AI-clinician collaboration via disagreement prediction: A decision pipeline and retrospective analysis of real-world radiologist-AI interactions.

Clinical decision support tools can improve diagnostic performance or reduce variability, but they are also subject to post-deployment underperformance. Although using AI in an assistive setting offsets many concerns with autonomous AI in medicine, systems that present all predictions equivalently fail to protect against key AI safety concerns. We design a decision pipeline that supports the diagnostic model with an ecosystem of models, integrating disagreement prediction, clinical significance categorization, and prediction quality modeling to guide prediction presentation. We characterize disagreement using data from a deployed chest X-ray interpretation aid and compare clinician burden in this proposed pipeline to the diagnostic model in isolation. The average disagreement rate is 6.5%, and the expected burden reduction is 4.8%, even if 5% of disagreements on urgent findings receive a second read. We conclude that, in our production setting, we can adequately balance risk mitigation with clinician burden if disagreement false positives are reduced.

Development of an artificial intelligence-derived histologic signature associated with adjuvant gemcitabine treatment outcomes in pancreatic cancer.

Pancreatic ductal adenocarcinoma (PDAC) has been left behind in the evolution of personalized medicine. Predictive markers of response to therapy are lacking in PDAC despite various histological and transcriptional classification schemes. We report an artificial intelligence (AI) approach to histologic feature examination that extracts a signature predictive of disease-specific survival (DSS) in patients with PDAC receiving adjuvant gemcitabine. We demonstrate that this AI-generated histologic signature is associated with outcomes following adjuvant gemcitabine, while three previously developed transcriptomic classification systems are not (n = 47). We externally validate this signature in an independent cohort of patients treated with adjuvant gemcitabine (n = 46). Finally, we demonstrate that the signature does not stratify survival outcomes in a third cohort of untreated patients (n = 161), suggesting that the signature is specifically predictive of treatment-related outcomes but is not generally prognostic. This imaging analysis pipeline has promise in the development of actionable markers in other clinical settings where few biomarkers currently exist.

Effectiveness of Kanna photoscreener in detecting amblyopia risk factors.

Purpose: Amblyopia is a significant public health problem. Photoscreeners have been shown to have significant potential for screening; however, most are limited by cost and display low accuracy. The purpose of this study was validate a novel artificial intelligence (AI) and machine learning-based facial photoscreener "Kanna," and to determine its effectiveness in detecting amblyopia risk factors. Methods: A prospective study that included 654 patients aged below 18 years was conducted in our outpatient clinic. Using an android smartphone, three images of each the participants' face were captured by trained optometrists in dark and ambient light conditions and uploaded onto Kanna. Deep learning was used to create an amblyopia risk score based on our previous study. The algorithm generates a risk dashboard consisting of six values: five normalized risk scores for ptosis, strabismus, hyperopia, myopia and media opacities; and one binary value denoting if a child is "at-risk" or "not at-risk." The presence of amblyopia risk factors (ARF) as determined on the ophthalmic examination was compared with the Kanna photoscreener. Results: Correlated patient data for 654 participants were analyzed. The mean age of the study population was 7.87 years. The algorithm had an F-score, 85.9%; accuracy, 90.8%; sensitivity, 83.6%; specificity, 94.5%; positive predictive value, 88.4%; and negative predictive value, 91.9% in identifying amblyopia risk factors. The P value for the amblyopia risk calculation was 8.5 × 10-142 implying strong statistical significance. Conclusion: The Kanna photo-based screener that uses deep learning to analyze photographs is an effective alternative for screening children for amblyopia risk factors.

Application of deep learning and image processing analysis of photographs for amblyopia screening.

Purpose: Photo screeners and autorefractors have been used to screen children for amblyopia risk factors (ARF) but are limited by cost and efficacy. We looked for a deep learning and image processing analysis-based system to screen for ARF. Methods: An android smartphone was used to capture images using a specially coded application that modified the camera setting. An algorithm was developed to process images taken in different light conditions in an automated manner to predict the presence of ARF. Deep learning and image processing models were used to segment images of the face. Light settings and distances were tested to obtain the necessary features. Deep learning was thereafter used to formulate normalized risks using sigmoidal models for each ARF creating a risk dashboard. The model was tested on 54 young adults and results statistically analyzed. Results: A combination of low-light and ambient-light images was needed for screening for exclusive ARF. The algorithm had an F-Score of 73.2% with an accuracy of 79.6%, a sensitivity of 88.2%, and a specificity of 75.6% in detecting the ARF. Conclusion: Deep-learning and image-processing analysis of photographs acquired from a smartphone are useful in screening for ARF in children and young adults for a referral to doctors for further diagnosis and treatment.