Strong versus Weak Data Labeling for Artificial Intelligence Algorithms in the Measurement of Geographic Atrophy.

Purpose: To gain an understanding of data labeling requirements to train deep learning models for measurement of geographic atrophy (GA) with fundus autofluorescence (FAF) images. Design: Evaluation of artificial intelligence (AI) algorithms. Subjects: The Age-Related Eye Disease Study 2 (AREDS2) images were used for training and cross-validation, and GA clinical trial images were used for testing. Methods: Training data consisted of 2 sets of FAF images; 1 with area measurements only and no indication of GA location (Weakly labeled) and the second with GA segmentation masks (Strongly labeled). Main Outcome Measures: Bland-Altman plots and scatter plots were used to compare GA area measurement between ground truth and AI measurements. The Dice coefficient was used to compare accuracy of segmentation of the Strong model. Results: In the cross-validation AREDS2 data set (n = 601), the mean (standard deviation [SD]) area of GA measured by human grader, Weakly labeled AI model, and Strongly labeled AI model was 6.65 (6.3) mm2, 6.83 (6.29) mm2, and 6.58 (6.24) mm2, respectively. The mean difference between ground truth and AI was 0.18 mm2 (95% confidence interval, [CI], -7.57 to 7.92) for the Weakly labeled model and -0.07 mm2 (95% CI, -1.61 to 1.47) for the Strongly labeled model. With GlaxoSmithKline testing data (n = 156), the mean (SD) GA area was 9.79 (5.6) mm2, 8.82 (4.61) mm2, and 9.55 (5.66) mm2 for human grader, Strongly labeled AI model, and Weakly labeled AI model, respectively. The mean difference between ground truth and AI for the 2 models was -0.97 mm2 (95% CI, -4.36 to 2.41) and -0.24 mm2 (95% CI, -4.98 to 4.49), respectively. The Dice coefficient was 0.99 for intergrader agreement, 0.89 for the cross-validation data, and 0.92 for the testing data. Conclusions: Deep learning models can achieve reasonable accuracy even with Weakly labeled data. Training methods that integrate large volumes of Weakly labeled images with small number of Strongly labeled images offer a promising solution to overcome the burden of cost and time for data labeling. Financial Disclosures: Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

Feline herpesvirus-1-related multiple respiratory eosinophilic nodules in an adult cat receiving long-term oral prednisolone.

Case summary: A 10-year-old male castrated domestic shorthair cat was presented for evaluation of a 3-day history of increased inspiratory effort. The cat had received prednisolone 1 mg/kg PO q24h for 1 year due to chronic diarrhea. On physical examination, the patient exhibited severe stridor, intermittent open-mouth breathing and bilateral mucopurulent nasal discharge. Subcutaneous emphysema was palpated over the dorsal cervical region. Mild hypoventilation (PvCO2 55.1 mmHg; approximate reference interval 35-45 mmHg) was identified. Cervicothoracic radiographs showed marked gas tracking within cervical soft tissues with concurrent laryngeal thickening, pulmonary nodules, a bronchial pulmonary pattern, pneumomediastinum and aerophagia. The cat was hospitalized and treated overnight with oxygen and intravenous fluid therapy before anesthesia the next day. On laryngoscopy, a large tracheal mass was observed arising from the right subglottic region and was removed using biopsy forceps. CT revealed an additional mass at the level of the tracheal bifurcation causing marked luminal narrowing of the trachea and proximal main bronchi. The cat made a good initial recovery, although moderate stridor persisted. Five days later, the cat was re-examined due to recurrence of respiratory distress and orthopnea, and the owner elected euthanasia. Histopathology revealed severe nodular obstructive eosinophilic plasmacytic laryngotracheitis with intranuclear inclusion bodies positive for feline herpesvirus-1 on immunohistochemistry. Relevance and novel information: This report describes the presentation and management of a cat with respiratory distress secondary to intratracheal eosinophilic masses caused by feline herpesvirus-1. Although the outcome was ultimately unsatisfactory, to the authors' knowledge, this clinical presentation has not been previously reported.

Integrating physiology into the first two years of a new osteopathic medical school curriculum.

The integrated curriculum is a hot topic in curriculum reform in undergraduate medical education; however, there are varying definitions of the term, and resources guiding the integration of specific disciplines throughout the first 2 years of undergraduate medical education in a learner-centered curriculum are limited. Our first class matriculated into our osteopathic medical school in 2017, and since then we have developed and implemented a learner-centered, integrated curriculum that begins on day one for our learners. This paper will discuss our experience with the development and implementation of the University of the Incarnate Word School of Osteopathic Medicine (UIWSOM) curriculum with specific emphasis on how we incorporate physiology into it.

Identifying cancer-associated leukocyte profiles using high-resolution flow cytometry screening and machine learning.

Background: Machine learning (ML) is a valuable tool with the potential to aid clinical decision making. Adoption of ML to this end requires data that reliably correlates with the clinical outcome of interest; the advantage of ML is that it can model these correlations from complex multiparameter data sets that can be difficult to interpret conventionally. While currently available clinical data can be used in ML for this purpose, there exists the potential to discover new "biomarkers" that will enhance the effectiveness of ML in clinical decision making. Since the interaction of the immune system and cancer is a hallmark of tumor establishment and progression, one potential area for cancer biomarker discovery is through the investigation of cancer-related immune cell signatures. Hence, we hypothesize that blood immune cell signatures can act as a biomarker for cancer progression. Methods: To probe this, we have developed and tested a multiparameter cell-surface marker screening pipeline, using flow cytometry to obtain high-resolution systemic leukocyte population profiles that correlate with detection and characterization of several cancers in murine syngeneic tumor models. Results: We discovered a signature of several blood leukocyte subsets, the most notable of which were monocyte subsets, that could be used to train CATboost ML models to predict the presence and type of cancer present in the animals. Conclusions: Our findings highlight the potential utility of a screening approach to identify robust leukocyte biomarkers for cancer detection and characterization. This pipeline can easily be adapted to screen for cancer specific leukocyte markers from the blood of cancer patient.

Clinical Outcomes of Angiotensin Converting Enzyme Inhibitors and Angiotensin II Receptor Blockers in COVID-19 Patients With Pre-existing Cardiac Comorbidities: A Literature Review.

The growing research regarding the implementation of angiotensin-converting enzyme-2 inhibitors (ACEi) and angiotensin receptor blockers (ARBs) in the treatment of COVID-19 in patients with pre-existing cardiac comorbidities has become a large topic of discussion since the onset of the pandemic. Previous research primarily associates positive outcomes to the use of these drug classes due to their mechanism of action, which involves the downregulation of angiotensin I-converting enzyme 2 (ACE2) in the renin-angiotensin-aldosterone-system (RAAS) pathway, inflammatory mediators, and cytokines. Thus, these medications can convey preventative and protective effects in patients suffering from a SARS-CoV-2 infection. While we explored the studies that supported the positive outcomes of the use of these drugs in the first half of this review, we also expanded on the limitations of these studies in the latter portion. We also further explored the contradictory studies that indicated that using these antihypertensives can paradoxically increase the severity of COVID-19 infection as well. The studies in support of the use of these medications should consider epigenetic variations, ACE2 variants and acknowledge inherent genetic variations in certain ethnic groups as some have a predisposition for a severe COVID-19 infection. Additionally, mortality rates need to be taken into consideration in these studies as they naturally differ throughout the trajectory of the COVID-19 pandemic. While some studies are in support of the use of these antihypertensives despite other studies suggesting otherwise, further research is needed to explore the long-term effects of these antihypertensives and observe whether they are truly beneficial or not in reducing the severity of COVID-19 infections.

Implementation of a Large-Scale Image Curation Workflow Using Deep Learning Framework.

Purpose: The curation of images using human resources is time intensive but an essential step for developing artificial intelligence (AI) algorithms. Our goal was to develop and implement an AI algorithm for image curation in a high-volume setting. We also explored AI tools that will assist in deploying a tiered approach, in which the AI model labels images and flags potential mislabels for human review. Design: Implementation of an AI algorithm. Participants: Seven-field stereoscopic images from multiple clinical trials. Methods: The 7-field stereoscopic image protocol includes 7 pairs of images from various parts of the central retina along with images of the anterior part of the eye. All images were labeled for field number by reading center graders. The model output included classification of the retinal images into 8 field numbers. Probability scores (0-1) were generated to identify misclassified images, with 1 indicating a high probability of a correct label. Main Outcome Measures: Agreement of AI prediction with grader classification of field number and the use of probability scores to identify mislabeled images. Results: The AI model was trained and validated on 17 529 images and tested on 3004 images. The pooled agreement of field numbers between grader classification and the AI model was 88.3% (kappa, 0.87). The pooled mean probability score was 0.97 (standard deviation [SD], 0.08) for images for which the graders agreed with the AI-generated labels and 0.77 (SD, 0.19) for images for which the graders disagreed with the AI-generated labels (P < 0.0001). Using receiver operating characteristic curves, a probability score of 0.99 was identified as a cutoff for distinguishing mislabeled images. A tiered workflow using a probability score of < 0.99 as a cutoff would include 27.6% of the 3004 images for human review and reduce the error rate from 11.7% to 1.5%. Conclusions: The implementation of AI algorithms requires measures in addition to model validation. Tools to flag potential errors in the labels generated by AI models will reduce inaccuracies, increase trust in the system, and provide data for continuous model development.

The Opioid Epidemic in America: Pandemic Impacts.

There is still an ongoing epidemic of opioid abuse in the United States. It has resulted in a major healthcare crisis and led to an alarming number of daily overdose deaths. Strong interventions from the medical community and the government had made progress toward control. Tragically, the COVID-19 pandemic made the death rate dramatically worse. The purpose of this article is to reexamine the origins of this situation and provide updates from the pandemic period. The authors review the medical-legal risks physicians face when prescribing pain-relieving medications for their orthopaedic patients. Methods for providing responsible pain management plans for patients are suggested.

What Is Your Diagnosis?

In collaboration with the American College of Veterinary Radiology.

Radiographic and ultrasonographic appearance of pneumonia in a frog.

A 12-year-old female African Clawed Frog (Xenopus laevis) displayed lethargy and anorexia for 1 week. Radiographs detected a moderately reduced volume of aerated lung with lobulated margins with the left lung predominantly affected. Transcelomic ultrasound revealed marked consolidation of the left lung. Postmortem examination confirmed suppurative and histiocytic pneumonia, worse in the left lung, caused by Mycobacterium chelonae. Given that amphibian pulmonary consolidation and celomic effusion can have similar radiographic findings, ultrasound may be helpful to differentiate.

The Opioid Epidemic: Risk Evaluation and Management Strategies for Prescribing Opioids.

Abuse of opioids has had and continues to have a devastating impact on public health and safety in the United States, and the use of opioids has increased dramatically in the last two decades. The purpose of this chapter is to examine the roots of this tragic state of affairs and what may be done about it moving forward. The authors review the medical-legal risks physicians face when prescribing pain relieving medications for their patients. Strategies are offered for staying out of trouble while providing quality pain management for patients.

Diagnostic imaging characteristics of canine infectious sacroiliitis.

Infectious sacroiliitis has not been described in dogs. This retrospective case series describes the presentation, diagnostic imaging characteristics, and outcomes of 2 canine patients with infectious sacroiliitis. Selection criteria included presentation with back pain from 2010 to 2017, diagnostic imaging of the sacroiliac joints, and short- and long-term response to antibiotic therapy. Medical records, and magnetic resonance imaging (MRI) were reviewed by a Board-certified veterinary radiologist, a neurologist, and a small animal intern. Two dogs met the inclusion criteria. Magnetic resonance imaging revealed unilaterally wide and irregular sacroiliac joint spaces, with juxta-articular soft tissue contrast enhancement and bone marrow edema. One patient had a communicating abscess of the psoas muscle, which cultured positive for Pasteurella canis. Following treatment with pain relief medications and antibiotics, both patients made a complete clinical recovery, with no signs of lameness 2 to 4 weeks after cessation of treatment, and no lameness reported by the owner afterwards. Infectious sacroiliitis should be considered when dogs are presented with lumbosacral pain.

Caractéristiques de l'imagerie diagnostique de la sacro-iliite infectieuse canine. La sacro-iliite infectieuse n'a pas été décrite chez les chiens. Cette série de cas rétrospectifs décrit la présentation, les caractéristiques de l'imagerie diagnostique et les résultats de deux patients canins atteints de sacro-iliite infectieuse. Les critères de sélection incluaient la présentation avec des douleurs au dos de 2010 à 2017, l'imagerie diagnostique des articulations sacro-iliaques et une réponse à court et à long terme à la thérapie aux antibiotiques. Les dossiers médicaux et l'imagerie à résonance magnétique (IRM) ont été évalués par un radiologiste vétérinaire agréé, un neurologue et un interne en médecine des petits animaux. Deux chiens satisfaisaient aux critères d'inclusion. L'imagerie à résonance magnétique a révélé des espaces d'articulation sacro-iliaque unilatéralement larges et irréguliers avec une augmentation de contraste des tissus mous juxta-articulaires et de l'oedème de la moelle osseuse. Le patient 2 avait un abcès communiquant du muscle psoas qui a révélé la présence de Pasteurella canis à la culture. Après un traitement avec des médicaments analgésiques et des antibiotiques, les deux patients ont connu un rétablissement clinique complet sans signes de boiterie de 2 à 4 semaines après l'arrêt du traitement et aucune boiterie n'a été signalée par la suite par le propriétaire. La sacro-iliite infectieuse devrait être considérée lorsque les chiens sont présentés avec des douleurs lombo-sacrées.(Traduit par Isabelle Vallières).

Preserved limb function following subtotal iliopsoas muscle and femoral nerve resection in a dog with low grade intramuscular chondrosarcoma.

A 9-year-old spayed female Labrador retriever dog was diagnosed with a grade 1 chondrosarcoma associated with the right iliopsoas muscle. A computed tomography scan of the patient revealed the mass to begin at the level of L5, continuing to the insertion of the iliopsoas muscle on the femur. Surgery was performed to remove the mass. The femoral nerve was encased within the mass and the nerve was transected and removed, along with the iliopsoas muscle, with minimal intraoperative complications. With regular physiotherapy, the patient was able to independently walk, run, and jump, with mild functional lameness after 6 months.

Fonction préservée des membres après une résection partielle du muscle ilio-psoas et du nerf fémoral chez un chien atteint d'un chondrosarcome intramusculaire de faible grade. Une chienne Labrador retriever stérilisée âgée de 9 ans a été diagnostiquée avec un chondrosarcome de grade 1 associé au muscle ilio-psoas droit. Une tomodensitométrie de la patiente a révélé une masse commençant au niveau de L5 et se poursuivant vers l'insertion de l'ilio-psoas sur le fémur. Une chirurgie a été réalisée pour enlever la masse. Le nerf fémoral était enveloppé par la masse et le nerf a été coupé et enlevé, ainsi que le muscle ilio-psoas, avec des complications peropératoires minimes. Grâce à de la physiothérapie régulière, après 6 mois le patient était capable de marcher, courir et sauter de façon indépendante, avec une légère boiterie fonctionnelle.(Traduit par Isabelle Vallières).

Complementary and Integrative Medicine in Lung Cancer: Questions and Challenges.

Lung cancer represents 13% of all cancers, making it the second most common type of malignancy in the United States. Lung cancer is the leading cause of cancer death in men and women in the United States and accounts for nearly 18% of all deaths from cancer. Because of its high mortality rate, lung cancer is associated with an increased rate of distress. Patients use various strategies to cope with this distress during and after cancer treatments, and complementary and integrative medicine (CIM) has become a common coping strategy. This review covers major questions and challenges of incorporating CIM during and beyond treatment for lung cancer. The questions revolve around determining the value of nutrition and nutritional supplements, assessing the role of exercise, addressing the mind-body connection, enhancing the benefit of immunotherapy, and determining the benefit of incorporating complementary therapies such as acupuncture and homeopathy. This review may provide a basis for discussion that can enhance patient-doctor dialogue regarding the use of CIM during and after treatment for lung cancer.

Influence of Age on Postconcussive Postural Control Measures and Future Implications for Assessment.

OBJECTIVE: To examine the influence of age, sex, attention-deficit hyperactivity disorder (ADHD) status, previous history of concussion, and days since injury on postconcussion postural control assessment in adolescents who have suffered a concussion. DESIGN: Prospective cohort study. SETTING: Hospital-based outpatient clinic. PARTICIPANTS: Seventy-one participants (42 males; 29 females) with mean age 14.14 ± 2.44. INDEPENDENT VARIABLES: Age, sex, previous concussion history, ADHD status, total and severity of postconcussion symptoms, and days since injury. MAIN OUTCOME MEASURES: Total Balance Error Scoring System score, path length, center-of-pressure (COP) area, sample entropy, and Romberg quotient. RESULTS: Pearson product-moment correlation coefficients were calculated to test for potential associations between the continuous participant characteristics and the postural control variables. Spearman correlation was used to test the association between symptom severity and the postural control variables. Standard multiple regression was used to model the extent to which participant characteristics accounted for the variance in the postural sway variables. Age was significantly associated with all of the postural sway variables except COP area for the eyes open condition and sample entropy in the anterior-posterior direction for the eyes closed condition. Sex, ADHD status, and previous concussion history did not significantly predict postural control scores. CONCLUSIONS: Age significantly influences scores on common postconcussion postural control assessments. CLINICAL RELEVANCE: This study demonstrates that age is a critical factor that needs to be accounted for to improve the clinical appropriateness and utility of current postconcussion postural control assessments.

Almond "Appetizer" Effect on Glucose Tolerance Test (GTT) Results.

BACKGROUND: The extent to which glucose intolerance can be acutely improved with dietary modification is unclear. The purpose of this study was to test the effect of ingesting a low-calorie almond preload ("appetizer") 30 minutes before oral glucose tolerance testing in glucose-intolerant individuals without diabetes. METHODS: Twenty adults with prediabetes or isolated 1-hour glucose ≥160 mg/dL underwent 2 fasting oral glucose tolerance tests (GTTs)-1 standard GTT and 1 GTT 30 minutes after eating a half ounce (12) of dry-roasted almonds. Fourteen participants met 1 or more prediabetes diagnostic criteria; 6 had only elevated 1-hour glucose ≥160 mg/dL. RESULTS: The mean 1-hour plasma glucose after the almond preload was 37.1 mg/dL (19.4%) lower (154.6 vs 191.7; P < .001) than in the standard GTT. The almond preload reduced the area under the glucose curve by 15.5% (P < .001). Eight individuals had a marked hypoglycemic effect (glucose reduced by 45 to 110 mg/dL); 4 had a moderate hypoglycemic effect (22-32 mg/dL). CONCLUSION: A low-calorie almond "appetizer" showed promise as an option for decreasing postprandial hyperglycemia in individuals with prediabetes or isolated 1-hour postprandial hyperglycemia. Further study is needed to confirm and refine the role of such a premeal appetizer in the self-care of prediabetes.