Artificial intelligence and machine learning in cancer imaging.

An increasing array of tools is being developed using artificial intelligence (AI) and machine learning (ML) for cancer imaging. The development of an optimal tool requires multidisciplinary engagement to ensure that the appropriate use case is met, as well as to undertake robust development and testing prior to its adoption into healthcare systems. This multidisciplinary review highlights key developments in the field. We discuss the challenges and opportunities of AI and ML in cancer imaging; considerations for the development of algorithms into tools that can be widely used and disseminated; and the development of the ecosystem needed to promote growth of AI and ML in cancer imaging.

Diagnostic accuracy of whole-body MRI versus standard imaging pathways for metastatic disease in newly diagnosed non-small-cell lung cancer: the prospective Streamline L trial.

BACKGROUND: Whole-body magnetic resonance imaging (WB-MRI) could be an alternative to multi-modality staging of non-small-cell lung cancer (NSCLC), but its diagnostic accuracy, effect on staging times, number of tests needed, cost, and effect on treatment decisions are unknown. We aimed to prospectively compare the diagnostic accuracy and efficiency of WB-MRI-based staging pathways with standard pathways in NSCLC. METHODS: The Streamline L trial was a prospective, multicentre trial done in 16 hospitals in England. Eligible patients were 18 years or older, with newly diagnosed NSCLC that was potentially radically treatable on diagnostic chest CT (defined as stage IIIb or less). Exclusion criteria were severe systemic disease, pregnancy, contraindications to MRI, or histologies other than NSCLC. Patients underwent WB-MRI, the result of which was withheld until standard staging investigations were complete and the first treatment decision made. The multidisciplinary team recorded its treatment decision based on standard investigations, then on the WB-MRI staging pathway (WB-MRI plus additional tests generated), and finally on all tests. The primary outcome was difference in per-patient sensitivity for metastases between standard and WB-MRI staging pathways against a consensus reference standard at 12 months, in the per-protocol population. Secondary outcomes were difference in per-patient specificity for metastatic disease detection between standard and WB-MRI staging pathways, differences in treatment decisions, staging efficiency (time taken, test number, and costs) and per-organ sensitivity and specificity for metastases and per-patient agreement for local T and N stage. This trial is registered with the International Standard Randomised Controlled Trial registry, number ISRCTN50436483, and is complete. FINDINGS: Between Feb 26, 2013, and Sept 5, 2016, 976 patients were screened for eligibility. 353 patients were recruited, 187 of whom completed the trial; 52 (28%) had metastasis at baseline. Pathway sensitivity was 50% (95% CI 37-63) for WB-MRI and 54% (41-67) for standard pathways, a difference of 4% (-7 to 15, p=0·73). No adverse events related to imaging were reported. Specificity did not differ between WB-MRI (93% [88-96]) and standard pathways (95% [91-98], p=0·45). Agreement with the multidisciplinary team's final treatment decision was 98% for WB-MRI and 99% for the standard pathway. Time to complete staging was shorter for WB-MRI (13 days [12-14]) than for the standard pathway (19 days [17-21]); a 6-day (4-8) difference. The number of tests required was similar WB-MRI (one [1-1]) and standard pathways (one [1-2]). Mean per-patient costs were £317 (273-361) for WBI-MRI and £620 (574-666) for standard pathways. INTERPRETATION: WB-MRI staging pathways have similar accuracy to standard pathways, and reduce the staging time and costs. FUNDING: UK National Institute for Health Research.

Duty of candour: implications for radiologists.

This article addresses duty of candour in relation to radiologists. The legislation underpinning duty of candour, definition of terms and the requirements for its application are explained. Consideration is given as to the differences between duty of candour and openness of discussion. The issue of how duty of candour should apply to radiologists (both diagnostic and interventional) is then addressed.

Complications of pulmonary hypertension: a pictorial review.

Pulmonary hypertension (PH) is a rare disease with a significant morbidity and mortality if untreated. The disease has a multifactorial aetiology and is often associated with insidious onset of signs and symptoms. Multimodality imaging is often required for establishing the diagnosis, evaluating the underlying haemodynamic compromise and follow-up after institution of therapy. The range of potential complications associated with PH vary widely. We aimed to summarize the imaging findings of complications that the radiologist should be familiar with.

Changes in renal function in elderly patients following intravenous iodinated contrast administration: a retrospective study.

Background. Contrast-induced nephropathy (CIN) is a recognised complication of intravascular administration of iodinated contrast media (ICM). Previous studies suggest a higher incidence in the elderly, but no large study has assessed this to date. We set out to assess changes in creatinine in elderly inpatients following computed tomography (CT) examination and compare those who received intravenous contrast to those who did not. Methods. Using the Radiology Information System in two teaching hospitals, inpatients over the age of seventy who had a CT examination and a baseline creatinine were identified and their follow-up creatinine levels were analysed. Elderly inpatients who underwent a noncontrast CT over the same period were used as controls. Results. 677 elderly inpatients who received ICM were compared with 487 controls. 9.2% of patients who received ICM developed acute kidney injury (AKI) compared to 3.5% of inpatient controls (P < 0.0001). Patients with higher baseline eGFR had a higher incidence of post-CT AKI. Conclusions. The incidence of post-CT AKI is higher in patients who received IV ICM compared to those who did not; the difference may be partly attributable to contrast-induced nephropathy. This suggests that the incidence of CIN in the elderly may not be as high as previously thought.

Interstitial lung disease and ANCA-associated vasculitis: a retrospective observational cohort study.

OBJECTIVES: ANCA-associated vasculitis and interstitial lung disease (ILD) are uncommon conditions. The occurrence of both diseases in the same patient is increasingly recognized. Our aim was to ascertain the characteristics and outcomes of patients with ILD and ANCA-associated vasculitis. METHODS: A retrospective observational cohort study was performed. Patients who presented to the Hammersmith Hospital, London, with ANCA-associated vasculitis [granulomatosis with polyangiitis (Wegener's), microscopic polyangiitis (MPA) or Churg-Strauss syndrome] who also had ILD were included. Following hospital discharge, all patients were followed up in a multi-disciplinary vasculitis clinic. We recorded patient demographics, diagnostic tests, treatment, complications and mortality. RESULTS: ILD was observed in 2.7% (n = 14) of our patients with ANCA-associated vasculitis (n = 510); all had MPO-ANCA and a clinical diagnosis of MPA, giving a prevalence of 7.2% in patients with MPA (n = 194). There was no significant difference in survival between patients with MPA and ILD and those with MPA alone. CONCLUSION: It is important that physicians are aware of this clinical association and the presence of ILD should be considered in all patients with ANCA-associated vasculitis, especially those with MPO-ANCA. The possibility that patients with ILD may subsequently develop features of systemic vasculitis should also be remembered.

Abnormal myocardial insulin signalling in type 2 diabetes and left-ventricular dysfunction.

AIMS: Whole body and myocardial insulin resistance are features of non-insulin-dependent diabetes mellitus (NIDDM) and left-ventricular dysfunction (LVD). We determined whether abnormalities of insulin receptor substrate-1 (IRS1), IRS1-associated PI3K (IRS1-PI3K), and glucose transporter 4 (GLUT4) contribute to tissue-specific insulin resistance. METHODS AND RESULTS: We collected skeletal muscle (n = 27) and myocardial biopsies (n = 24) from control patients (n = 7), patients with NIDDM (n = 9) and patients with LVD (n = 8), who were characterized by euglycaemic-hyperinsulinaemic clamp and positron emission tomography. Comparative studies were carried out in three mouse models. We demonstrate an unrecognized reduction of IRS1 in skeletal muscle of LVD patients and an unexpected increase in cardiac IRS1-PI3K activity in NIDDM and LVD patients. In NIDDM, there was a concomitant reduction in sarcolemmal GLUT4, whereas in patients with LVD sarcolemmal GLUT4 was increased. We confirm activation of IRS1-PI3K and reduction in sarcolemmal GLUT4 in the insulin resistant ob/ob mouse heart where we also demonstrate perturbation of GLUT4 docking and fusion. A direct relationship between PI3K and GLUT4 was demonstrated in mice expressing activated PI3K in the heart and increased GLUT4 at the sarcolemma was confirmed in a mouse model of LVD. CONCLUSION: Our data show that the mechanisms of myocardial insulin resistance are different between NIDDM and LVD.

Iodine-123 uptake in vertebral haemangiomas in a patient with papillary thyroid carcinoma.

We present a case of a 58-year-old woman with papillary carcinoma of the thyroid and elevated thyroglobulin. Whole body 123I scan with SPECT images demonstrated focal uptake in the thoracic spine, reported as bone metastases. Subsequent (18)FDG PET and (99m)Tc HDP bone were normal. MRI and CT scans confirmed the presence of vertebral haemangiomas corresponding to the uptake seen on the (123)I scan. False-positive uptake of (123)I in benign vertebral haemangiomas should be considered in the differential diagnosis of focal vertebral uptake.

Pulmonary emboli and extramedullary haematopoiesis in beta-thalassaemia intermedia.

We present a case of a 52-year-old woman with beta-thalassaemia intermedia under investigation for pulmonary hypertension. V/Q scan demonstrated bilateral pulmonary emboli. In addition there was evidence of central, posterior matched defects. CT pulmonary angiogram confirmed the presence of extramedullary haematopoiesis (EMH) in the paravertebral region. Although a few case reports of pulmonary emboli in thalassaemia intermedia have been published, to the best of our knowledge there have been no reports of matched defects caused by extramedullary haematopoiesis. This may be mistaken for enlarged pulmonary arteries, which are also seen in patients with pulmonary hypertension.

A comparison of conventional film, CR hard copy and PACS soft copy images of the chest: analyses of ROC curves and inter-observer agreement.

STUDY OBJECTIVE: The aim of this study was to determine whether the accuracy of diagnosis of a spectrum of chest pathology was affected by the imaging technique used, and to compare conventional film/screen, hard copy computed (phosphor plate) radiography (CR) and soft copy CR (PACS) images. MATERIALS AND METHODS: For each of 44 patients who had a CT examination of the thorax, PA and lateral chest radiographs were produced using conventional film, hard copy CR and soft copy PACS images. Five radiologists independently scored all images for the presence of abnormalities. The data were analysed in two stages using the result of the CT scan as the reference standard diagnosis: firstly, to investigate differences in abnormality scores between image modalities and observers using ROC analysis; secondly, to investigate the agreement of the diagnoses with the reference standard by the analysis of kappa scores. RESULTS: The ROC analyses and comparison of kappa scores showed no differences between image modalities (P=0.72, P=0.87), but highly significant differences between observers (P<0.001, P=0.003). CONCLUSION: The detection of chest lesions did not vary between conventional film, CR hard copy and PACS soft copy images. For all three image types, there were statistically significant differences between observers.