, but fewer than 30% live... Read more
Drone AED Delivery Could Shorten Time to Out-of-Hospital Cardiac Arrest Defibrillation
Out-of-hospital cardiac arrest requires prompt cardiopulmonary resuscitation and defibrillation. Yet automated external defibrillators (AEDs) may be difficult to reach quickly, particularly in rural areas... Read moreSurgical Techniques
view channel
Labor Induction Guide Improves Patient Satisfaction in Randomized Trial
Labor induction uses medications or other methods to start labor, but patients may encounter information that conflicts with clinical guidance. This misinformation can create confusion and mistrust when... Read more
First Human Implant Performed With New Open Mitral Annuloplasty Ring
In mitral valve repair, clinicians aim to preserve the mitral annulus’s three-dimensional motion while providing enough support for effective remodeling. Ring preferences vary between open and closed ... Read more
Handheld Device Improves Targeted Cell Collection for Early Ovarian Cancer Assessment
High-grade serous ovarian cancer is the most common form of ovarian cancer, and many cases are thought to originate in the fallopian tubes. Because precursor lesions can be microscopic, they are difficult... Read morePatient Care
view channel
EHR Model Flags High-Risk Periods in Patients with Metastatic Breast Cancer
Determining when patients with metastatic breast cancer are nearing the end of life remains difficult, often leading to intensive interventions with limited benefit. Prognostic uncertainty can delay goals-of-care... Read more
Virtual Reality Helps Claustrophobic Patients Complete Cardiac PET/CT Scans
Anxiety and claustrophobia—intense fear of confined spaces—often prevent patients from completing cardiac positron emission tomography/computed tomography (PET/CT). Aborted scans delay diagnosis, consume... Read moreHealth IT
view channel
Healthcare Database Helps Close Long COVID Surveillance Gap
Long COVID is a chronic condition that occurs after SARS-CoV-2 infection and lasts at least three months. It can follow severe illness, but it can also affect anyone who has been infected.... Read more
New Tablet Platform Extends Stroke Neurorehabilitation Beyond the Clinic
Continuity and intensity of neurorehabilitation after stroke are critical for functional recovery, but access to ongoing therapy at home often remains limited. Hospitals and outpatient programs therefore... Read morePoint of Care
view channel
Continuous Glucose Monitoring Identifies Cardiometabolic Risk in Adults Without Diabetes
Dysglycemia—abnormal blood glucose regulation—can fluctuate throughout the day and often escape conventional screening. Clinicians typically rely on fasting plasma glucose and hemoglobin A1c, which offer... Read more
Microneedle Patch Enables At-Home Monitoring of Acute Kidney Injury
Kidney disease is often silent in its early stages, making timely detection difficult outside clinical settings. Early biomarkers such as neutrophil gelatinase–associated lipocalin can require blood draws... Read moreBusiness
view channel
Medtronic Receives FDA Clearance for Vessel-Sealing Instrument for Robotic Surgery
As robotic-assisted surgery expands across U.S. hospitals, teams increasingly seek energy instruments with the familiarity and performance of tools used in open and laparoscopic procedures.... Read more
Robotic Single-Port System Gains CE Mark for Transvaginal Gynecologic Procedures
Intuitive (Sunnyvale, CA, USA) announced that it has received CE mark approval for use of the da Vinci SP Single Port surgical system in transvaginal gynecologic procedures, marking the first such indication... Read more







