Patrick Campbell

Could artificial intelligence (AI) enhance the ability of clinicians to detect prediabetes and diabetes?

Findings from a recently published study show that 

adjunctive use of DiaBeats, a machine learning-based algorithm, with electrocardiogram (ECG) recordings

 may provide an alternative noninvasive imaging approach to screening for prediabetes and diabetes.

“In theory, our study provides a relatively inexpensive, non-invasive, and accurate alternative (to current diagnostic methods) which can be used as a gatekeeper to effectively detect diabetes and prediabetes early in its course,” wrote study co-author Anoop R. Kulkarni, Ph.D., who is affiliated with Innotomy Consulting in Bengaluru, India, anbd colleagues. “Nevertheless, adoption of this algorithm into routine practice will need robust validation on external, independent data sets.”

As the intrinsic risk of prediabetes has become increasingly recognized by the medical community, identification of low-cost, noninvasive methods for diagnosis or risk stratification for prediabetes and future diabetes has become paramount. With this in mind, a team from the Late Medical Research Foundation in Nagpur, India and M&H Research LLC sought to assess whether leveraging machine learning could identify prediabetes or diabetes from noninvasive cardiovascular imaging as the cardiovascular system bears some of the earliest signs of damage from the diabetic process.

To do so, investigators designed their study as an analysis using data from the Diabetes in Sindhi Families in Nagpur (DISFIN) study. A cross-sectional observational study aimed at estimating prevalence of type 2 diabetes in Sindhi families of Nagpur, 1262 of the 1462 individuals enrolled in the DISFIN study were included in the current analysis. From these 1262 individuals, investigators obtained data related to 10,461 time-aligned heartbeats recorded digitally.

This cohort had a mean age of 48 years and 61 percent were female. The overall prevalence of type 2 diabetes, prediabetes, and insulin resistance was 30 percent, 14 percent, and 35 percent respectively. Investigators also pointed out hypertension, obesity, and dyslipidemia were present among 51 percent, 40 percent, and 36 percent of participants.

For the purpose of analysis, the 10,461 time-aligned heartbeats were split into a training set, a validation set, and an independent test set, which included 8892, 523, and 1,046 beats, respectively. These recordings were processed with median filtering, band-pass filtering, and standard scaling. Investigators pointed out minority oversampling was performed to balance the training dataset prior to initiation of training and extreme gradient boosting was used to train the classifier that used the signal-processed ECG as input and predicted the membership to no diabetes, prediabetes, or type 2 diabetes classes.

Upon analysis, investigators found the DiaBeats algorithm predicted the classes with 97.1 percent precision, 96.2 percent recall, 96.8 percent accuracy, and 96.6 percent F1 score in the independent test set. Investigators noted a low calibration error (0.06) was observed with the calibrated model and the feature importance maps suggested that leads III, aVL, V4, V5, and V6 contributed most to the classification performance. Additionally, the predictions made by the algorithm matched the clinical expectations based on biological mechanisms of cardiac involvement in diabetes, according to investigators.

“Machine-learning-based DiaBeats algorithm using ECG signal data accurately predicted diabetes-related classes. This algorithm can help in early detection of diabetes and prediabetes after robust validation in external datasets,” investigators added.

Articles

Emerging research suggests the machine learning-based DiaBeats algorithm could facilitate early detection of prediabetes or diabetes.

Study Says Combination of AI and Electrocardiogram May Enhance Detection of Diabetes Risk

(Editor’s note: This article has been adapted from its original publication on our sister site 

Preoperative planning with 3D computed tomography (CT) imaging may be beneficial for surgeons looking to improve the success and efficiency of procedures for left atrial appendage occlusion (LAAO).

In a recently published retrospective analysis of nearly 500 Watchman (Boston Scientific) implantations

, researchers found that preoperative use of 3D CT not only facilitated higher procedure success rates in comparison to standalone transesophageal echocardiogram (TEE), it also reduced procedure time and implant size changes.

“The standard method for imaging the heart to guide LAAO procedures is 2-dimensional transesophageal echocardiogram, which uses ultrasound waves to make a detailed picture of the heart,” said the study’s senior investigator Dee Dee Wang, MD, the director of structural heart imaging at Henry Ford Hospital, 

. “This study aimed to assess the value of adding 3-dimensional CT imaging to that process versus using only TEE imaging to make that detailed picture. Our findings indicate significant benefit by adding CT imaging, which uses X-ray to help create a more comprehensive three-dimensional image of the heart.”

For the retrospective review of 485 Watchman implantation procedures, the study authors noted the primary outcome of interest was the rate of successful device implantation without major peri-device leaks, which was defined as a leak greater than 5 mm. Secondary outcomes of interest included major adverse events, total procedural time, radiation dose, total contrast used, number and types of delivery sheath used, number of devices used, number of partial recaptures, and risk of significant peri‐device leak, and device‐related thrombus at follow‐up imaging.

Out of the 485 Watchman implantations, additional CT preoperative planning was used for 328 procedures and surgeons used stand-alone TEE guidance in 157 cases. Compared to those in the stand-alone TEE guidance arm, patients in the CT group had a significantly lower BMI (28.7±6.2 vs 30.9±6.8; 

=.001) and higher HAS‐BLED score (3.2±1.0 vs 2.9±1.0; 

=.006). However, there were no significant differences observed in their baseline clinical profiles, including the baseline creatinine levels.

Upon analysis, results indicated additional preprocedural CT planning was associated with a significantly higher successful device implantation rate (98.5% vs 94.9%; 

=0.02), a shorter procedural time (median, 45.5 vs 51.0 minutes; 

=.03), and a less frequent change of device size (5.6% vs 12.1%; 

=.01) and particularly device upsize (4% vs 9.4%; 

=.02). Investigators pointed out there was no significant difference in the risk of major adverse events (2.1% vs 1.9%; 

=.87). Additionally, only 1 significant peri‐device leak (0.2%) and 5 device‐related thrombi were detected in follow‐up (1.2%), with no intergroup differences observed.

“CT imaging allows us to take all guesswork out of device implantation. We know that we can safely close the appendage and have a success (rate) of 98 percent when imaging is available,” said William O’Neill, MD, director of the Henry Ford Center for Structural Disease.

Emerging research shows the use of preoperative computed tomography (CT) led to improved success rates with device implantation, reduced procedure time and less device size changes in comparison to standalone transesophageal echocardiogram (TEE) for left atrial appendage occlusion (LAAO) procedures.

Study Suggests Benefits of Preoperative CT for Left Atrial Appendage Occlusion Procedures

Cardiac computed tomography (CT) may provide a safer option than coronary angiography for assessing patients with stable chest pain and suspected coronary artery disease (CAD), according to 

In the randomized multicenter DISCHARGE trial, which involved over 3,500 patients with stable chest pain, researchers found that initial cardiac CT and coronary angiography had similar cardiovascular risks, but the 

use of CT had a significantly lower risk of major, procedure-related complications

“The trial confirmed that the CT-based examination is safe in patients with stable chest pain and suspected coronary artery disease,” said Marc Dewey, MD, head of the Department of Radiology on Campus Charité Mitte. “Among the patients referred for cardiac catheterization and included in this trial, the risk of major cardiovascular events was found to be similar in both the CT and catheterization groups, occurring in 2.1% and 3.0% of patients, respectively. The incidence of major, procedure-related complications was found to be lower in patients managed with an initial CT strategy.”

Conducted across 26 European medical centers from October 3, 2015 to April 12, 2019, the DISCHARGE trial enrolled 3,667 patients who were randomized in a 1:1 ratio to undergo CT or invasive coronary angiography. Of the 3,667 enrolled, 3,561 were included in a modified intention-to-treat analysis.

To be considered eligible for the study, patients needed to be at least 30 years of age and referred for invasive coronary angiography at a study site because of stable chest pain with an intermediate pretest probability of obstructive CAD. Investigators noted a web-based system was used to ensure concealment of group assignments after eligibility criteria had been checked.

As part of the study protocol, it was recommended that patients in either arm without obstructive CAD be discharged from the trial center back to their referring physician for further treatment while those with obstructive CAD were treated according to guidelines.

The primary outcome of interest for the study was incidence of major adverse cardiovascular events, which was a composite of cardiovascular death, nonfatal MI, or nonfatal stroke. Secondary outcomes of interest included major procedure-related complications occurring during or within the first 48 hours after CT or invasive coronary angiography.

Of the 3,561 people included in the intention-to-treat analysis, 1,808 represented the CT group and 1,753 patients were in the invasive coronary angiography group. Baseline characteristics of patients were similar between the study groups and the median follow-up time was 3.5 (IQR, 2.9-4.2) years.

During the follow-up period, a total of 90 major adverse cardiovascular events occurred among the study cohort, with events occurring among 2.1% in the CT group and 3.0% in the invasive coronary angiography group (HR, 0.70 [95% CI, 0.46-1.07]; 

=.10). When assessing major procedure-related complications, 42 such events were identified by investigators, with these events occurring in 0.5% of the CT group and 1.9% of the invasive coronary angiography group (HR, 0.26 [95% CI, 0.13-0.55]). Assessments of angina among study participants during the final four weeks of follow-up indicated angina was reported by 8.8% of the CT group and 7.5% of the invasive coronary angiography group (HR, 1.17 [95% CI, 0.92-1.48]).

“Now that it has been standardized and quality tested as part of the DISCHARGE trial, this method could be made more widely available as part of the routine clinical care of people with intermediate CAD risk,” said Henryk Dreger, Deputy Head of Charité’s Department of Cardiology and Angiology.

For patients with stable chest pain and suspected coronary artery disease (CAD), emerging evidence shows that cardiac computed tomography (CT) has a lower risk of major procedure-related complications than catheterization.

Can Cardiac CT be a Viable Alternative for Diagnosing Coronary Artery Disease?

This article was first published on our sister site Endocrinology Network.

Data gathered from bone density scans (DEXA) shows that older women with higher levels of aortic calcification could also be at greater risk for increased falls and fractures. 

Based on a study of more than 1,000 women in Australia, researchers revealed data that suggests older women with evidence of aortic calcification could be at an increased 

risk of falls and potential for fractures

A look at fall-related hospitalizations over a 14.5-year period, results indicate each unit increase in abdominal aortic calcification (AAC) was associated with a 3-percent increase in relative risk for fall-related hospitalization.

"We know that AAC identifies women at a higher risk of heart attacks and strokes, but our research now shows that it also identifies women at a higher falls risk, independent of other falls risk factors and muscle strength," Joshua Lewis, PhD, a National Heart Foundation Future Leader Fellow at Edith Cowan University, 

Funded by the Rebecca L. Cooper Medical Research Foundation, the current study was conducted Lewis and colleagues from Edith Cowan University with an interest in identifying whether an association existed between AAC and increased risk of long-term fall-related hospitalizations in aging women. To do so, investigators designed their study as an analysis of data from within the Perth Longitudinal Study of Aging in Women (PLSAW).

A 15-year longitudinal study that began in 1998-1999, PLSAW examined the impact of various environmental, anatomical, physiological, metabolic, and genetic factors on health outcomes in aging women and provided investigators with information related to a cohort of 1,053 older women with a mean age of 75.0±2.6 years meeting criterion for inclusion in their study. All women included in the study had AAC assessed from lateral spine images obtained from dual-energy X-ray absorptiometry and were scored using the AAC 24 semiquantitative method. Of note, presence of any AAC was defined as an AAC 24 of 1 or more.

Upon analysis, investigators identified 413 (39.2 percent) women who experienced a fall-related hospitalization during the follow-up period. In adjusted models, results suggested each 1 unit increase in baseline AAC24 was associated with a 3-percent increase in relative risk for a fall-related hospitalization (HJR, 1.03; 95 percent CI, 1.01-1.07).

When compared to women with no AAC, results indicated those with any AAC were at a 40-percent greater risk for fall-related hospitalizations in minimally adjusted models. In fully adjusted models comparing the same groups, women with any AAC were at a 39-percent increased risk for fall-related hospitalizations. Additionally, investigators pointed out these relationships were not attenuated by inclusion of measurements of muscle function.

"If we can capture an additional scan to look for evidence of AAC at the same time, we can potentially identify and prevent future harmful falls,” said Abadi Gebre, a co-investigator and Ph.D. candidate at Edith Cowan University. "We often wait until a person suffers a fall to intervene and at that point the damage is already done."

Abdominal aortic calcification is associated with a higher risk of injurious fall-related hospitalizations in older Australian women

For more coverage based on industry expert insights and research, subscribe to the Diagnostic Imaging e-Newsletter 

Assessment of more than 1,000 women in Australia shows abdominal aortic calcification linked to increased risk of fall-related hospitalizations.

DEXA Scans Show Aortic Calcification Predicts Fall Risk in Older Women

This story first appeared on Diagnostic Imaging's sister site 

In patients who are at risk for a heart attack or stroke, coronary artery calcium (CAC) scoring could help providers identify patients who could benefit the most from using aspirin as a primary intervention. 

A new study from a team of investigators from the Preventive Cardiology Program at the University of Texas Southwestern outlines this potential new use for CAC scoring. The team, led by Amit Khera, M.D., professor of internal medicine and director of the Preventive Cardiology Program, used data from the Dallas Heart Study.  

“Aspirin use is not a one-size-fits-all therapy,” Khera said 

. “CAC scanning can be a valuable tool to help us tailor care to help more patients avoid a first heart attack or stroke.”

CT Scans for Arterial Plaque Better Predictor for Heart Attack Than Stroke

With guideline changes altering the use of aspirin in prevention settings, a new emphasis has been placed on predictors and factors associated with increased bleeding when used for primary prevention. In recent years, a multitude of studies have concluded CAC scoring was useful for identifying patients at an increased atherosclerotic cardiovascular disease (ASCVD) risk. With this in mind, Khera and colleagues designed the current study to examine whether CAC scoring could help determine the net estimated effect of aspirin at different risk thresholds.

Using the 6,101-patient Dallas Heart Study as a data source, investigators performed a search of patients free from ASCVD and not taking aspirin at baseline. From this search, investigators identified a cohort of 2,191 patients for inclusion in the current study. This cohort had a mean age of 44 (SD, 9.1) years, 57 percent were women, and 47 percent were Black.

Of note, all patients in the Dallas Heart Study were invited to three sequential visits to collect health information, including various imaging studies and electron beam computed tomography for CAC scoring. Of the 2,191 patients included in the study, 1,063 had a CAC of 0, 967 had a score of 1-99, and 161 had a score of 100 or greater.

The primary outcomes of the study were first bleeding event and first ASCVD event associated with hospitalization or death. Bleeding events were identified through the use of ICD-9 codes and the ASCVD outcome was defined as a composite of nonfatal myocardial infarction, coronary heart disease death, and nonfatal and fatal stroke.

Over a follow-up period lasting a mean of 12.2 (1.9) years, 116 first major bleeding events and 123 ASCVD events were observed. When assessing types of bleeding events, investigators found 81 (70 percent) were gastrointestinal and 18 were intracerebral. Investigators pointed out 24 (21 percent) of the 116 events were fatal bleeds.

Upon analysis, results indicated those with a CAC score of 1-99 were at a greater risk for ASCVD (HR, 4.8; 95 percent CI, 2.8-8.2; 

 <.01) and bleeding events (HR, 1.6; 95 percent C I, 1.1-2.4; 

=.02) when compared to their counterparts with a score of 0. Results indicated those with a CAC score of 100 or greater were also at an increased risk of ASCVD (HR, 5.3; 95 percent CI, 3.6-7.9; 

 <.01) and bleeding events (HR, 2.6; 95 percent CI< 1.5-4.3; 

 <.01) when compared to those with a score of 0. However, investigators pointed out these association between bleeding events and CAC was attenuated in multivariable-adjusted models.

Using meta-analysis estimates for the effects of aspirin, irrespective of CAC, investigators found aspirin use was estimated to result in net harm for individuals at low and intermediate (less than 5 percent) 10-year ASCVD risk. For those considered to be at high-risk (20 percent or greater), aspirin use was associated with a net benefit.

When assessing bleeding risk, results suggested a CAC score of 100 or greater in patients considered to be at low bleeding risk indicated net benefit from aspirin use, but only in those at borderline or higher (5 percent or greater) 10-year ASCVD risk. Among individuals at higher bleeding risk, aspirin use was associated with net harm regardless of CAC and ASCVD risk.

, Miguel Cainzos-Achirica, MD, MPH, PhD, and Philip Greenland, MD, applaud investigators of the current study for adding to a growing body of evidence suggesting CAC scores could help guide aspirin use for primary prevention, but cautions that more conclusive studies are needed before overinterpreting data.

For more coverage based on industry expert insights and research, subscribe to the Diagnostic Imaging e-Newsletter

“Until further studies become available, the overall consistent observations from the DHS and the MESA have important clinical implications…Although it has been proposed that clinical risk scores could aid in the identification of optimal candidates for aspirin therapy, CAC scores seem to be a better tool to inform risk management discussions involving aspirin, leading to a safer allocation in candidates who are appropriate and motivated—particularly in middle-aged populations,” wrote the pair.

Value of Coronary Artery Calcium Scanning in Association With the Net Benefit of Aspirin in Primary Prevention of Atherosclerotic Cardiovascular Disease