United States

Background 

Early melanoma detection has profound clinical impact with distant stages having a
32% 5-year relative survival rate. Traditionally, the automation of melanoma detection prior to biopsy has relied on broad-spectrum photography, thresholding in the frequency domain or analysis of electrical impedance. The advent of deep learning (DL) has enabled automation of dermoscopy.
However, most existing end-to-end DL approaches require annotated data, which is time-consuming to produce and cannot propose novel predictive markers. To address this, we utilized the Semantic Clustering by Adopting Nearest Neighbors (SCAN) algorithm, an unsupervised learning approach that decouples feature learning and clustering, enabling the automated detection of dermatoscopic features that predict disease.

Methods 

We employed the SCAN algorithm to analyze 7991 training and 2024 testing
dermatoscopic images of melanoma or benign nevi from the International Skin Imaging Collaboration dataset. Images are almost all Fitzpatrick type I-III. The model was trained to learn semantically meaningful features that were employed to build clusters of similar images.

Results 

The final clusters exhibited varying levels of homogeneity, with several clusters composed exclusively of one class (melanoma/benign nevi). Several highly predictive clusters for melanoma showed an absence of color uniformity, which is a Menzies method criterion to establish melanoma diagnosis. This indicates latent-space morphology congruent with established evidence-based practice.

Conclusion 

The application of the SCAN algorithm to dermatoscopic images presents a promising methodology for unsupervised learning in melanoma detection. Previous work has shown that AI models that offer explanations to clinicians, such as photos of similar lesions in that continuous area of the model’s latent space, boost physician satisfaction. The FDA approval process also prefers explainable paradigms to ensure model predictability spanning the input space. This study presents an unsupervised approach to melanoma detection via dermoscopy, which enables explainable AI and automated detection of features that predict disease. By providing explainability, the research addresses a key challenge in machine learning paradigms’ FDA approval pathway. Future research is warranted to see if analysis of nevi on dark skin (Fitzpatrick type &gt;3) demonstrates alternate dermoscopic features, as has been previously reported in dermoscopic
literature.

AMA Research Challenge 2023

End-to-End Semantic Clustering in Dermoscopy of Melanocytic Lesions

ICPR 2020 is the premier international conference on Pattern Recognition, gathering leading researchers from around the world to present their latest achievements in Pattern Analysis and Recognition, applied Artificial Intelligence and Machine Learning. The topic of the year is " Putting Artificial intelligence to work on Patterns". 
For the first time, ICPR has employed a two-round review process similar to journal submissions, with new papers allowed to be submitted in either the first or second round and papers submitted in the first round and not accepted allowed to be revised and re-submitted for second round review. 

In this edition of the **ICPR 2020** Conference, there are 1407 Oral/Poster presentations organized in five Tracks: 
- Artificial Intelligence, Machine Learning for Pattern Analysis; 
- Biometrics, Human Analysis and Behavior Understanding; Computer Vision, Robotics and Intelligent Systems; 
- Document and Media Analysis; and 
- Image and Signal Processing. 

The **ICPR 2020** Main Conference program also includes 7 Keynotes from prominent world-renowned scientists, 8 Tutorials, 28 Demos, 8 Challenges, 2 Panels, as well as an Industry/Sponsor Exhibition. In addition, 41 Workshops on a range of important current and emerging research topics complete the ICPR 2020 program. 


ICPR 2020

Background 

Early melanoma detection has profound clinical impact with distant stages having a
32% 5-year relative survival rate. Traditionally, the automation of melanoma detection prior to biopsy has relied on broad-spectrum photography, thresholding in the frequency domain or analysis of electrical impedance. The advent of deep learning (DL) has enabled automation of dermoscopy.
However, most existing end-to-end DL approaches require annotated data, which is time-consuming to produce and cannot propose novel predictive markers. To address this, we utilized the Semantic Clustering by Adopting Nearest Neighbors (SCAN) algorithm, an unsupervised learning approach that decouples feature learning and clustering, enabling the automated detection of dermatoscopic features that predict disease.

Methods 

We employed the SCAN algorithm to analyze 7991 training and 2024 testing
dermatoscopic images of melanoma or benign nevi from the International Skin Imaging Collaboration dataset. Images are almost all Fitzpatrick type I-III. The model was trained to learn semantically meaningful features that were employed to build clusters of similar images.

Results 

The final clusters exhibited varying levels of homogeneity, with several clusters composed exclusively of one class (melanoma/benign nevi). Several highly predictive clusters for melanoma showed an absence of color uniformity, which is a Menzies method criterion to establish melanoma diagnosis. This indicates latent-space morphology congruent with established evidence-based practice.

Conclusion 

The application of the SCAN algorithm to dermatoscopic images presents a promising methodology for unsupervised learning in melanoma detection. Previous work has shown that AI models that offer explanations to clinicians, such as photos of similar lesions in that continuous area of the model’s latent space, boost physician satisfaction. The FDA approval process also prefers explainable paradigms to ensure model predictability spanning the input space. This study presents an unsupervised approach to melanoma detection via dermoscopy, which enables explainable AI and automated detection of features that predict disease. By providing explainability, the research addresses a key challenge in machine learning paradigms’ FDA approval pathway. Future research is warranted to see if analysis of nevi on dark skin (Fitzpatrick type >3) demonstrates alternate dermoscopic features, as has been previously reported in dermoscopic
literature.

poster

Welcome to the **largest national, multi-specialty research** event featuring research from nearly 700 medical students, residents, fellows, and international medical graduates.

Use the buttons below or navigation menu to access the event.
 


[![](https://assets.underline.io/uploads/markdown_image/1/image/3e33a8c754cc49610b30950617d813e7.png)](https://underline.io/events/434/sessions?searchGroup=all)
 
[![](https://assets.underline.io/uploads/markdown_image/1/image/551404d429c4e274f9a54cbdd7081918.png)](https://underline.io/events/434/posters?searchGroup=all)
 
[![](https://assets.underline.io/uploads/markdown_image/1/image/677d92f566b880cebd36905fc738fc10.png)](https://underline.io/events/434/lounge)
 
[![](https://assets.underline.io/uploads/markdown_image/1/image/c7392833c7261c3874c5b43fa2f4df19.png)](https://underline.io/events/434/help-desk)
 

#### Event Details

The 2023 American Medical Association Research Challenge includes a poster symposium hall and a semifinals hall with the top scored research. **Log in at any time throughout the event to:**

* **View posters and presentations** on a variety of topics and specialties
* **Score posters in the semifinals** to help decide the top five finalists (AMA member exclusive opportunity)
* **Connect with participants** and provide feedback to help advance their research
* **Attend education sessions** with physician experts on key research topics

At the conclusion of this event, the AMA will announce the top five finalists who will advance to the AMA Research Challenge finals, present their research to an elite panel of judges, and compete to win a $10,000 grand prize sponsored by Laurel Road. 

### Grand Prize Sponsor

AMA would like to thank Laurel Road for sponsoring a $10,000 grand prize for the winner of the AMA Research Challenge. Laurel Road is not involved in the judging or scoring process to select the winner. The winner will be announced on February 4, 2024 during the AMA Research Challenge finals. Payment of the grand prize will be directly from Laurel Road. See the AMA Research Challenge Official Rules for terms and conditions.

![](https://assets.underline.io/uploads/markdown_image/1/image/611e331d4a0eba0920f240cfc04b4d77.png)

**There is no registration fee for this event! **

Register to explore research in a variety of topics. AMA members get exclusive access to score posters in the semifinals.

**Please use this link [here](https://www.ama-assn.org/about/events/ama-research-challenge) to register for free.**

Registration Is Required

Join the AMA Research Challenge poster symposium and semifinals to explore top research posters, score posters in the semifinals (an AMA member exclusive opportunity) and attend education sessions with experienced physicians to learn tips to advance your research and career. 

Alopecia areata (AA) is an autoimmune condition that leads to patchy, nonscarring hair loss. The etiology still remains unknown, and the condition can be debilitating for patients, impacting their psychosocial wellbeing. Various triggers have been reported ranging from genetic predispositions, infections, and environmental factors. Medications have also been thought to be an inciting factor in AA. Using the FDA adverse event reporting system (FAERS), all cases reporting AA as an adverse event were used to capture associated medications and patient characteristics. Descriptive statistics were conducted along with logistic regression analysis using patient sex as the primary outcome. There were 1,331 AA cases reported as an adverse event with medication use. With regards to the top 10 medications, 29.2% of AA cases were seen with docetaxel, 8.3% were with adalimumab, 4.7% with trastuzumab, 4.4% with cyclophosphamide, 4.3% with infliximab, 4.1% with carboplatin, 3.4% with dupilumab, 2.7% with levonorgestrel, 2.6% with etanercept, and 2.6% with from pertuzumab. Monoclonal antibodies accounted for 6 out of the top 10 drugs with the highest number of AA cases. Males were more likely to report AA when taking adalimumab (OR 1.79, p = 0.04) and dupilumab (OR 2.56, p = 0.03) compared to females. Individuals between 42-64 years old accounted for 46.7% of AA cases. Lastly, females in older age groups showed greater odds of experiencing AA compared to males (OR 1.03, p < 0.01). Based on the FAERS, there has been a steady rise in AA cases and monoclonal antibodies were the most frequently cited medication class tied to AA. Given the dearth of literature and many unanswered questions surrounding AA, our novel large dataset study highlights the most associated medications, identifies an increase in reported cases over the years, and captures age/sex related demographic trends. An awareness of notable medications and patient demographics adds granularity to the spectrum of potential AA triggers and may increase a provider’s index of suspicion to improve clinical outcomes in the future.

A Cross-Sectional Analysis of Medications Used by Patients Reporting Alopecia Areata on the FDA Adverse Event Reporting System

Objective: To determine the effect of video and direct laryngoscopy on first-pass success rates for out-of-hospital orotracheal intubation.
Methods: MEDLINE, Embase, and Cochrane databases were searched from inception to January 2023. Out-of-hospital studies comparing video and direct laryngoscopy on either first-pass or overall intubation success were included. A random effects meta-analysis was performed with a primary outcome of first-pass success stratified by clinician type and laryngoscope blade geometry. The secondary outcomes were overall intubation success stratified by clinician type, and intubation time. All hypotheses and subgroup analyses were determined a priori.
Results: Twenty-five studies involving 35,489 intubations met inclusion criteria. Substantial heterogeneity (>75%) precluded reporting point estimates for nearly all analyses. For our primary outcome, video laryngoscopy was associated with improved first-pass success in 3/5 physician studies, 4/6 critical care paramedic/registered nurse studies, and 7/10 paramedic studies. Video laryngoscope devices with Macintosh blade geometry were associated with improved first-pass success in 7/10 studies, while devices with hyperangulated geometry were associated with improved first-pass success in 3/7 studies. Overall intubation success was greater with video laryngoscopy in 2/6 studies in the physician subgroup and 9/10 studies in the paramedic subgroup. Video laryngoscopy was not associated with overall intubation success among critical care paramedics/nurses (OR = 1.89, 0.96 to 3.72, I2 = 34%). Lastly, 4/5 studies found video laryngoscopy to be associated with longer intubation times.
Conclusions: We found substantial heterogeneity among out-of-hospital studies comparing video laryngoscopy to direct laryngoscopy on first-pass success, overall success, or intubation time. This heterogeneity was not explained with stratification by study design, clinician type, video laryngoscope blade geometry, or leave-one-out meta-analysis. A majority of studies showed that video laryngoscopy was associated with improved first pass success in all subgroups, but only for paramedics and not physicians when looking at overall success. This improvement was more common in studies that used Macintosh blades than those that used hyperangulated blades. Future research should explore the heterogeneity identified in our analysis with an emphasis on differences in training, clinical milieu, and specific video laryngoscope devices.


Out-of-Hospital Video Laryngoscopy: Should We Be Convinced?
A Systematic Review & Meta-Analysis

Background:
As we navigate the post-COVID era, it is essential to confront the health challenges that arose from our changed lifestyles. The World Health Organization (WHO) predicts that by 2025, one out of five adults globally will be obese. Researchers have identified a connection between obesity and mild cognitive impairment. Cognitive impairment is a major public health issue with rising costs for governments and communities. This study aims to evaluate the impact of obesity on cognitive function 

Methods:
A systematic review was undertaken, following the PRISMA 2020 guidelines. Initially, 1250 potentially eligible studies were discovered, out of which 12 duplicates were removed. A thorough assessment of titles and abstracts, 35 full-text articles carefully reviewed, and ultimately, 21 eligible studies incorporated. The search encompassed databases such as PubMed, ScienceDirect, and Google Scholar, focusing on the most recent 20-year period. Articles involving animal models were omitted, and only those freely available in the English Language were examined for inclusion.

Results: 
Based on the literature reviewed, obesity is significantly associated with an increased risk of cognitive impairment in younger and middle-aged individuals but not in older age. Older people with obesity may have a reduced risk of incident dementia. Studies show that obese individuals aged ≥65 years had lower risk ratios for incident dementia compared to normal-weight individuals (0.83 and 0.88, respectively). Another study indicates lower cognitive function in obese school children compared to healthy ones. 

Conclusion:
Obesity is a complex, multifactorial, but preventable disease. Evidence indicates that obesity is associated with cognitive impairment, increasing the risk of dementia later in life. Pathophysiological mechanisms, such as neural integrity reduction and inflammation, contribute to cognitive decline. Addressing obesity through public health interventions may help prevent cognitive impairment. More research is needed to understand the association fully, especially regarding body weight and dementia risk in older age. The lack of a universally accepted cognitive test approach hinders early detection of mild cognitive impairment.


Investigating Obesity-Related Cognitive Decline

Background
Type 2 diabetes mellitus (T2DM) may be a long-term sequela of Mycobacterium tuberculosis (M.tb) infection by mechanisms yet to be fully unraveled. Using the 2011-2012 NHANES, we evaluated (i) the association between latent tuberculosis infection (LTBI) and T2DM among adults (≥20 years old), and (ii) the extent to which this association is mediated by insulin resistance and β-cell failure.

Methods
T2DM was assessed by fasting plasma glucose and HbA1c; homoeostasis model assessment 2 (HOMA2) was used to estimate β-cell function (HOMA2-B) and insulin resistance (HOMA2-IR); while LTBI status was ascertained by interferon-γ release assay. Quintile and probit regression models were used to examine associations between LTBI and, respectively, HOMA2 indices, and prevalent T2DM. Mediation analyses using the counterfactual framework were then performed to quantify the contribution of each of β-cell dysfunction and insulin resistance to the association of LTBI with T2DM. Analyses were adjusted for sociodemographic, behavioral and clinical confounders.

Results
The analytic sample included 2,235 adults. Compared to TB uninfected individuals, those with LTBI were older [median (IQR): 60 (46, 69) vs. 44 (30, 61) years old], and more frequently male [58.5% vs. 48.9%]. LTBI was associated with increased insulin resistance [adjusted mean difference (AMD) (95% CI) in median: 0.08 (0.01, 0.22); p=0.048] but not β-cell function [AMD in median: -2.6 (-9.4, 4.3); p=0.46]. Similarly, T2DM was more prevalent among those with LTBI [adjusted prevalence (95% CI): 18.3 (13.1, 23.5)%] than peers without TB [11.2 (9.7, 12.9)%], corresponding to an adjusted absolute risk difference of 7.4 (2.3, 13.4)%. In mediation analyses, insulin resistance and β-cell function mediated 22.7 (12.2, 68.9)% and 2.7 (-1.4, 9.7)%, respectively, of the total effect of the association between LTBI and T2DM.

Conclusion
Definitive prospective studies examining incident T2DM following tuberculosis are warranted. Notwithstanding, our findings suggest infection with M.tb may be a novel risk factor for T2DM, likely driven by an increase in insulin resistance.

Insulin resistance, and not β-cell impairment, mediates association between latent tuberculosis infection and type II diabetes mellitus among US adults.

Background
There is a higher burden of diabetes related lower extremity amputation in the Lower Rio Grande Valley area of Texas compared to the Texas average. Hispanics have a higher proportion of this complication in the region. Yet, there is a dearth of research on factors associated with the risk of diabetes associated amputation in Hispanic populations. This study evaluates factors associated with diabetic related amputations in Hispanic patients with diabetic foot ulcer in the Rio Grande Valley area of Texas.

Methods
Our study was a retrospective study using patients’ medical records. Subjects were categorized based on whether they had at least one amputation (outcome) during the period of analyses. Baseline demographic, lifestyle, and metabolic characteristics of participants were determined. Characteristics of interest were insulin use, neuropathy, nephropathy, peripheral arterial disease, hypertension, history of cerebrovascular accident, smoking status, osteomyelitis, previous history of diabetic foot ulcer, previous history of amputation, hyperlipidemia, and glycated hemoglobin levels. Chi-square tests were used to analyze the association between these characteristics and amputation status. Logistic regression modeling was used to estimate the odds ratios for amputation and their 95% confidence intervals by each of the independent variables. We adjusted for age and gender in our multivariable analyses. Statistical analyses were carried out using Stata software.

Results
This study included 120 subjects, all of whom had diabetic foot ulcer. Sixty (50%) of the study subjects had diabetic foot amputation and the rest did not. The variables found to be significantly associated with amputation were peripheral arterial disease (p = 0.006), osteomyelitis (p < 0.0001), previous history of amputation (p < 0.0001), and hyperlipidemia (p = 0.013). However, previous diabetic ulcer, neuropathy, nephropathy, hypertension, smoking status, the use of insulin, retinopathy, and levels of hemoglobin A1c were not significantly associated with amputation.

Conclusion
Diabetic foot amputations have untold effects on the lives of affected individuals, and the broader society. Thus, factors promoting this event should be readily prevented and managed. In the studied population, peripheral arterial disease, osteomyelitis, previous amputation, and hyperlipidemia were significantly linked to diabetic foot amputations. Hence, while more research is required, medical professionals should actively seek out and adequately manage diabetic Hispanic patients with these risk factors to prevent this complication.

Factors associated with diabetic foot amputations in Hispanic patients in Rio Grande Valley, Texas.

Background
The bi-directional relationship between depression and prescription opioids is well-established. Evidence shows that adults with depression and non-cancer pain are more likely to remain on opioids and are at risk for opioid use disorder. In contrast, less is known about persistent opioid use and change in anxiety. Among chronic pain patients, opioid users are at higher risk for developing anxiety symptoms compared to those receiving other forms of pain management. However, it is unknown if continued use leads to worsening anxiety when compared to quitting. This study aims to determine if persistent opioid use is a risk factor for developing or worsening generalized anxiety disorder (GAD).

Methods
This prospective cohort study involved non-cancer pain patients (n=361) 18-70 years old starting a period of 30 to 90-day opioid use. Patients from Saint Louis University’s medical practice and the Henry Ford Health System self-administered web surveys or participated in telephone interviews. Difference in difference (D.I.D) analysis determined if change in opioid use was associated with change in GAD-7 scores. Outcome measures included mean GAD-7 score, change in GAD-7 score between baseline and 6 months, and threshold change from a GAD-7 score<10 before and after adjusting for age, race, gender, depression, PTSD, anhedonia, substance use disorder, smoking, and alcohol use.

Results
At baseline, those who persistently used opioids had significantly higher GAD scores [6.4 (±5.9)] compared to those who quit [4.7 (±5.3)]. However, there was no significant difference in GAD score at 6 months between persistent users [5.3 (±5.5)] and quitters [5.2 (±5.8)]. The crude D-I-D analysis revealed no differences in GAD score change from baseline to 6 months between the groups (D.I.D = -1.58; 95% CI: -3.38, -0.23). An adjusted analysis revealed similarly insignificant results (D.I.D = -1.16; 95% CI: -2.94,0.62). Furthermore, among patients with negative GAD (score<10) at baseline, there was no significant difference in risk of GAD (score>=10) at 6 months, whether they persisted or not [crude RR = 1.10; 95% CI: 0.55,2.20); adjusted RR = 1.01 95% CI: 0.53,1.93)].

Conclusion
Patients who persisted vs. quit opioid use at 6 months did not differ in association with worsening or new-onset GAD. However, the results reveal a trend for persistent opioid users to have lower 6-month GAD scores. Additional analysis with a larger sample is needed to determine the characteristics of patients who persist yet experience a decrease in GAD symptoms compared to those who experience constant or worsening GAD.

Persistent Opioid Use and the Development of Generalized Anxiety Disorder

Background
Crohn’s disease (CD) and ulcerative colitis (UC) are inflammatory disorders of the GI tract that may cause symptoms leading to the alteration of dietary intake to control symptoms. There are recommended diets for treatment of inflammatory bowel disease (IBD), but many diets have low-quality evidence for efficacy. However, in an effort to relieve symptoms, patients or families may self-restrict certain foods, leading to the development of restrictive eating behaviors.
In 2013, avoidant-restrictive food intake disorder (ARFID) was added to the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) to identify patients who actively restrict food intake leading to insufficient caloric/nutritional intake and causing at least one of the following complications: weight loss, nutritional deficiency, dependence on nutritional supplements, or psychosocial impairment. 
As ARFID is a relatively new diagnosis, the prevalence of this diagnosis in chronic disease states has not been well-characterized. A recent adult study measured the prevalence of ARFID in an adult IBD population, examined nutritional risk in the context of ARFID risk, and identified risk factors for ARFID in IBD patients. 28/161 (17%) patients had a positive ARFID risk score, as measured by the Nine-item ARFID Screening tool (NIAS).
Currently, there are no studies evaluating the prevalence of ARFID in pediatric patients with IBD. We estimate that the prevalence is high based on anecdotal food avoidances, particularly during a flare of symptoms. We also hypothesize that patients who have high ARFID risk scores may be more at risk for nutrition deficiencies. 
Methods
This is a prospective, observational cohort study of pediatric patients (age 0-17) with inflammatory bowel disease. Fifty patients will be enrolled. Participants complete surveys about eating behaviors and nutritional status after scheduled clinic visits. Study outcomes measured are ARFID risk score (calculated by the NIAS) and nutritional status (height and weight velocity z-scores, mid-upper arm circumference z-scores). 
Results
Based on recent guidelines, a positive ARFID risk score is defined as a score that meets at least one of the following criteria: ≥10 in the picky eating sub-score, ≥9 in the appetite sub-score, ≥10 in the fear sub-score, ≥24 in the total NIAS score Preliminary results suggest a prevalence of ARFID in participants based off a positive ARFID screen to be 21% (8/38 patients), compared to the estimated prevalence of ARFID in community pediatric populations of 2%.
Conclusion
We anticipate that the results of this study can help clinicians address nutritional deficiencies in pediatric IBD patients before it becomes detrimental to patients’ well-being. With the knowledge that pediatric IBD patients are more at risk of ARFID, their needs can be predicted by clinicians and dietitians to ensure proper care of patients.

Evaluating the Prevalence of Avoidant Restrictive Food Intake Disorder in Pediatric Patients with Inflammatory Bowel Disease

Background
Routine hepatic ultrasonography is the most used clinical imaging study in diagnosing chronic liver diseases (CLDs), including assessing hepatic steatosis, a common presentation of CLDs. However, data remain lacking on the performance of hepatic ultrasonography in assessing steatosis and factors that may impact the quality of its performance. This study evaluated the association of hepatic ultrasonography for the assessment of hepatic steatosis with histological hepatic steatosis and other related factors.
Methods
A single-center retrospective study of 188 patients with various CLDs, regular follow-ups, and liver biopsy at the UCIMC Liver Clinic.
Results
The mean age±SD was 57±13 years old. 93 patients were male (49.5%), and 44 (23.4%) had a BMI ≥ 30. Chronic hepatitis C (CHC, 20.8%) and B (17.5%), and non-alcoholic fatty liver disease (NAFLD, 29.0%) were the main CLDs. In this cohort, 89/187 (47.6%) patients had biopsy-proven hepatic steatosis (i.e., ≥ 5% of hepatocytes), 129/188 (68.6%) had ultrasonographic reports of hepatic steatosis, and 67/184 (36.4%) had ultrasonographic reports of hepatomegaly (i.e., ≥ 16.5 cm). The sensitivity and specificity of hepatic ultrasonography for assessing hepatic steatosis, compared to histological steatosis ≥ 5% (gold standard), was 85.4% (76/89 patients) and 46.9% (46/98 patients), respectively. Chi-square analysis showed that hepatic steatosis on ultrasonography was significantly associated with biopsy-proven hepatic steatosis (76/128, 59.4%, p<.01). Hepatic steatosis on ultrasonography was also significantly associated with age ≥ 65 years old (48/129, 37.2%, p=.02), histological evidence of ballooning degeneration (57/86, 66.3%, p<.01), Mallory-Denk bodies (20/60, 33.3%, p<.01), and portal/periportal fibrosis (66/98, 67.3%, p<.03). The frequency of hepatic steatosis diagnosed on ultrasonography was higher in those with NAFLD (45/92, 48.9%) than those with CHC (25/92, 27.2%) or CHB (22/92, 23.9%), but the difference was not significant (p=.08). Ultrasonographic-diagnosed hepatic steatosis was not significantly associated with BMI ≥ 30 (31/129, 24%, p=.76), elevated AST or ALT to ≥ 35 IU/L for males or ≥ 25 IU/L for females (90/111, 81.1%, p=.06 or 84/111, 75.7%, p=.77, respectively). In multivariate analysis, ultrasonographic-diagnosed hepatic steatosis was significantly associated with the presence of ballooning degeneration (95% CI: 1.25-20.72; p=.02) and portal/periportal fibrosis (95% CI: 1.29-11.14; p=.02) on liver biopsy, but not biopsy-proven hepatic steatosis (95% CI: 0.17-1.94; p=.37) or hepatomegaly on ultrasonography (95% CI: 0.71-7.33; p=.17).
Conclusion
Routine hepatic ultrasonography has good sensitivity, but low specificity in the assessment of hepatic steatosis. Histological hepatocytic ballooning degeneration and portal/periportal fibrosis were independently associated with hepatic steatosis detected by ultrasonography.

Association of Ultrasonographic Assessment of Hepatic Steatosis with the Presence Of Histological Steatosis and the Related Factors

There is very sparse data about prevalence
and outcome of gastric intestinal
metaplasia (GIM) in children. Our aim is to
review clinical and endoscopic findings in
children with GIM diagnosed by gastric
biopsy (bx) undergoing
esophagogastroduodenoscopy (EGD) for
various indications at our institution; also
review and compare above findings in
patients with repeat EGD in patients with
GIM.

Gastric Intestinal Metaplasia (GIM) in Children-How Common It Is And What Happens to It-A Pilot Project

Background:
Gastric cancer is among the top five cancer-burden in the United States and worldwide. The aim of our study is to investigate the racial variations in gastric carcinoma among individuals aged 50 years and above.

Methods:
We conducted a retrospective study using the SEER database, 17 registries, Nov 2021 Sub (2000- 2019) to identify individuals diagnosed with gastric cancer of 50 years and above age group. Age adjustment was done using the 2000 US population as a standard population. We collected racial/ethnic information related to incidence and incidence-based mortality of stomach cancer in our desired age groups.

Results:
This study included a total population of 109,352 patients of 50 years and above age with stomach cancer, out of which 71.6% were white, 12.9% were black, 14.2% were Asian or Pacific Islanders (API), and 0.9% were 1 American 1 Indian/ 1 Alaskan native (AI/AN). Based on incidence, all-cause mortality was 79.2% in white population, 78.9% in black, 71.5% in API and 82.8% in AI/AN. The incidence rate among white, black, API, and AI/AN were 20.6, 35, 36.3, and 21.3 per 100,000 population respectively, whereas the incidence-based mortality among the above races were 16.4, 28.9, 26.7, and 18.7 per 100,000 standard population. According to race, the incidence in white was 71.6% with incidence-based mortality of 79.2%. In Black incidence was 12.9% with incidence-based mortality of 78.9%. In American Indian incidence was 0.9% and incidence-based mortality was 82.8% and for Asian
incidence was 14.2% and incidence-based mortality of 71.5%.

Conclusion:
Our study outlines the racial disparities in the incidence and mortality of gastric carcinoma among individuals with gastric carcinoma among individuals aged 50 years and above. Although, gastric carcinoma is diagnosed mostly in the white population followed by Asian or Pacific islanders, black and then American Indians. The incidence rate and incidence-based mortality rates were highest among
Asian/Pacific Islanders and Black compared to that of white and American Indian/Alaskan natives. These findings help to address the disparities through interventions and healthcare policies targeted to improve gastric cancer incidence and outcomes across a diverse racial or ethnic population in the United
States. Further research is warranted to explore the underlying factors contributing to the racial disparities in gastric cancer.

Premium content

Next from AMA Research Challenge 2023

A Cross-Sectional Analysis of Medications Used by Patients Reporting Alopecia Areata on the FDA Adverse Event Reporting System

Similar lecture

A Comparison of Neural Network Approaches for Melanoma Classification