Curation of myeloma observational study MALIMAR using XNAT: solving the challenges posed by real-world data.

OBJECTIVES: MAchine Learning In MyelomA Response (MALIMAR) is an observational clinical study combining "real-world" and clinical trial data, both retrospective and prospective. Images were acquired on three MRI scanners over a 10-year window at two institutions, leading to a need for extensive curation. METHODS: Curation involved image aggregation, pseudonymisation, allocation between project phases, data cleaning, upload to an XNAT repository visible from multiple sites, annotation, incorporation of machine learning research outputs and quality assurance using programmatic methods. RESULTS: A total of 796 whole-body MR imaging sessions from 462 subjects were curated. A major change in scan protocol part way through the retrospective window meant that approximately 30% of available imaging sessions had properties that differed significantly from the remainder of the data. Issues were found with a vendor-supplied clinical algorithm for "composing" whole-body images from multiple imaging stations. Historic weaknesses in a digital video disk (DVD) research archive (already addressed by the mid-2010s) were highlighted by incomplete datasets, some of which could not be completely recovered. The final dataset contained 736 imaging sessions for 432 subjects. Software was written to clean and harmonise data. Implications for the subsequent machine learning activity are considered. CONCLUSIONS: MALIMAR exemplifies the vital role that curation plays in machine learning studies that use real-world data. A research repository such as XNAT facilitates day-to-day management, ensures robustness and consistency and enhances the value of the final dataset. The types of process described here will be vital for future large-scale multi-institutional and multi-national imaging projects. CRITICAL RELEVANCE STATEMENT: This article showcases innovative data curation methods using a state-of-the-art image repository platform; such tools will be vital for managing the large multi-institutional datasets required to train and validate generalisable ML algorithms and future foundation models in medical imaging. KEY POINTS: • Heterogeneous data in the MALIMAR study required the development of novel curation strategies. • Correction of multiple problems affecting the real-world data was successful, but implications for machine learning are still being evaluated. • Modern image repositories have rich application programming interfaces enabling data enrichment and programmatic QA, making them much more than simple "image marts".

Development and Evaluation of Machine Learning in Whole-Body Magnetic Resonance Imaging for Detecting Metastases in Patients With Lung or Colon Cancer: A Diagnostic Test Accuracy Study.

OBJECTIVES: Whole-body magnetic resonance imaging (WB-MRI) has been demonstrated to be efficient and cost-effective for cancer staging. The study aim was to develop a machine learning (ML) algorithm to improve radiologists' sensitivity and specificity for metastasis detection and reduce reading times. MATERIALS AND METHODS: A retrospective analysis of 438 prospectively collected WB-MRI scans from multicenter Streamline studies (February 2013-September 2016) was undertaken. Disease sites were manually labeled using Streamline reference standard. Whole-body MRI scans were randomly allocated to training and testing sets. A model for malignant lesion detection was developed based on convolutional neural networks and a 2-stage training strategy. The final algorithm generated lesion probability heat maps. Using a concurrent reader paradigm, 25 radiologists (18 experienced, 7 inexperienced in WB-/MRI) were randomly allocated WB-MRI scans with or without ML support to detect malignant lesions over 2 or 3 reading rounds. Reads were undertaken in the setting of a diagnostic radiology reading room between November 2019 and March 2020. Reading times were recorded by a scribe. Prespecified analysis included sensitivity, specificity, interobserver agreement, and reading time of radiology readers to detect metastases with or without ML support. Reader performance for detection of the primary tumor was also evaluated. RESULTS: Four hundred thirty-three evaluable WB-MRI scans were allocated to algorithm training (245) or radiology testing (50 patients with metastases, from primary 117 colon [n = 117] or lung [n = 71] cancer). Among a total 562 reads by experienced radiologists over 2 reading rounds, per-patient specificity was 86.2% (ML) and 87.7% (non-ML) (-1.5% difference; 95% confidence interval [CI], -6.4%, 3.5%; P = 0.39). Sensitivity was 66.0% (ML) and 70.0% (non-ML) (-4.0% difference; 95% CI, -13.5%, 5.5%; P = 0.344). Among 161 reads by inexperienced readers, per-patient specificity in both groups was 76.3% (0% difference; 95% CI, -15.0%, 15.0%; P = 0.613), with sensitivity of 73.3% (ML) and 60.0% (non-ML) (13.3% difference; 95% CI, -7.9%, 34.5%; P = 0.313). Per-site specificity was high (>90%) for all metastatic sites and experience levels. There was high sensitivity for the detection of primary tumors (lung cancer detection rate of 98.6% with and without ML [0.0% difference; 95% CI, -2.0%, 2.0%; P = 1.00], colon cancer detection rate of 89.0% with and 90.6% without ML [-1.7% difference; 95% CI, -5.6%, 2.2%; P = 0.65]). When combining all reads from rounds 1 and 2, reading times fell by 6.2% (95% CI, -22.8%, 10.0%) when using ML. Round 2 read-times fell by 32% (95% CI, 20.8%, 42.8%) compared with round 1. Within round 2, there was a significant decrease in read-time when using ML support, estimated as 286 seconds (or 11%) quicker ( P = 0.0281), using regression analysis to account for reader experience, read round, and tumor type. Interobserver variance suggests moderate agreement, Cohen κ = 0.64; 95% CI, 0.47, 0.81 (with ML), and Cohen κ = 0.66; 95% CI, 0.47, 0.81 (without ML). CONCLUSIONS: There was no evidence of a significant difference in per-patient sensitivity and specificity for detecting metastases or the primary tumor using concurrent ML compared with standard WB-MRI. Radiology read-times with or without ML support fell for round 2 reads compared with round 1, suggesting that readers familiarized themselves with the study reading method. During the second reading round, there was a significant reduction in reading time when using ML support.

Repeatability and test-retest reproducibility of mean apparent diffusion coefficient measurements of focal and diffuse disease in relapsed multiple myeloma at 3T whole body diffusion-weighted MRI (WB-DW-MRI).

OBJECTIVE: To assess the test-retest reproducibility and intra/interobserver agreement of apparent diffusion coefficient (ADC) measurements of myeloma lesions using whole body diffusion-weighted MRI (WB-DW-MRI) at 3T MRI. METHODS: Following ethical approval, 11 consenting patients with relapsed multiple myeloma were prospectively recruited and underwent baseline WB-DW-MRI. For a single bed position, axial DWI was repeated after a short interval to permit test-retest measurements.Mean ADC measurement was performed by two experienced observers. Intra- and interobserver agreement and test-retest reproducibility were assessed, using coefficient of variation (CV) and interclass correlation coefficient (ICC) measures, for diffuse and focal lesions (small ≤10 mm and large >10 mm). RESULTS: 47 sites of disease were outlined (23 focal, 24 diffuse) in different bed positions (pelvis = 22, thorax = 20, head and neck = 5). For all lesions, there was excellent intraobserver agreement with ICC of 0.99 (0.98-0.99) and COV of 5%. For interobserver agreement, ICC was 0.89 (0.8-0.934) and COV was 17%. There was poor interobserver agreement for diffuse disease (ICC = 0.46) and small lesions (ICC = 0.54).For test-retest reproducibility, excellent ICC (0.916) and COV (14.5%) values for mean ADC measurements were observed. ICCs of test-retest were similar between focal lesions (0.83) and diffuse infiltration (0.80), while ICCs were higher in pelvic (0.95) compared to thoracic (0.81) region and in small (0.96) compared to large (0.8) lesions. CONCLUSION: ADC measurements of focal lesions in multiple myeloma are repeatable and reproducible, while there is more variation in ADC measurements of diffuse disease in patients with multiple myeloma. ADVANCES IN KNOWLEDGE: Mean ADC measurements are repeatable and reproducible in focal lesions in multiple myeloma, while the ADC measurements of diffuse disease in multiple myeloma are more subject to variation. The evidence supports the future potential role of ADC measurements as predictive quantitative biomarker in multiple myeloma.

Incidental uptake of fluorodeoxyglucose in the Waldeyer's ring and risk of oropharyngeal malignancy.

PURPOSE: Fluorodeoxyglucose (FDG) positron emission tomography (PET) is increasingly used to diagnose and stage malignancy. The aim of this article is to investigate the significance of incidental FDG uptake in the Waldeyer's ring and to assess its value in predicting clinically occult oropharyngeal malignancy. METHODS: All FDG-PET/CT scans performed in Imperial College NHS Foundation Trust, UK between January 2012 and November 2018 were included. Patients with known or suspected oropharyngeal malignancy or lymphoma were excluded. Minimum follow-up was 12 months. RESULTS: A total of 724 scans revealed oropharyngeal uptake of FDG. Of these, 102 were included in the study. Most patients (62.1%) were scanned as part of staging for other malignancies. Oropharyngeal FDG uptake was asymmetrical in 57.3% of the cases. Uptake was more common in the tonsils (56.3%), followed by the tongue base (31.1%) and both sites (12.6%). In 41.7% of reports, appearance was described as likely physiological; however, 52.4% of reports advised direct visualisation, clinical correlation or ENT opinion. Only 24.3% (25/102) of patients were referred and seen by ENT, 14.6% (15/102) of which had an interval PET scan and 8.7% (9/102) proceeded to tissue diagnosis. There was one oropharyngeal cancer identified and one unexpected metastasis from esophageal cancer. CONCLUSION: Incidental uptake on PET/CT in the oropharynx is common. However, malignancy is rare (1.9%) and, when present, is associated with high SUVmax and asymmetrical uptake. Imaging results must be correlated clinically. These patients should be seen by an ENT specialist yet most may not require further investigations.

Pictorial review of whole body MRI in myeloma: emphasis on diffusion-weighted imaging.

There have been major advances in myeloma imaging over the past few years with focal lesions on imaging now forming part of the disease defining criteria. Whole body diffusion-weighted MRI (WB-MRI) is considered the most sensitive technique for the detection of focal active lesions. This pictorial review will focus on imaging the spectrum of myelomatous disorders on WB-MRI including diffusion and Dixon sequences. The typical imaging patterns of disease are demonstrated including in the contexts of staging, presumed solitary plasmacytoma, smouldering myeloma and examples of paramedullary and extramedullary disease. The utility of diffusion-weighted imaging in response assessment is a major advantage and this will be exemplified here.

Immunotherapy-related adverse effects on 18F-FDG PET/CT imaging.

18F-Fluorodeoxyglucose positron emission tomography/CT imaging plays a key role in oncological imaging including in staging, radiotherapy planning, treatment response and recurrence assessment. Immunotherapies represent a major advance in cancer therapy for a number of tumours with resulting survival benefit. However, a wide range of immune related adverse events (irAEs), some of which can be apparent on imaging, have been reported. These involve many organ systems but particularly endocrine, cutaneous and gastrointestinal systems. Early detection of irAEs is essential to aid diagnosis and management of patients and to reduce associated morbidity. In addition, it is important to not mistake treatment related effects for disease.This pictorial review aims to identify common irAEs and changes seen on 18F-fluorodeoxyglucose positron emission tomography/CT.

Interobserver agreement of the visual Herder scale for the assessment of solitary pulmonary nodules on 18F Fluorodeoxyglucose PET/computed tomography.

BACKGROUND: British Thoracic Society guidelines recommend 18F Fluorodeoxyglucose (F-FDG) PET/computed tomography (CT) for solitary pulmonary nodule (SPN) follow-up in high-risk individuals or if the CT Brock score is >10%. Nodule tracer uptake is assessed visually in comparison to the surrounding lung tissue and mediastinal blood pool (Herder score). This score is used to calculate the risk of malignancy and guide patient management. Despite its widespread use, there have been no studies to date looking at interobserver agreement using the Herder scale. PATIENTS AND METHODS: F-FDG PET/CT studies of 100 consecutive patients imaged for the evaluation of SPN were retrospectively analysed. Anonymized images were reviewed independently by three Consultant Nuclear Medicine Radiologists and the Herder score was documented, along with a confidence score graded 1-3, where 1 indicated 'not at all confident' and 3 indicated 'very confident'. Interobserver agreement was assessed using interclass correlation coefficient modelling. RESULTS: There was complete reviewer agreement in 81% cases, and interclass correlation with Cronbach's alpha was excellent at 0.973 (95% confidence interval, 0.962-0.981). The agreement between pairs of reviewers was good and confidence scores using the Herder scale were high, with reviewers giving a confidence score of 3 in an average of 78% of cases. CONCLUSION: Our study suggests excellent interobserver agreement for use of the Herder scale in evaluating SPNs. Reviewer confidence scores were high reflecting high confidence in the use of the Herder scale for evaluating SPN.

Clinical 18F-FDG and amyloid brain positron emission tomography/CT in the investigation of cognitive impairment: where are we now?

The number of people living with dementia is increasing, but as yet there remains no cure or disease-modifying treatment. This review aims to help readers understand the role of 18F-FDG PET/CT imaging in the investigation of cognitive impairment and how the advent of amyloid PET/CT imaging may hold the key to radically changing management of the most common form of dementia - Alzheimer's disease. The indications for 18F-FDG PET/CT and amyloid PET/CT imaging in cognitive impairment are outlined. Additionally, the mechanisms of action, technique, patient preparation and acquisition parameters for both are detailed. We conclude by providing a framework for interpreting 18F-FDG PET/CT and amyloid PET/CT imaging in the more common conditions that lead to cognitive impairment conditions with tips on avoiding pitfalls in interpretation.

Imaging in myeloma with focus on advanced imaging techniques.

In recent years, there have been major advances in the imaging of myeloma with whole body MRI incorporating diffusion-weighted imaging, emerging as the most sensitive modality. Imaging is now a key component in the work-up of patients with a suspected diagnosis of myeloma. The International Myeloma Working Group now specifies that more than one focal lesion on MRI or lytic lesion on whole body low-dose CT or fludeoxyglucose (FDG) PET/CT fulfil the criteria for bone damage requiring therapy. The recent National Institute for Health and Care Excellence myeloma guidelines recommend imaging in all patients with suspected myeloma. In addition, there is emerging data supporting the use of functional imaging techniques (WB-DW MRI and FDG PET/CT) to predict outcome and evaluate response to therapy. This review summarises the imaging modalities used in myeloma, the latest guidelines relevant to imaging and future directions.

Discovery of pre-therapy 2-deoxy-2-18F-fluoro-D-glucose positron emission tomography-based radiomics classifiers of survival outcome in non-small-cell lung cancer patients.

PURPOSE: The aim of this multi-center study was to discover and validate radiomics classifiers as image-derived biomarkers for risk stratification of non-small-cell lung cancer (NSCLC). PATIENTS AND METHODS: Pre-therapy PET scans from a total of 358 Stage I-III NSCLC patients scheduled for radiotherapy/chemo-radiotherapy acquired between October 2008 and December 2013 were included in this seven-institution study. A semi-automatic threshold method was used to segment the primary tumors. Radiomics predictive classifiers were derived from a training set of 133 scans using TexLAB v2. Least absolute shrinkage and selection operator (LASSO) regression analysis was used for data dimension reduction and radiomics feature vector (FV) discovery. Multivariable analysis was performed to establish the relationship between FV, stage and overall survival (OS). Performance of the optimal FV was tested in an independent validation set of 204 patients, and a further independent set of 21 (TESTI) patients. RESULTS: Of 358 patients, 249 died within the follow-up period [median 22 (range 0-85) months]. From each primary tumor, 665 three-dimensional radiomics features from each of seven gray levels were extracted. The most predictive feature vector discovered (FVX) was independent of known prognostic factors, such as stage and tumor volume, and of interest to multi-center studies, invariant to the type of PET/CT manufacturer. Using the median cut-off, FVX predicted a 14-month survival difference in the validation cohort (N = 204, p = 0.00465; HR = 1.61, 95% CI 1.16-2.24). In the TESTI cohort, a smaller cohort that presented with unusually poor survival of stage I cancers, FVX correctly indicated a lack of survival difference (N = 21, p = 0.501). In contrast to the radiomics classifier, clinically routine PET variables including SUVmax, SUVmean and SUVpeak lacked any prognostic information. CONCLUSION: PET-based radiomics classifiers derived from routine pre-treatment imaging possess intrinsic prognostic information for risk stratification of NSCLC patients to radiotherapy/chemo-radiotherapy.

What's New in Imaging for Gynecologic Cancer?

Magnetic resonance imaging (MRI) is the optimal modality for local staging of gynecological tumors. Advances in functional MRI with diffusion-weighted and dynamic contrast-enhanced sequences provide more detailed information regarding tumor cellularity, vascularity, and viability. Fluorodeoxyglucose positron emission tomography/computed tomography (FDG PET/CT) now has an established role in imaging for gynecological cancers, particularly staging of locally advanced cervical cancers and pre-salvage exenterative therapy in relapsed gynecologic tumors. Novel PET tracers, targeting other aspects of tumor biology, are being evaluated although none are currently in routine clinical use. New PET/MR scanners have the potential to combine the strengths of both modalities in one sitting. This review covers advances in gynecologic imaging concentrating on cervical, endometrial, and ovarian cancers.

Clinical PET Imaging in Prostate Cancer.

Prostate cancer is the second most common cancer in men worldwide, with a wide spectrum of biologic behavior ranging from indolent low-risk disease to highly aggressive castration-resistant prostate cancer. Conventional imaging with computed tomography, magnetic resonance imaging, and bone scintigraphy is limited for the detection of nodal disease and distant bone metastases. In addition, advances in the available therapeutic options, both localized and systemic, drive the requirement for precise diagnostic and prognostic tools to refine the individual therapeutic approach at various times in the management of patients with prostate cancer. Positron emission tomography (PET) has a rapidly evolving role in the assessment of prostate cancer, particularly in the scenario of biochemical relapse. Fluorine 18 (18F) fluorodeoxyglucose, the most widely available PET tracer, has limitations, particularly in indolent prostate cancer. In the past decade, several PET tracers with specific molecular targets have reached the clinical domain. These tracers include 18F-sodium fluoride, which is a bone-specific biomarker of osteoblastic activity; 18F-choline and carbon 11-choline, which are directed at cell membrane metabolism; gallium 68-prostate-specific membrane antigen ligands; and, more recently, an amino acid analog, 18F-fluciclovine (anti-1-amino-3-18F-fluorocyclobutane-1-carboxylic acid; also known as FACBC), which is also directed at cell membrane turnover. The mechanisms of actions of the clinically available PET tracers are reviewed, as well as their role in the imaging of prostate cancer with reference to relevant guidelines and the technical and imaging pearls and pitfalls of these tracers. ©RSNA, 2017.