Machine Learning for Adrenal Gland Segmentation and Classification of Normal and Adrenal Masses at CT.

Background Adrenal masses are common, but radiology reporting and recommendations for management can be variable. Purpose To create a machine learning algorithm to segment adrenal glands on contrast-enhanced CT images and classify glands as normal or mass-containing and to assess algorithm performance. Materials and Methods This retrospective study included two groups of contrast-enhanced abdominal CT examinations (development data set and secondary test set). Adrenal glands in the development data set were manually segmented by radiologists. Images in both the development data set and the secondary test set were manually classified as normal or mass-containing. Deep learning segmentation and classification models were trained on the development data set and evaluated on both data sets. Segmentation performance was evaluated with use of the Dice similarity coefficient (DSC), and classification performance with use of sensitivity and specificity. Results The development data set contained 274 CT examinations (251 patients; median age, 61 years; 133 women), and the secondary test set contained 991 CT examinations (991 patients; median age, 62 years; 578 women). The median model DSC on the development test set was 0.80 (IQR, 0.78-0.89) for normal glands and 0.84 (IQR, 0.79-0.90) for adrenal masses. On the development reader set, the median interreader DSC was 0.89 (IQR, 0.78-0.93) for normal glands and 0.89 (IQR, 0.85-0.97) for adrenal masses. Interreader DSC for radiologist manual segmentation did not differ from automated machine segmentation (P = .35). On the development test set, the model had a classification sensitivity of 83% (95% CI: 55, 95) and specificity of 89% (95% CI: 75, 96). On the secondary test set, the model had a classification sensitivity of 69% (95% CI: 58, 79) and specificity of 91% (95% CI: 90, 92). Conclusion A two-stage machine learning pipeline was able to segment the adrenal glands and differentiate normal adrenal glands from those containing masses. © RSNA, 2022 Online supplemental material is available for this article.

Implementation of Clinical Artificial Intelligence in Radiology: Who Decides and How?

As the role of artificial intelligence (AI) in clinical practice evolves, governance structures oversee the implementation, maintenance, and monitoring of clinical AI algorithms to enhance quality, manage resources, and ensure patient safety. In this article, a framework is established for the infrastructure required for clinical AI implementation and presents a road map for governance. The road map answers four key questions: Who decides which tools to implement? What factors should be considered when assessing an application for implementation? How should applications be implemented in clinical practice? Finally, how should tools be monitored and maintained after clinical implementation? Among the many challenges for the implementation of AI in clinical practice, devising flexible governance structures that can quickly adapt to a changing environment will be essential to ensure quality patient care and practice improvement objectives.

Use of Artificial Intelligence in Clinical Neurology.

Artificial intelligence is already innovating in the provision of neurologic care. This review explores key artificial intelligence concepts; their application to neurologic diagnosis, prognosis, and treatment; and challenges that await their broader adoption. The development of new diagnostic biomarkers, individualization of prognostic information, and improved access to treatment are among the plethora of possibilities. These advances, however, reflect only the tip of the iceberg for the ways in which artificial intelligence may transform neurologic care in the future.

Multicenter Assessment of CT Pneumonia Analysis Prototype for Predicting Disease Severity and Patient Outcome.

To perform a multicenter assessment of the CT Pneumonia Analysis prototype for predicting disease severity and patient outcome in COVID-19 pneumonia both without and with integration of clinical information. Our IRB-approved observational study included consecutive 241 adult patients (> 18 years; 105 females; 136 males) with RT-PCR-positive COVID-19 pneumonia who underwent non-contrast chest CT at one of the two tertiary care hospitals (site A: Massachusetts General Hospital, USA; site B: Firoozgar Hospital Iran). We recorded patient age, gender, comorbid conditions, laboratory values, intensive care unit (ICU) admission, mechanical ventilation, and final outcome (recovery or death). Two thoracic radiologists reviewed all chest CTs to record type, extent of pulmonary opacities based on the percentage of lobe involved, and severity of respiratory motion artifacts. Thin-section CT images were processed with the prototype (Siemens Healthineers) to obtain quantitative features including lung volumes, volume and percentage of all-type and high-attenuation opacities (≥ -200 HU), and mean HU and standard deviation of opacities within a given lung region. These values are estimated for the total combined lung volume, and separately for each lung and each lung lobe. Multivariable analyses of variance (MANOVA) and multiple logistic regression were performed for data analyses. About 26% of chest CTs (62/241) had moderate to severe motion artifacts. There were no significant differences in the AUCs of quantitative features for predicting disease severity with and without motion artifacts (AUC 0.94-0.97) as well as for predicting patient outcome (AUC 0.7-0.77) (p > 0.5). Combination of the volume of all-attenuation opacities and the percentage of high-attenuation opacities (AUC 0.76-0.82, 95% confidence interval (CI) 0.73-0.82) had higher AUC for predicting ICU admission than the subjective severity scores (AUC 0.69-0.77, 95% CI 0.69-0.81). Despite a high frequency of motion artifacts, quantitative features of pulmonary opacities from chest CT can help differentiate patients with favorable and adverse outcomes.

Using DICOM Metadata for Radiological Image Series Categorization: a Feasibility Study on Large Clinical Brain MRI Datasets.

The growing interest in machine learning (ML) in healthcare is driven by the promise of improved patient care. However, how many ML algorithms are currently being used in clinical practice? While the technology is present, as demonstrated in a variety of commercial products, clinical integration is hampered by a lack of infrastructure, processes, and tools. In particular, automating the selection of relevant series for a particular algorithm remains challenging. In this work, we propose a methodology to automate the identification of brain MRI sequences so that we can automatically route the relevant inputs for further image-related algorithms. The method relies on metadata required by the Digital Imaging and Communications in Medicine (DICOM) standard, resulting in generalizability and high efficiency (less than 0.4 ms/series). To support our claims, we test our approach on two large brain MRI datasets (40,000 studies in total) from two different institutions on two different continents. We demonstrate high levels of accuracy (ranging from 97.4 to 99.96%) and generalizability across the institutions. Given the complexity and variability of brain MRI protocols, we are confident that similar techniques could be applied to other forms of radiological imaging.

Primary care provider perspectives on the value of opportunistic CT screening.

BACKGROUND: Clinical adoption of AI applications requires stakeholders see value in their use. AI-enabled opportunistic-CT-screening (OS) capitalizes on incidentally-detected findings within CTs for potential health benefit. This study evaluates primary care providers' (PCP) perspectives on OS. METHODS: A survey was distributed to US Internal and Family Medicine residencies. Assessed were familiarity with AI and OS, perspectives on potential value/costs, communication of results, and technology implementation. RESULTS: 62 % of respondents (n = 71) were in Family Medicine, 64.8 % practiced in community hospitals. Although 74.6 % of respondents had heard of AI/machine learning, 95.8 % had little-to-no familiarity with OS. The majority reported little-to-no trust in AI. Reported concerns included AI accuracy (74.6 %) and unknown liability (73.2 %). 78.9 % of respondents reported that OS applications would require radiologist oversight. 53.5 % preferred OS results be included in a separate "screening" section within the Radiology report, accompanied by condition risks and management recommendations. The majority of respondents reported results would likely affect clinical management for all queried applications, and that atherosclerotic cardiovascular disease risk, abdominal aortic aneurysm, and liver fibrosis should be included within every CT report regardless of reason for examination. 70.5 % felt that PCP practices are unlikely to pay for OS. Added costs to the patient (91.5 %), the healthcare provider (77.5 %), and unknown liability (74.6 %) were the most frequently reported concerns. CONCLUSION: PCP preferences and concerns around AI-enabled OS offer insights into clinical value and costs. As AI applications grow, feedback from end-users should be considered in the development of such technology to optimize implementation and adoption. Increasing stakeholder familiarity with AI may be a critical prerequisite first step before stakeholders consider implementation.

No code machine learning: validating the approach on use-case for classifying clavicle fractures.

PURPOSE: We created an infrastructure for no code machine learning (NML) platform for non-programming physicians to create NML model. We tested the platform by creating an NML model for classifying radiographs for the presence and absence of clavicle fractures. METHODS: Our IRB-approved retrospective study included 4135 clavicle radiographs from 2039 patients (mean age 52 ± 20 years, F:M 1022:1017) from 13 hospitals. Each patient had two-view clavicle radiographs with axial and anterior-posterior projections. The positive radiographs had either displaced or non-displaced clavicle fractures. We configured the NML platform to automatically retrieve the eligible exams using the series' unique identification from the hospital virtual network archive via web access to DICOM Objects. The platform trained a model until the validation loss plateaus. Once the testing was complete, the platform provided the receiver operating characteristics curve and confusion matrix for estimating sensitivity, specificity, and accuracy. RESULTS: The NML platform successfully retrieved 3917 radiographs (3917/4135, 94.7 %) and parsed them for creating a ML classifier with 2151 radiographs in the training, 100 radiographs for validation, and 1666 radiographs in testing datasets (772 radiographs with clavicle fracture, 894 without clavicle fracture). The network identified clavicle fracture with 90 % sensitivity, 87 % specificity, and 88 % accuracy with AUC of 0.95 (confidence interval 0.94-0.96). CONCLUSION: A NML platform can help physicians create and test machine learning models from multicenter imaging datasets such as the one in our study for classifying radiographs based on the presence of clavicle fracture.

Evaluation of an artificial intelligence model for identification of mass effect and vasogenic edema on computed tomography of the head.

BACKGROUND AND PURPOSE: Mass effect and vasogenic edema are critical findings on CT of the head. This study compared the accuracy of an artificial intelligence model (Annalise Enterprise CTB) to consensus neuroradiologist interpretations in detecting mass effect and vasogenic edema. MATERIALS AND METHODS: A retrospective standalone performance assessment was conducted on datasets of non-contrast CT head cases acquired between 2016 and 2022 for each finding. The cases were obtained from patients aged 18 years or older from five hospitals in the United States. The positive cases were selected consecutively based on the original clinical reports using natural language processing and manual confirmation. The negative cases were selected by taking the next negative case acquired from the same CT scanner after positive cases. Each case was interpreted independently by up to three neuroradiologists to establish consensus interpretations. Each case was then interpreted by the AI model for the presence of the relevant finding. The neuroradiologists were provided with the entire CT study. The AI model separately received thin (≤1.5mm) and/or thick (>1.5 and ≤5mm) axial series. RESULTS: The two cohorts included 818 cases for mass effect and 310 cases for vasogenic edema. The AI model identified mass effect with sensitivity 96.6% (95% CI, 94.9-98.2) and specificity 89.8% (95% CI, 84.7-94.2) for the thin series, and 95.3% (95% CI, 93.5-96.8) and 93.1% (95% CI, 89.1-96.6) for the thick series. It identified vasogenic edema with sensitivity 90.2% (95% CI, 82.0-96.7) and specificity 93.5% (95% CI, 88.9-97.2) for the thin series, and 90.0% (95% CI, 84.0-96.0) and 95.5% (95% CI, 92.5-98.0) for the thick series. The corresponding areas under the curve were at least 0.980. CONCLUSIONS: The assessed AI model accurately identified mass effect and vasogenic edema in this CT dataset. It could assist the clinical workflow by prioritizing interpretation of abnormal cases, which could benefit patients through earlier identification and subsequent treatment. ABBREVIATIONS: AI = artificial intelligence; AUC = area under the curve; CADt = computer assisted triage devices; FDA = Food and Drug Administration; NPV = negative predictive value; PPV = positive predictive value; SD = standard deviation.

Neuropsychiatric lupus: classification criteria in neuroimaging studies.

This systematic review described the criteria and main evaluations methods procedures used to classify neuropsychiatric systemic lupus erythematosus (NPSLE) patients. Also, within the evaluations methods, this review aimed to identify the main contributions of neuropsychological measurements in neuroimaging studies. A search was conducted in PubMed, EMBASE and SCOPUS databases with the terms related to neuropsychiatric syndromes, systemic lupus erythematosus, and neuroimaging techniques. Sixty-six abstracts were found; only 20 were completely analyzed and included. Results indicated that the 1999 American College of Rheumatology (ACR) criteria is the most used to classify NPSLE samples together with laboratorial, cognitive, neurological and psychiatric assessment procedures. However, the recommended ACR assessment procedures to classify NPSLE patients are being used incompletely, especially the neuropsychological batteries. Neuropsychological instruments and neuroimaging techniques have been used mostly to characterize NPSLE samples, instead of contributing to their classifications. The most described syndromes in neuroimaging studies have been seizure/cerebrovascular disease followed by cognitive dysfunctions as well as headache disorder.

Automatic segmentation and measurement of tracheal collapsibility in tracheomalacia.

PURPOSE: To assess feasibility of automated segmentation and measurement of tracheal collapsibility for detecting tracheomalacia on inspiratory and expiratory chest CT images. METHODS: Our study included 123 patients (age 67 ± 11 years; female: male 69:54) who underwent clinically indicated chest CT examinations in both inspiration and expiration phases. A thoracic radiologist measured anteroposterior length of trachea in inspiration and expiration phase image at the level of maximum collapsibility or aortic arch (in absence of luminal change). Separately, another investigator separately processed the inspiratory and expiratory DICOM CT images with Airway Segmentation component of a commercial COPD software (IntelliSpace Portal, Philips Healthcare). Upon segmentation, the software automatically estimated average lumen diameter (in mm) and lumen area (sq.mm) both along the entire length of trachea and at the level of aortic arch. Data were analyzed with independent t-tests and area under the receiver operating characteristic curve (AUC). RESULTS: Of the 123 patients, 48 patients had tracheomalacia and 75 patients did not. Ratios of inspiration to expiration phases average lumen area and lumen diameter from the length of trachea had the highest AUC of 0.93 (95% CI = 0.88-0.97) for differentiating presence and absence of tracheomalacia. A decrease of ≥25% in average lumen diameter had sensitivity of 82% and specificity of 87% for detecting tracheomalacia. A decrease of ≥40% in the average lumen area had sensitivity and specificity of 86% for detecting tracheomalacia. CONCLUSION: Automatic segmentation and measurement of tracheal dimension over the entire tracheal length is more accurate than a single-level measurement for detecting tracheomalacia.

Auto-Detection of Motion Artifacts on CT Pulmonary Angiograms with a Physician-Trained AI Algorithm.

Purpose: Motion-impaired CT images can result in limited or suboptimal diagnostic interpretation (with missed or miscalled lesions) and patient recall. We trained and tested an artificial intelligence (AI) model for identifying substantial motion artifacts on CT pulmonary angiography (CTPA) that have a negative impact on diagnostic interpretation. Methods: With IRB approval and HIPAA compliance, we queried our multicenter radiology report database (mPower, Nuance) for CTPA reports between July 2015 and March 2022 for the following terms: "motion artifacts", "respiratory motion", "technically inadequate", and "suboptimal" or "limited exam". All CTPA reports were from two quaternary (Site A, n = 335; B, n = 259) and a community (C, n = 199) healthcare sites. A thoracic radiologist reviewed CT images of all positive hits for motion artifacts (present or absent) and their severity (no diagnostic effect or major diagnostic impairment). Coronal multiplanar images from 793 CTPA exams were de-identified and exported offline into an AI model building prototype (Cognex Vision Pro, Cognex Corporation) to train an AI model to perform two-class classification ("motion" or "no motion") with data from the three sites (70% training dataset, n = 554; 30% validation dataset, n = 239). Separately, data from Site A and Site C were used for training and validating; testing was performed on the Site B CTPA exams. A five-fold repeated cross-validation was performed to evaluate the model performance with accuracy and receiver operating characteristics analysis (ROC). Results: Among the CTPA images from 793 patients (mean age 63 ± 17 years; 391 males, 402 females), 372 had no motion artifacts, and 421 had substantial motion artifacts. The statistics for the average performance of the AI model after five-fold repeated cross-validation for the two-class classification included 94% sensitivity, 91% specificity, 93% accuracy, and 0.93 area under the ROC curve (AUC: 95% CI 0.89-0.97). Conclusion: The AI model used in this study can successfully identify CTPA exams with diagnostic interpretation limiting motion artifacts in multicenter training and test datasets. Clinical relevance: The AI model used in the study can help alert technologists about the presence of substantial motion artifacts on CTPA, where a repeat image acquisition can help salvage diagnostic information.

Addressing the Challenges of Implementing Artificial Intelligence Tools in Clinical Practice: Principles From Experience.

The multitude of artificial intelligence (AI)-based solutions, vendors, and platforms poses a challenging proposition to an already complex clinical radiology practice. Apart from assessing and ensuring acceptable local performance and workflow fit to improve imaging services, AI tools require multiple stakeholders, including clinical, technical, and financial, who collaborate to move potential deployable applications to full clinical deployment in a structured and efficient manner. Postdeployment monitoring and surveillance of such tools require an infrastructure that ensures proper and safe use. Herein, the authors describe their experience and framework for implementing and supporting the use of AI applications in radiology workflow.

Radiologist-Trained AI Model for Identifying Suboptimal Chest-Radiographs.

RATIONALE AND OBJECTIVES: Suboptimal chest radiographs (CXR) can limit interpretation of critical findings. Radiologist-trained AI models were evaluated for differentiating suboptimal(sCXR) and optimal(oCXR) chest radiographs. MATERIALS AND METHODS: Our IRB-approved study included 3278 CXRs from adult patients (mean age 55 ± 20 years) identified from a retrospective search of CXR in radiology reports from 5 sites. A chest radiologist reviewed all CXRs for the cause of suboptimality. The de-identified CXRs were uploaded into an AI server application for training and testing 5 AI models. The training set consisted of 2202 CXRs (n = 807 oCXR; n = 1395 sCXR) while 1076 CXRs (n = 729 sCXR; n = 347 oCXR) were used for testing. Data were analyzed with the Area under the curve (AUC) for the model's ability to classify oCXR and sCXR correctly. RESULTS: For the two-class classification into sCXR or oCXR from all sites, for CXR with missing anatomy, AI had sensitivity, specificity, accuracy, and AUC of 78%, 95%, 91%, 0.87(95% CI 0.82-0.92), respectively. AI identified obscured thoracic anatomy with 91% sensitivity, 97% specificity, 95% accuracy, and 0.94 AUC (95% CI 0.90-0.97). Inadequate exposure with 90% sensitivity, 93% specificity, 92% accuracy, and AUC of 0.91 (95% CI 0.88-0.95). The presence of low lung volume was identified with 96% sensitivity, 92% specificity, 93% accuracy, and 0.94 AUC (95% CI 0.92-0.96). The sensitivity, specificity, accuracy, and AUC of AI in identifying patient rotation were 92%, 96%, 95%, and 0.94 (95% CI 0.91-0.98), respectively. CONCLUSION: The radiologist-trained AI models can accurately classify optimal and suboptimal CXRs. Such AI models at the front end of radiographic equipment can enable radiographers to repeat sCXRs when necessary.

Head CT deep learning model is highly accurate for early infarct estimation.

Non-contrast head CT (NCCT) is extremely insensitive for early (< 3-6 h) acute infarct identification. We developed a deep learning model that detects and delineates suspected early acute infarcts on NCCT, using diffusion MRI as ground truth (3566 NCCT/MRI training patient pairs). The model substantially outperformed 3 expert neuroradiologists on a test set of 150 CT scans of patients who were potential candidates for thrombectomy (60 stroke-negative, 90 stroke-positive middle cerebral artery territory only infarcts), with sensitivity 96% (specificity 72%) for the model versus 61-66% (specificity 90-92%) for the experts; model infarct volume estimates also strongly correlated with those of diffusion MRI (r2 > 0.98). When this 150 CT test set was expanded to include a total of 364 CT scans with a more heterogeneous distribution of infarct locations (94 stroke-negative, 270 stroke-positive mixed territory infarcts), model sensitivity was 97%, specificity 99%, for detection of infarcts larger than the 70 mL volume threshold used for patient selection in several major randomized controlled trials of thrombectomy treatment.

Predictive values of AI-based triage model in suboptimal CT pulmonary angiography.

PURPOSE: We evaluated and compared performance of an acute pulmonary embolism (PE) triaging artificial intelligence (PE-AI) model in suboptimal and optimal CT pulmonary angiography (CTPA). METHODS: In an IRB approved, retrospective study we identified 104 consecutive, suboptimal CTPA which were deemed as suboptimal for PE evaluation in radiology reports due to motion, artifacts and/or inadequate contrast enhancement. We enriched this dataset, with additional 226 optimal CTPA (over same timeframe as suboptimal CTPA) with and without PE. Two thoracic radiologists (ground truth) independently reviewed all 330 CTPA for adequacy (to assess PE down to distal segmental level), reason for suboptimal CTPA (artifacts or poor contrast enhancement), as well as for presence and location of PE. CT values (HU) were measured in the main pulmonary artery. Same attributes were assessed in 80 patients who had repeat or follow-up CTPA following suboptimal CTPA. All CTPA were processed with the PE-AI (Aidoc). RESULTS: Among 104 suboptimal CTPA (mean age ± standard deviation 56 ± 15 years), 18/104 (17%) were misclassified as suboptimal for PE evaluation in their radiology reports but relabeled as optimal on ground truth evaluation. Of 226 optimal CTPA, 47 (21%) were reclassified as suboptimal CTPA. PEs were present in 97/330 CTPA. PE-AI had similar performance on suboptimal CTPA (sensitivity 100%; specificity 89%; AUC 0.89, 95% CI 0.80-0.98) and optimal CTPA (sensitivity 96%; specificity 92%; AUC 0.87, 95% CI 0.81-0.93). CONCLUSION: Suboptimal CTPA examinations do not impair the performance of PE-AI triage model; AI retains clinically meaningful sensitivity and high specificity regardless of diagnostic quality.

Computer-assisted Reporting and Decision Support Increases Compliance with Follow-up Imaging and Hormonal Screening of Adrenal Incidentalomas.

OBJECTIVE: To assess the impact of using a computer-assisted reporting and decision support (CAR/DS) tool at the radiologist point-of-care on ordering provider compliance with recommendations for adrenal incidentaloma workup. METHOD: Abdominal CT reports describing adrenal incidentalomas (2014 - 2016) were retrospectively extracted from the radiology database. Exclusion criteria were history of cancer, suspected functioning adrenal tumor, dominant nodule size < 1 cm or ≥ 4 cm, myelolipomas, cysts, and hematomas. Multivariable logistic regression models were employed to predict follow-up imaging (FUI) and hormonal screening orders as a function of patient age and sex, nodule size, and CAR/DS use. CAR/DS reports were compared to conventional reports regarding ordering provider compliance with, frequency, and completeness of, guideline-warranted recommendations for FUI and hormonal screening of adrenal incidentalomas using Chi-square test. RESULT: Of 174 patients (mean age 62.4; 51.1% women) with adrenal incidentalomas, 62% (108/174) received CAR/DS-based recommendations versus 38% (66/174) unassisted recommendations. CAR/DS use was an independent predictor of provider compliance both with FUI (Odds Ratio [OR]=2.47, p = 0.02) and hormonal screening (OR=2.38, p = 0.04). CAR/DS reports recommended FUI (97.2%,105/108) and hormonal screening (87.0%,94/108) more often than conventional reports (respectively, 69.7% [46/66], 3.0% [2/66], both p <0.0001). CAR/DS recommendations more frequently included instructions for FUI time, protocol, and modality than conventional reports (all p <0.001). CONCLUSION: Ordering providers were at least twice as likely to comply with report recommendations for FUI and hormonal evaluation of adrenal incidentalomas generated using CAR/DS versus unassisted reporting. CAR/DS-directed recommendations were more adherent to guidelines than those generated without.

FDA-regulated AI Algorithms: Trends, Strengths, and Gaps of Validation Studies.

RATIONALE AND OBJECTIVES: To assess key trends, strengths, and gaps in validation studies of the Food and Drug Administration (FDA)-regulated imaging-based artificial intelligence/machine learning (AI/ML) algorithms. MATERIALS AND METHODS: We audited publicly available details of regulated AI/ML algorithms in imaging from 2008 until April 2021. We reviewed 127 regulated software (118 AI/ML) to classify information related to their parent company, subspecialty, body area and specific anatomy type, imaging modality, date of FDA clearance, indications for use, target pathology (such as trauma) and findings (such as fracture), technique (CAD triage, CAD detection and/or characterization, CAD acquisition or improvement, and image processing/quantification), product performance, presence, type, strength and availability of clinical validation data. Pertaining to validation data, where available, we recorded the number of patients or studies included, sensitivity, specificity, accuracy, and/or receiver operating characteristic area under the curve, along with information on ground-truthing of use-cases. Data were analyzed with pivot tables and charts for descriptive statistics and trends. RESULTS: We noted an increasing number of FDA-regulated AI/ML from 2008 to 2021. Seventeen (17/118) regulated AI/ML algorithms posted no validation claims or data. Just 9/118 reviewed AI/ML algorithms had a validation dataset sizes of over 1000 patients. The most common type of AI/ML included image processing/quantification (IPQ; n = 59/118), and triage (CADt; n = 27/118). Brain, breast, and lungs dominated the targeted body regions of interest. CONCLUSION: Insufficient public information on validation datasets in several FDA-regulated AI/ML algorithms makes it difficult to justify clinical applications since their generalizability and presence of bias cannot be inferred.

Radiologist-Trained and -Tested (R2.2.4) Deep Learning Models for Identifying Anatomical Landmarks in Chest CT.

(1) Background: Optimal anatomic coverage is important for radiation-dose optimization. We trained and tested (R2.2.4) two (R3-2) deep learning (DL) algorithms on a machine vision tool library platform (Cognex Vision Pro Deep Learning software) to recognize anatomic landmarks and classify chest CT as those with optimum, under-scanned, or over-scanned scan length. (2) Methods: To test our hypothesis, we performed a study with 428 consecutive chest CT examinations (mean age 70 ± 14 years; male:female 190:238) performed at one of the four hospitals. CT examinations from two hospitals were used to train the DL classification algorithms to identify lung apices and bases. The developed algorithms were then tested on the data from the remaining two hospitals. For each CT, we recorded the scan lengths above and below the lung apices and bases. Model performance was assessed with receiver operating characteristics (ROC) analysis. (3) Results: The two DL models for lung apex and bases had high sensitivity, specificity, accuracy, and areas under the curve (AUC) for identifying under-scanning (100%, 99%, 99%, and 0.999 (95% CI 0.996-1.000)) and over-scanning (99%, 99%, 99%, and 0.998 (95%CI 0.992-1.000)). (4) Conclusions: Our DL models can accurately identify markers for missing anatomic coverage and over-scanning in chest CTs.

Validation pipeline for machine learning algorithm assessment for multiple vendors.

A standardized objective evaluation method is needed to compare machine learning (ML) algorithms as these tools become available for clinical use. Therefore, we designed, built, and tested an evaluation pipeline with the goal of normalizing performance measurement of independently developed algorithms, using a common test dataset of our clinical imaging. Three vendor applications for detecting solid, part-solid, and groundglass lung nodules in chest CT examinations were assessed in this retrospective study using our data-preprocessing and algorithm assessment chain. The pipeline included tools for image cohort creation and de-identification; report and image annotation for ground-truth labeling; server partitioning to receive vendor "black box" algorithms and to enable model testing on our internal clinical data (100 chest CTs with 243 nodules) from within our security firewall; model validation and result visualization; and performance assessment calculating algorithm recall, precision, and receiver operating characteristic curves (ROC). Algorithm true positives, false positives, false negatives, recall, and precision for detecting lung nodules were as follows: Vendor-1 (194, 23, 49, 0.80, 0.89); Vendor-2 (182, 270, 61, 0.75, 0.40); Vendor-3 (75, 120, 168, 0.32, 0.39). The AUCs for detection of solid (0.61-0.74), groundglass (0.66-0.86) and part-solid (0.52-0.86) nodules varied between the three vendors. Our ML model validation pipeline enabled testing of multi-vendor algorithms within the institutional firewall. Wide variations in algorithm performance for detection as well as classification of lung nodules justifies the premise for a standardized objective ML algorithm evaluation process.