Assessment of colonoscopy skill using machine learning to measure quality: Proof-of-concept and initial validation.

Background and study aims Low-quality colonoscopy increases cancer risk but measuring quality remains challenging. We developed an automated, interactive assessment of colonoscopy quality (AI-CQ) using machine learning (ML). Methods Based on quality guidelines, metrics selected for AI development included insertion time (IT), withdrawal time (WT), polyp detection rate (PDR), and polyps per colonoscopy (PPC). Two novel metrics were also developed: HQ-WT (time during withdrawal with clear image) and WT-PT (withdrawal time subtracting polypectomy time). The model was pre-trained using a self-supervised vision transformer on unlabeled colonoscopy images and then finetuned for multi-label classification on another mutually exclusive colonoscopy image dataset. A timeline of video predictions and metric calculations were presented to clinicians in addition to the raw video using a web-based application. The model was externally validated using 50 colonoscopies at a second hospital. Results The AI-CQ accuracy to identify cecal intubation was 88%. IT ( P = 0.99) and WT ( P = 0.99) were highly correlated between manual and AI-CQ measurements with a median difference of 1.5 seconds and 4.5 seconds, respectively. AI-CQ PDR did not significantly differ from manual PDR (47.6% versus 45.5%, P = 0.66). Retroflexion was correctly identified in 95.2% and number of right colon evaluations in 100% of colonoscopies. HQ-WT was 45.9% of, and significantly correlated with ( P = 0.85) WT time. Conclusions An interactive AI assessment of colonoscopy skill can automatically assess quality. We propose that this tool can be utilized to rapidly identify and train providers in need of remediation.

Assistive AI in Lung Cancer Screening: A Retrospective Multinational Study in the United States and Japan.

Purpose To evaluate the impact of an artificial intelligence (AI) assistant for lung cancer screening on multinational clinical workflows. Materials and Methods An AI assistant for lung cancer screening was evaluated on two retrospective randomized multireader multicase studies where 627 (141 cancer-positive cases) low-dose chest CT cases were each read twice (with and without AI assistance) by experienced thoracic radiologists (six U.S.-based or six Japan-based radiologists), resulting in a total of 7524 interpretations. Positive cases were defined as those within 2 years before a pathology-confirmed lung cancer diagnosis. Negative cases were defined as those without any subsequent cancer diagnosis for at least 2 years and were enriched for a spectrum of diverse nodules. The studies measured the readers' level of suspicion (on a 0-100 scale), country-specific screening system scoring categories, and management recommendations. Evaluation metrics included the area under the receiver operating characteristic curve (AUC) for level of suspicion and sensitivity and specificity of recall recommendations. Results With AI assistance, the radiologists' AUC increased by 0.023 (0.70 to 0.72; P = .02) for the U.S. study and by 0.023 (0.93 to 0.96; P = .18) for the Japan study. Scoring system specificity for actionable findings increased 5.5% (57% to 63%; P < .001) for the U.S. study and 6.7% (23% to 30%; P < .001) for the Japan study. There was no evidence of a difference in corresponding sensitivity between unassisted and AI-assisted reads for the U.S. (67.3% to 67.5%; P = .88) and Japan (98% to 100%; P > .99) studies. Corresponding stand-alone AI AUC system performance was 0.75 (95% CI: 0.70, 0.81) and 0.88 (95% CI: 0.78, 0.97) for the U.S.- and Japan-based datasets, respectively. Conclusion The concurrent AI interface improved lung cancer screening specificity in both U.S.- and Japan-based reader studies, meriting further study in additional international screening environments. Keywords: Assistive Artificial Intelligence, Lung Cancer Screening, CT Supplemental material is available for this article. Published under a CC BY 4.0 license.

Significance of Intraoperative Medication Data and Predictive Model Selection for Predicting Postoperative First-Time Atrial Fibrillation.

Atrial fibrillation (AF) is the most common sustained cardiac arrhythmia in clinical practice and has a well-established association with coronary artery bypass graft (CABG) surgery. Being able to predict post-operative AF (POAF) may improve surgical outcomes. This study retrospectively assembled a large cohort of 3,807 first-time CABG patients with no prior AF to study factors that contribute to occurrence of POAF, in addition to testing models that may predict its incidence. Several clinical features with established relevance to POAF were extracted from the EHR, along with a record of medications administered intra-operatively. Tests of performance with logistic regression, decision tree, and neural network predictive models showed slight improvements when incorporating medication information. Analysis of the clinical and medications data indicate that there may be effects contributing to POAF incidence captured in the medication administration records. Our results show that improved predictive performance is achievable by incorporating a record of medications administered intra-operatively.

An artificial intelligence platform provides an accurate interpretation of esophageal motility from Functional Lumen Imaging Probe Panometry studies.

BACKGROUND: Functional lumen imaging probe (FLIP) Panometry is performed at the time of sedated endoscopy and evaluates esophageal motility in response to distension. This study aimed to develop and test an automated artificial intelligence (AI) platform that could interpret FLIP Panometry studies. METHODS: The study cohort included 678 consecutive patients and 35 asymptomatic controls that completed FLIP Panometry during endoscopy and high-resolution manometry (HRM). "True" study labels for model training and testing were assigned by experienced esophagologists per a hierarchical classification scheme. The supervised, deep learning, AI model generated FLIP Panometry heatmaps from raw FLIP data and based on convolutional neural networks assigned esophageal motility labels using a two-stage prediction model. Model performance was tested on a 15% held-out test set (n = 103); the remainder of the studies were utilized for model training (n = 610). KEY RESULTS: "True" FLIP labels across the entire cohort included 190 (27%) "normal," 265 (37%) "not normal/not achalasia," and 258 (36%) "achalasia." On the test set, both the Normal/Not normal and the achalasia/not achalasia models achieved an accuracy of 89% (with 89%/88% recall, 90%/89% precision, respectively). Of 28 patients with achalasia (per HRM) in the test set, 0 were predicted as "normal" and 93% as "achalasia" by the AI model. CONCLUSIONS: An AI platform provided accurate interpretation of FLIP Panometry esophageal motility studies from a single center compared with the impression of experienced FLIP Panometry interpreters. This platform may provide useful clinical decision support for esophageal motility diagnosis from FLIP Panometry studies performed at the time of endoscopy.

Deep learning-based artificial intelligence model for identifying swallow types in esophageal high-resolution manometry.

BACKGROUND: This study aimed to build and evaluate a deep learning, artificial intelligence (AI) model to automatically classify swallow types based on raw data from esophageal high-resolution manometry (HRM). METHODS: HRM studies on patients with no history of esophageal surgery were collected including 1,741 studies with 26,115 swallows labeled by swallow type (normal, hypercontractile, weak-fragmented, failed, and premature) by an expert interpreter per the Chicago Classification. The dataset was stratified and split into train/validation/test datasets for model development. Long short-term memory (LSTM), a type of deep-learning AI model, was trained and evaluated. The overall performance and detailed per-swallow type performance were analyzed. The interpretations of the supine swallows in a single study were further used to generate an overall classification of peristalsis. KEY RESULTS: The LSTM model for swallow type yielded accuracies from the train/validation/test datasets of 0.86/0.81/0.83. The model's interpretation for study-level classification of peristalsis yielded accuracy of 0.88 in the test dataset. Among model misclassification, 535/698 (77%) swallows and 25/35 (71%) studies were to adjacent categories, for example, normal to weak or normal to ineffective, respectively. CONCLUSIONS AND INFERENCES: A deep-learning AI model can automatically and accurately identify the Chicago Classification swallow types and peristalsis classification from raw HRM data. While future work to refine this model and incorporate overall manometric diagnoses are needed, this study demonstrates the role that AI will serve in the interpretation and classification of esophageal HRM studies.

Amalgamation of cloud-based colonoscopy videos with patient-level metadata to facilitate large-scale machine learning.

Background and study aims Storage of full-length endoscopic procedures is becoming increasingly popular. To facilitate large-scale machine learning (ML) focused on clinical outcomes, these videos must be merged with the patient-level data in the electronic health record (EHR). Our aim was to present a method of accurately linking patient-level EHR data with cloud stored colonoscopy videos. Methods This study was conducted at a single academic medical center. Most procedure videos are automatically uploaded to the cloud server but are identified only by procedure time and procedure room. We developed and then tested an algorithm to match recorded videos with corresponding exams in the EHR based upon procedure time and room and subsequently extract frames of interest. Results Among 28,611 total colonoscopies performed over the study period, 21,170 colonoscopy videos in 20,420 unique patients (54.2 % male, median age 58) were matched to EHR data. Of 100 randomly sampled videos, appropriate matching was manually confirmed in all. In total, these videos represented 489,721 minutes of colonoscopy performed by 50 endoscopists (median 214 colonoscopies per endoscopist). The most common procedure indications were polyp screening (47.3 %), surveillance (28.9 %) and inflammatory bowel disease (9.4 %). From these videos, we extracted procedure highlights (identified by image capture; mean 8.5 per colonoscopy) and surrounding frames. Conclusions We report the successful merging of a large database of endoscopy videos stored with limited identifiers to rich patient-level data in a highly accurate manner. This technique facilitates the development of ML algorithms based upon relevant patient outcomes.

Estimation of Instantaneous Oxygen Uptake During Exercise and Daily Activities Using a Wearable Cardio-Electromechanical and Environmental Sensor.

OBJECTIVE: To estimate instantaneous oxygen uptake VO2 with a small, low-cost wearable sensor during exercise and daily activities in order to enable monitoring of energy expenditure (EE) in uncontrolled settings. We aim to do so using a combination of seismocardiogram (SCG), electrocardiogram (ECG) and atmospheric pressure (AP) signals obtained from a minimally obtrusive wearable device. METHODS: In this study, subjects performed a treadmill protocol in a controlled environment and an outside walking protocol in an uncontrolled environment. During testing, the COSMED K5 metabolic system collected gold standard breath-by-breath (BxB) data and a custom-built wearable patch placed on the mid-sternum collected SCG, ECG and AP signals. We extracted features from these signals to estimate the BxB VO2 data obtained from the COSMED system. RESULTS: In estimating instantaneous VO2, we achieved our best results on the treadmill protocol using a combination of SCG (frequency) and AP features (RMSE of 3.68 ± 0.98 ml/kg/min and R2 of 0.77). For the outside protocol, we achieved our best results using a combination of SCG (frequency), ECG and AP features (RMSE of 4.3 ± 1.47 ml/kg/min and R2 of 0.64). In estimating VO2 consumed over one minute intervals during the protocols, our median percentage error was 15.8[Formula: see text] for the treadmill protocol and 20.5[Formula: see text] for the outside protocol. CONCLUSION: SCG, ECG and AP signals from a small wearable patch can enable accurate estimation of instantaneous VO2 in both controlled and uncontrolled settings. SCG signals capturing variation in cardio-mechanical processes, AP signals, and state of the art machine learning models contribute significantly to the accurate estimation of instantaneous VO2. SIGNIFICANCE: Accurate estimation of VO2 with a low cost, minimally obtrusive wearable patch can enable the monitoring of VO2 and EE in everyday settings and make the many applications of these measurements more accessible to the general public.

Non-Invasive Wearable Patch Utilizing Seismocardiography for Peri-Operative Use in Surgical Patients.

OBJECTIVE: Optimizing peri-operative fluid management has been shown to improve patient outcomes and the use of stroke volume (SV) measurement has become an accepted tool to guide fluid therapy. The Transesophageal Doppler (TED) is a validated, minimally invasive device that allows clinical assessment of SV. Unfortunately, the use of the TED is restricted to the intra-operative setting in anesthetized patients and requires constant supervision and periodic adjustment for accurate signal quality. However, post-operative fluid management is also vital for improved outcomes. Currently, there is no device regularly used in clinics that can track patient's SV continuously and non-invasively both during and after surgery. METHODS: In this paper, we propose the use of a wearable patch mounted on the mid-sternum, which captures the seismocardiogram (SCG) and electrocardiogram (ECG) signals continuously to predict SV in patients undergoing major surgery. In a study of 12 patients, hemodynamic data was recorded simultaneously using the TED and wearable patch. Signal processing and regression techniques were used to derive SV from the signals (SCG and ECG) captured by the wearable patch and compare it to values obtained by the TED. RESULTS: The results showed that the combination of SCG and ECG contains substantial information regarding SV, resulting in a correlation and median absolute error between the predicted and reference SV values of 0.81 and 7.56 mL, respectively. SIGNIFICANCE: This work shows promise for the proposed wearable-based methodology to be used as an alternative to TED for continuous patient monitoring and guiding peri-operative fluid management.

Cardiac Function Monitoring for Patients Undergoing Cancer Treatments Using Wearable Seismocardiography: A Proof-of-Concept Study.

Advances in cancer therapeutics have dramatically improved the survival rate and quality of life in patients affected by various cancers, but have been accompanied by treatment-related cardiotoxicity, e.g. left ventricular (LV) dysfunction and/or overt heart failure (HF). Cardiologists thus need to assess cancer treatment-related cardiotoxic risks and have close followups for cancer survivors and patients undergoing cancer treatments using serial echocardiography exams and cardiovascular biomarkers testing. Unfortunately, the cost-prohibitive nature of echocardiography has made these routine follow-ups difficult and not accessible to the growing number of cancer survivors and patients undergoing cancer treatments. There is thus a need to develop a wearable system that can yield similar information at a minimal cost and can be used for remote monitoring of these patients. In this proof-of-concept study, we have investigated the use of wearable seismocardiography (SCG) to monitor LV function non-invasively for patients undergoing cancer treatment. A total of 12 subjects (six with normal LV relaxation, five with impaired relaxation and one with pseudo-normal relaxation) underwent routine echocardiography followed by a standard six-minute walk test. Wearable SCG and electrocardiogram signals were collected during the six-minute walk test and, later, the signal features were compared between subjects with normal and impaired LV relaxation. Pre-ejection period (PEP) from SCG decreased significantly (p < 0.05) during exercise for the subjects with impaired relaxation compared to the subjects with normal relaxation, and changes in PEP/LV ejection time (LVET) were also significantly different between these two groups (p < 0.05). These results suggest that wearable SCG may enable monitoring of patients undergoing cancer treatments by assessing cardiotoxicity.

Wearable Patch-Based Estimation of Oxygen Uptake and Assessment of Clinical Status during Cardiopulmonary Exercise Testing in Patients With Heart Failure.

BACKGROUND: To estimate oxygen uptake (VO2) from cardiopulmonary exercise testing (CPX) using simultaneously recorded seismocardiogram (SCG) and electrocardiogram (ECG) signals captured with a small wearable patch. CPX is an important risk stratification tool for patients with heart failure (HF) owing to the prognostic value of the features derived from the gas exchange variables such as VO2. However, CPX requires specialized equipment, as well as trained professionals to conduct the study. METHODS AND RESULTS: We have conducted a total of 68 CPX tests on 59 patients with HF with reduced ejection fraction (31% women, mean age 55 ± 13 years, ejection fraction 0.27 ± 0.11, 79% stage C). The patients were fitted with a wearable sensing patch and underwent treadmill CPX. We divided the dataset into a training-testing set (n = 44) and a separate validation set (n = 24). We developed globalized (population) regression models to estimate VO2 from the SCG and ECG signals measured continuously with the patch. We further classified the patients as stage D or C using the SCG and ECG features to assess the ability to detect clinical state from the wearable patch measurements alone. We developed the regression and classification model with cross-validation on the training-testing set and validated the models on the validation set. The regression model to estimate VO2 from the wearable features yielded a moderate correlation (R2 of 0.64) with a root mean square error of 2.51 ± 1.12 mL · kg-1 · min-1 on the training-testing set, whereas R2 and root mean square error on the validation set were 0.76 and 2.28 ± 0.93 mL · kg-1 · min-1, respectively. Furthermore, the classification of clinical state yielded accuracy, sensitivity, specificity, and an area under the receiver operating characteristic curve values of 0.84, 0.91, 0.64, and 0.74, respectively, for the training-testing set, and 0.83, 0.86, 0.67, and 0.92, respectively, for the validation set. CONCLUSIONS: Wearable SCG and ECG can assess CPX VO2 and thereby classify clinical status for patients with HF. These methods may provide value in the risk stratification of patients with HF by tracking cardiopulmonary parameters and clinical status outside of specialized settings, potentially allowing for more frequent assessments to be performed during longitudinal monitoring and treatment.

International evaluation of an AI system for breast cancer screening.

Screening mammography aims to identify breast cancer at earlier stages of the disease, when treatment can be more successful1. Despite the existence of screening programmes worldwide, the interpretation of mammograms is affected by high rates of false positives and false negatives2. Here we present an artificial intelligence (AI) system that is capable of surpassing human experts in breast cancer prediction. To assess its performance in the clinical setting, we curated a large representative dataset from the UK and a large enriched dataset from the USA. We show an absolute reduction of 5.7% and 1.2% (USA and UK) in false positives and 9.4% and 2.7% in false negatives. We provide evidence of the ability of the system to generalize from the UK to the USA. In an independent study of six radiologists, the AI system outperformed all of the human readers: the area under the receiver operating characteristic curve (AUC-ROC) for the AI system was greater than the AUC-ROC for the average radiologist by an absolute margin of 11.5%. We ran a simulation in which the AI system participated in the double-reading process that is used in the UK, and found that the AI system maintained non-inferior performance and reduced the workload of the second reader by 88%. This robust assessment of the AI system paves the way for clinical trials to improve the accuracy and efficiency of breast cancer screening.

Seismocardiography and 4D flow MRI reveal impact of aortic valve replacement on chest acceleration and aortic hemodynamics.

Aortic valve replacement (AVR) is a common treatment for severe aortic valve disease, which can adversely affect blood flow in the aorta. Seismocardiography (SCG) measures physical vibrations at the exterior of the chest, which can be sensitive to altered cardiac function and flow dynamics. Magnetic resonance imaging (MRI) can image blood movement, and it can provide depiction and quantification of aortic flow. Here we present SCG and MRI measurements from before and after AVR and ascending aorta replacement, in the case of a woman with bicuspid aortic valve disease and a dilated ascending aorta. SCG measurements show elevated energy during systole indicating stenotic flow before surgery and lowered systolic energy levels after replacement with a prosthetic valve. MRI shows jetting, helical flow before surgery, and cohesive flow after.

Author Correction: End-to-end lung cancer screening with three-dimensional deep learning on low-dose chest computed tomography.

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

End-to-end lung cancer screening with three-dimensional deep learning on low-dose chest computed tomography.

With an estimated 160,000 deaths in 2018, lung cancer is the most common cause of cancer death in the United States1. Lung cancer screening using low-dose computed tomography has been shown to reduce mortality by 20-43% and is now included in US screening guidelines1-6. Existing challenges include inter-grader variability and high false-positive and false-negative rates7-10. We propose a deep learning algorithm that uses a patient's current and prior computed tomography volumes to predict the risk of lung cancer. Our model achieves a state-of-the-art performance (94.4% area under the curve) on 6,716 National Lung Cancer Screening Trial cases, and performs similarly on an independent clinical validation set of 1,139 cases. We conducted two reader studies. When prior computed tomography imaging was not available, our model outperformed all six radiologists with absolute reductions of 11% in false positives and 5% in false negatives. Where prior computed tomography imaging was available, the model performance was on-par with the same radiologists. This creates an opportunity to optimize the screening process via computer assistance and automation. While the vast majority of patients remain unscreened, we show the potential for deep learning models to increase the accuracy, consistency and adoption of lung cancer screening worldwide.

MyPectus: First-in-human pilot study of remote compliance monitoring of teens using dynamic compression bracing to correct pectus carinatum.

BACKGROUND: Patient compliance is a crucial determinant of outcomes in treatments involving medical braces, such as dynamic compression therapy for pectus carinatum (PC). We performed a pilot study to assess a novel, wireless, real-time monitoring system (MyPectus) to address noncompliance. METHODS: Eight patients (10-16years old) with moderately severe PC deformities underwent bracing. Each patient received a data logger device inserted in the compression brace to sense temperature and pressure. The data were transmitted via Bluetooth 4.0 to an iOS smartphone app, then synced to cloud-based storage, and presented to the clinician on a web-based dashboard. Patients received points for brace usage on the app throughout the 4-week study, and completed a survey to capture patient-reported usage patterns. RESULTS: In all 8 patients, the data logger sensed and recorded data, which connected through all MyPectus system components. There were occasional lapses in data collection because of technical difficulties, such as limited storage capacity. Patients reported positive feedback regarding points. CONCLUSIONS: The components of the MyPectus system recorded, stored, and provided data to patients and clinicians. The MyPectus system will inform clinicians about issues related to noncompliance: discrepancy between patient-reported and sensor-reported data regarding brace usage; real-time, actionable information; and patient motivation.

Antenatal maternally-administered phosphodiesterase type 5 inhibitors normalize eNOS expression in the fetal lamb model of congenital diaphragmatic hernia.

PURPOSE: Pulmonary hypertension (pHTN), a main determinant of survival in congenital diaphragmatic hernia (CDH), results from in utero vascular remodeling. Phosphodiesterase type 5 (PDE5) inhibitors have never been used antenatally to treat pHTN. The purpose of this study is to determine if antenatal PDE5 inhibitors can prevent pHTN in the fetal lamb model of CDH. METHODS: CDH was created in pregnant ewes. Postoperatively, pregnant ewes received oral placebo or tadalafil, a PDE5 inhibitor, until delivery. Near term gestation, lambs underwent resuscitations, and lung tissue was snap frozen for protein analysis. RESULTS: Mean cGMP levels were 0.53±0.11 in placebo-treated fetal lambs and 1.73±0.21 in tadalafil-treated fetal lambs (p=0.002). Normalized expression of eNOS was 82%±12% in Normal-Placebo, 61%±5% in CDH-Placebo, 116%±6% in Normal-Tadalafil, and 86%±8% in CDH-Tadalafil lambs. Normalized expression of ß-sGC was 105%±15% in Normal-Placebo, 82%±3% in CDH-Placebo, 158%±16% in Normal-Tadalafil, and 86%±8% in CDH-Tadalafil lambs. Endothelial NOS and ß-sGC were significantly decreased in CDH (p=0.0007 and 0.01 for eNOS and ß-sGC, respectively), and tadalafil significantly increased eNOS expression (p=0.0002). CONCLUSIONS: PDE5 inhibitors can cross the placental barrier. ß-sGC and eNOS are downregulated in fetal lambs with CDH. Antenatal PDE5 inhibitors normalize eNOS and may prevent in utero vascular remodeling in CDH.

Implantable ultra-low pulmonary pressure monitoring system for fetal surgery.

Congenital pulmonary hypoplasia is a devastating condition affecting fetal and newborn pulmonary physiology, resulting in great morbidity and mortality. The fetal lung develops in a fluid-filled environment. In this work, we describe a novel, implantable pressure sensing and recording device which we use to study the pressures present in the fetal pulmonary tree throughout gestation. The system achieves 0.18 cm H2O resolution and can record for twenty one days continuously at 256 Hz. Sample tracings of in vivo fetal lamb recordings are shown.

Dynamic tracheal occlusion improves lung morphometrics and function in the fetal lamb model of congenital diaphragmatic hernia.

BACKGROUND: Congenital diaphragmatic hernia (CDH) is associated with significant neonatal morbidity and mortality. Although prenatal complete tracheal occlusion (cTO) causes hypoplastic CDH lungs to enlarge, improved lung function has not been demonstrated. Furthermore, cTO interferes with the dynamic pressure change and fluid flow associated with fetal breathing. PURPOSE: The purpose of the study was to assess a novel dynamic tracheal occlusion (dTO) device that preserves pressure changes and fluid flow. METHODS: In this pilot study, CDH was created in fetal lambs at 65 days of gestational age (GA). At 110 days GA, a cTO device (n = 3) or a dTO device (n = 4) was placed in the fetal trachea. At 135 days GA, lambs were delivered and resuscitated. Unoperated lamb co-twins (n = 5), sham thoracotomy lambs (n = 2), and untreated CDH lambs (n = 3) served as controls. RESULTS: Tracheal opening pressure, lung volume, lung fluid total protein, and phospholipid were significantly higher in the cTO group than in the dTO and unoperated control groups. Maximal oxygenation and lung compliance were significantly lower in the cTO group when compared with the unoperated control and dTO groups. CONCLUSION: Preliminary results suggest that in the fetal lamb CDH model, dTO restores normal lung morphometrics and function, whereas cTO leads to enlarged but less functional lungs.