Idiopathic pulmonary fibrosis (IPF) is a fatal lung disease characterized by an unpredictable progressive declinein lung function. Natural history of IPF is unknown and the prediction of disease progression at the time ofdiagnosis is notoriously difficult. High resolution computed tomography (HRCT) has been used for the diagnosisof IPF, but not generally for monitoring purpose. The objective of this work is to develop a novel predictivemodel for the radiological progression pattern at voxel-wise level using only baseline HRCT scans. Mainly, thereare two challenges: (a) obtaining a data set of features for region of interest (ROI) on baseline HRCT scans andtheir follow-up status; and (b) simultaneously selecting important features from high-dimensional space, andoptimizing the prediction performance. We resolved the first challenge by implementing a study design andhaving an expert radiologist contour ROIs at baseline scans, depending on its progression status in follow-upvisits. For the second challenge, we integrated the feature selection with prediction by developing an algorithmusing a wrapper method that combines quantum particle swarm optimization to select a small number of featureswith random forest to classify early patterns of progression. We applied our proposed algorithm to analyzeanonymized HRCT images from 50 IPF subjects from a multi-center clinical trial. We showed that it yields aparsimonious model with 81.8% sensitivity, 82.2% specificity and an overall accuracy rate of 82.1% at the ROIlevel. These results are superior to other popular feature selections and classification methods, in that ourmethod produces higher accuracy in prediction of progression and more balanced sensitivity and specificity witha smaller number of selected features. Our work is the first approach to show that it is possible to use onlybaseline HRCT scans to predict progressive ROIs at 6 months to 1year follow-ups using artificial intelligence.

Background

Idiopathic pulmonary fibrosis (IPF) is a fatal interstitial lung disease characterized by an unpredictable decline in lung function. Predicting IPF progression from the early changes in lung function tests have known to be a challenge due to acute exacerbation. Although it is unpredictable, the neighboring regions of fibrotic reticulation increase during IPF's progression. With this clinical information, quantitative characteristics of high-resolution computed tomography (HRCT) and a statistical learning paradigm, the aim is to build a model to predict IPF progression.

Design

A paired set of anonymized 193 HRCT images from IPF subjects with 6-12 month intervals were collected retrospectively. The study was conducted in two parts: (1) Part A collects the ground truth in small regions of interest (ROIs) with labels of "expected to progress" or "expected to be stable" at baseline HRCT and develop a statistical learning model to classify voxels in the ROIs. (2) Part B uses the voxel-level classifier from Part A to produce whole-lung level scores of a single-scan total probability's (STP) baseline.

Methods

Using annotated ROIs from 71 subjects' HRCT scans in Part A, we applied Quantum Particle Swarm Optimization-Random Forest (QPSO-RF) to build the classifier. Then, 122 subjects' HRCT scans were used to test the prediction. Using Spearman rank correlations and survival analyses, we ascertained STP associations with 6-12 month changes in quantitative lung fibrosis and forced vital capacity.

Conclusion

This study can serve as a reference for collecting ground truth, and developing statistical learning techniques to predict progression in medical imaging.

Purpose

Domain knowledge (DK) acquired from prior studies is important for medical diagnosis. This paper leverages the population-level DK using an optimality design criterion to train a deep learning model in an end-to-end manner. In this study, the problem of interest is at the patient level to diagnose a subject with idiopathic pulmonary fibrosis (IPF) among subjects with interstitial lung disease (ILD) using a computed tomography (CT). IPF diagnosis is a complicated process with multidisciplinary discussion with experts and is subject to interobserver variability, even for experienced radiologists. To this end, we propose a new statistical method to construct a time/memory-efficient IPF diagnosis model using axial chest CT and DK, along with an optimality design criterion via a DK-enhanced loss function of deep learning.

Methods

Four state-of-the-art two-dimensional convolutional neural network (2D-CNN) architectures (MobileNet, VGG16, ResNet-50, and DenseNet-121) and one baseline 2D-CNN are implemented to automatically diagnose IPF among ILD patients. Axial lung CT images are retrospectively acquired from 389 IPF patients and 700 non-IPF ILD patients in five multicenter clinical trials. To enrich the sample size and boost model performance, we sample 20 three-slice samples (triplets) from each CT scan, where these three slices are randomly selected from the top, middle, and bottom of both lungs respectively. Model performance is evaluated using a fivefold cross-validation, where each fold was stratified using a fixed proportion of IPF vs non-IPF.

Results

Using DK-enhanced loss function increases the model performance of the baseline CNN model from 0.77 to 0.89 in terms of study-wise accuracy. Four other well-developed models reach satisfactory model performance with an overall accuracy >0.95 but the benefits brought on by the DK-enhanced loss function is not noticeable.

Conclusions

We believe this is the first attempt that (a) uses population-level DK with an optimal design criterion to train deep learning-based diagnostic models in an end-to-end manner and (b) focuses on patient-level IPF diagnosis. Further evaluation of using population-level DK on prospective studies is warranted and is underway.

Background

Deep learning (DL)-based systems have not yet been broadly implemented in clinical practice, in part due to unknown robustness across multiple imaging protocols.

Purpose

To this end, we aim to evaluate the performance of several previously developed DL-based models, which were trained to distinguish idiopathic pulmonary fibrosis (IPF) from non-IPF among interstitial lung disease (ILD) patients, under standardized reference CT imaging protocols. In this study, we utilized CT scans from non-IPF ILD subjects, acquired using various imaging protocols, to assess the model performance.

Methods

Three DL-based models, including one 2D and two 3D models, have been previously developed to classify ILD patients into IPF or non-IPF based on chest CT scans. These models were trained on CT image data from 389 IPF and 700 non-IPF ILD patients, retrospectively, obtained from five multicenter studies. For some patients, multiple CT scans were acquired (e.g., one at inhalation and one at exhalation) and/or reconstructed (e.g., thin slice and/or thick slice). Thus, for each patient, one CT image dataset was selected to be used in the construction of the classification model, so the parameters of that data set serve as the reference conditions. In one non-IPF ILD study, due to its specific study protocol, many patients had multiple CT image data sets that were acquired under both prone and supine positions and/or reconstructed under different imaging parameters. Therefore, to assess the robustness of the previously developed models under different (e.g., non-reference) imaging protocols, we identified 343 subjects from this study who had CT data from both the reference condition (used in model construction) and non-reference conditions (e.g., evaluation conditions), which we used in this model evaluation analysis. We reported the specificities from three model under the non-reference conditions. Generalized linear mixed effects model (GLMM) was utilized to identify the significant CT technical and clinical parameters that were associated with getting inconsistent diagnostic results between reference and evaluation conditions. Selected parameters include effective tube current-time product (known as "effective mAs"), reconstruction kernels, slice thickness, patient orientation (prone or supine), CT scanner model, and clinical diagnosis. Limitations include the retrospective nature of this study.

Results

For all three DL models, the overall specificity of the previously trained IPF diagnosis model decreased (p < 0.05 for two out of three models). GLMM further suggests that for at least one out of three models, mean effective mAs across the scan is the key factor that leads to the decrease in model predictive performance (p < 0.001); the difference of mean effective mAs between the reference and evaluation conditions (p = 0.03) and slice thickness (3 mm; p = 0.03) are flagged as significant factors for one out of three models; other factors are not statistically significant (p > 0.05).

Conclusion

Preliminary findings demonstrated the lack of robustness of IPF diagnosis model when the DL-based model is applied to CT series collected under different imaging protocols, which indicated that care should be taken as to the acquisition and reconstruction conditions used when developing and deploying DL models into clinical practice.

Purpose

Recent studies have demonstrated a lack of reproducibility of radiomic features in response to variations in CT parameters. In addition, reproducibility of radiomic features has not been well established in clinical datasets. We aimed to investigate the effects of a wide range of CT acquisition and reconstruction parameters on radiomic features in a realistic setting using clinical low dose lung cancer screening cases. We performed univariable and multivariable explorations to consider the effects of individual parameters and the simultaneous interactions between three different acquisition/reconstruction parameters of radiation dose level, reconstructed slice thickness, and kernel.

Method

A cohort of 89 lung cancer screening patients were collected that each had a solid lung nodule >4mm diameter. A computational pipeline was used to perform a simulation of dose reduction of the raw projection data, collected from patient scans. This was followed by reconstruction of raw data with weighted filter back projection (wFBP) algorithm and automatic lung nodule detection and segmentation using a computer-aided detection tool. For each patient, 36 different image datasets were created corresponding to dose levels of 100%, 50%, 25%, and 10% of the original dose level, three slice thicknesses of 0.6 mm, 1 mm, and 2 mm, as well as three reconstruction kernels of smooth, medium, and sharp. For each nodule, 226 well-known radiomic features were calculated at each image condition. The reproducibility of radiomic features was first evaluated by measuring the intercondition agreement of the feature values among the 36 image conditions. Then in a series of univariable analyses, the impact of individual CT parameters was assessed by selecting subsets of conditions with one varying and two constant CT parameters. In each subset, intraparameter agreements were assessed. Overall concordance correlation coefficient (OCCC) served as the measure of agreement. An OCCC ≥ 0.9 implied strong agreement and reproducibility of radiomic features in intercondition or intraparameter comparisons. Furthermore, the interaction of CT parameters in impacting radiomic feature values was investigated via ANOVA.

Results

All included radiomic features lacked intercondition reproducibility (OCCC < 0.9) among all the 36 conditions. Out of 226 radiomic features analyzed, only 17 and 18 features were considered reproducible (OCCC ≥ 0.9) to dose and kernel variation, respectively, within the corresponding condition subsets. Slice thickness demonstrated the largest impact on radiomic feature values where only one to five features were reproducible at a few condition subsets. ANOVA revealed significant interactions (P < 0.05) between CT parameters affecting the variability of >50% of radiomic features.

Conclusion

We systematically explored the multidimensional space of CT parameters in affecting lung nodule radiomic features. Univariable and multivariable analyses of this study not only showed the lack of reproducibility of the majority of radiomic features but also revealed existing interactions among CT parameters, meaning that the effect of individual CT parameters on radiomic features can be conditional upon other CT acquisition and reconstruction parameters. Our findings advise on careful radiomic feature selection and attention to the inclusion criteria for CT image acquisition protocols within the datasets of radiomic studies.

Genicular artery embolization (GAE) is a novel therapy to treat patients with symptomatic knee osteoarthritis (OA) by reducing synovial arterial hypervascularity. This study evaluates the safety and efficacy of GAE for the treatment of symptomatic knee OA. A prospective, single-center, open-label U.S. Food and Drug Administration-approved investigational device exemption study was conducted. Patients enrolled in the study were 40 to 80 years old, with moderate or severe knee OA (Kellgren-Lawrence grade 2, 3, or 4), who previously had failure of conservative therapy. Baseline pain (visual analog scale [VAS]) and symptom scores (Western Ontario and McMaster Universities Osteoarthritis Index [WOMAC]) were assessed. After femoral arterial access was achieved, GAE of 1, 2, or 3 genicular arteries supplying the location of the subject's pain, as determined by digital subtraction angiography and cone-beam computed tomography, was performed using 100-μm particles. Adverse events and symptoms scores were assessed at 1 week, 1 month, 3 months, 6 months, and 1 year after GAE. Over a 10-month period, 40 subjects were enrolled. The median age was 69 years (range, 49 to 80 years). The median body mass index was 29 kg/m2 (range, 19 to 44 kg/m2). Knee OA severity was grade 2 in 18% of the patients, grade 3 in 43%, and grade 4 in 40%. Technical success was achieved in 100% of the subjects. Transient skin discoloration and transient mild knee pain after the procedure were common and expected. Treatment-related adverse events included a groin hematoma requiring overnight observation in 1 subject, self-resolving focal skin ulceration in 7 subjects, and an asymptomatic small bone infarct on magnetic resonance imaging at 3 months in 2 subjects. The WOMAC total and VAS pain scores decreased by 61% and 67% at 12 months from a median baseline of 52 (of 96) and 8 (of 10), respectively. Twenty-seven patients (68%) had a reduction of ≥50% in both WOMAC total and VAS pain scores. This prospective trial demonstrates that GAE is effective and durable in reducing pain symptoms from moderate or severe knee OA that is refractory to other conservative therapy, with an acceptable safety profile. Therapeutic Level IV. See Instructions for Authors for a complete description of levels of evidence.

We aimed to validate quantitative high-resolution computed tomography (HRCT) imaging analyses of interstitial lung disease (ILD) in rheumatoid arthritis (RA) patients, and to delineate a broad spectrum of annual longitudinal changes of ILD severity in the RA-ILD cohorts. Retrospective cohort 1 (n = 26) had matched PFT results and prospective cohort 2 (n = 34) were followed for over two years with baseline serum specimen. Automated quantitative analysis of HRCT was expressed as the extent of ground-glass opacity, lung fibrosis, honeycombing, and their summation-the total extent of quantitative ILD (QILD). Higher QILD score was associated with lower pulmonary function especially for DLCO% (ρ = -0.433, p = 0.027). Higher serum level of Krebs von den Lungen 6 were significantly associated with high QILD scores (ρ = 0.400, p = 0.026). Regarding QILD score changes in whole lung, even a single point increase was significantly associated with interval progression detected by the radiologist. Four distinct patterns (improvement, worsening, convex-like, and concave-like) during the 24 months were described by QILD scores. Prolonged disease duration of ILD at baseline was significantly associated with worsening of QILD scores. QILD has the potential to reliably evaluate the dynamic severity changes in patients with RA-ILD.

Objectives

The purpose of this study was to define clinically appropriate, computer-aided lung nodule detection (CAD) requirements and protocols based on recent screening trials. In the following paper, we describe a CAD evaluation methodology based on a publically available, annotated computed tomography (CT) image data set, and demonstrate the evaluation of a new CAD system with the functionality and performance required for adoption in clinical practice.

Methods

A new automated lung nodule detection and measurement system was developed that incorporates intensity thresholding, a Euclidean Distance Transformation, and segmentation based on watersheds. System performance was evaluated against the Lung Imaging Database Consortium (LIDC) CT reference data set.

Results

The test set comprised thin-section CT scans from 108 LIDC subjects. The median (±IQR) sensitivity per subject was 100 (±37.5) for nodules ≥ 4 mm and 100 (±8.33) for nodules ≥ 8 mm. The corresponding false positive rates were 0 (±2.0) and 0 (±1.0), respectively. The concordance correlation coefficient between the CAD nodule diameter and the LIDC reference was 0.91, and for volume it was 0.90.

Conclusions

The new CAD system shows high nodule sensitivity with a low false positive rate. Automated volume measurements have strong agreement with the reference standard. Thus, it provides comprehensive, clinically-usable lung nodule detection and assessment functionality.

Key points

• CAD requirements can be based on lung cancer screening trial results. • CAD systems can be evaluated using publically available annotated CT image databases. • A new CAD system was developed with a low false positive rate. • The CAD system has reliable measurement tools needed for clinical use.

Purpose

Currently, available Computed Tomography dose metrics are mostly based on fixed tube current Monte Carlo (MC) simulations and/or physical measurements such as the size specific dose estimate (SSDE). In addition to not being able to account for Tube Current Modulation (TCM), these dose metrics do not represent actual patient dose. The purpose of this study was to generate and evaluate a dose estimation model based on the Generalized Linear Model (GLM), which extends the ability to estimate organ dose from tube current modulated examinations by incorporating regional descriptors of patient size, scanner output, and other scan-specific variables as needed.

Methods

The collection of a total of 332 patient CT scans at four different institutions was approved by each institution's IRB and used to generate and test organ dose estimation models. The patient population consisted of pediatric and adult patients and included thoracic and abdomen/pelvis scans. The scans were performed on three different CT scanner systems. Manual segmentation of organs, depending on the examined anatomy, was performed on each patient's image series. In addition to the collected images, detailed TCM data were collected for all patients scanned on Siemens CT scanners, while for all GE and Toshiba patients, data representing z-axis-only TCM, extracted from the DICOM header of the images, were used for TCM simulations. A validated MC dosimetry package was used to perform detailed simulation of CT examinations on all 332 patient models to estimate dose to each segmented organ (lungs, breasts, liver, spleen, and kidneys), denoted as reference organ dose values. Approximately 60% of the data were used to train a dose estimation model, while the remaining 40% was used to evaluate performance. Two different methodologies were explored using GLM to generate a dose estimation model: (a) using the conventional exponential relationship between normalized organ dose and size with regional water equivalent diameter (WED) and regional CTDI_vol as variables and (b) using the same exponential relationship with the addition of categorical variables such as scanner model and organ to provide a more complete estimate of factors that may affect organ dose. Finally, estimates from generated models were compared to those obtained from SSDE and ImPACT.

Results

The Generalized Linear Model yielded organ dose estimates that were significantly closer to the MC reference organ dose values than were organ doses estimated via SSDE or ImPACT. Moreover, the GLM estimates were better than those of SSDE or ImPACT irrespective of whether or not categorical variables were used in the model. While the improvement associated with a categorical variable was substantial in estimating breast dose, the improvement was minor for other organs.

Conclusions

The GLM approach extends the current CT dose estimation methods by allowing the use of additional variables to more accurately estimate organ dose from TCM scans. Thus, this approach may be able to overcome the limitations of current CT dose metrics to provide more accurate estimates of patient dose, in particular, dose to organs with considerable variability across the population.