IMG-11. A COMPUTERISED CLINICAL DECISION SUPPORT SYSTEM FOR DIAGNOSING CHILDREN’S BRAIN TUMOURS USING FUNCTIONAL IMAGING AND MACHINE LEARNING

INTRODUCTION: Magnetic resonance imaging is a key investigation in the diagnosis of childhood solid tumours. Advanced techniques such as diffusion weighted imaging (DWI), magnetic resonance spectroscopy (MRS) and perfusion imaging probe the underlying cellular, chemical and vascular nature of the disease. Coupled with machine learning these scanning methods show improvement in diagnostic accuracy compared with conventional imaging. Advanced image analysis is not routinely available in hospitals. We present a clinical decision support system (CDSS) developed for advanced MR analysis and interpretation. METHOD: The CDSS was developed in house. The Children’s Cancer and Leukaemia Group Functional Imaging Group (CCLGFIG) Database, a national resource, was used to provide a repository of cases together with their advanced imaging and machine learning diagnostic classifiers. A new case is displayed alongside cases in the repository with known diagnoses, including summary statistics for relevant diagnostic categories. The CDSS was made available to radiologists, in their clinical environment for technical and clinical evaluation. Structured interviews were undertaken. The CDSS was developed as a computer app for multi-centre distribution. RESULTS: 436 MRS, 240 DWI and 85 perfusion cases were available for building repositories. Machine learning classifiers showed diagnostic accuracies for the major childhood brain tumour types of 85-95%. Comparison of MRS with a data repository was found to improve non-invasive diagnosis. Results from the CDSS can be uploaded to the CCLGFIG to support multicentre research. Positive feedback on the CDSS from clinicians included: ready access to advanced analysis; simple and efficient integration into clinical workflow; and assisted interpretation of advanced analysis. DISCUSSION: Advanced MR analysis techniques provide improved non-invasive diagnostic accuracy but are difficult to implement on clinical systems due to technical, infrastructure and training limitations. CONCLUSION: We have successfully released a CDSS for paediatric cancer within the hospital environment and assessed its suitability for clinical use.

Automated Modular Magnetic Resonance Imaging Clinical Decision Support System (MIROR): An Application in Pediatric Cancer Diagnosis

BACKGROUND

Advances in magnetic resonance imaging and the introduction of clinical decision support systems has underlined the need for an analysis tool to extract and analyze relevant information from magnetic resonance imaging data to aid decision making, prevent errors, and enhance health care.

OBJECTIVE

The aim of this study was to design and develop a modular medical image region of interest analysis tool and repository (MIROR) for automatic processing, classification, evaluation, and representation of advanced magnetic resonance imaging data.

METHODS

The clinical decision support system was developed and evaluated for diffusion-weighted imaging of body tumors in children (cohort of 48 children, with 37 malignant and 11 benign tumors). Mevislab software and Python have been used for the development of MIROR. Regions of interests were drawn around benign and malignant body tumors on different diffusion parametric maps, and extracted information was used to discriminate the malignant tumors from benign tumors.

RESULTS

Using MIROR, the various histogram parameters derived for each tumor case when compared with the information in the repository provided additional information for tumor characterization and facilitated the discrimination between benign and malignant tumors. Clinical decision support system cross-validation showed high sensitivity and specificity in discriminating between these tumor groups using histogram parameters.

CONCLUSIONS

MIROR, as a diagnostic tool and repository, allowed the interpretation and analysis of magnetic resonance imaging images to be more accessible and comprehensive for clinicians. It aims to increase clinicians' skillset by introducing newer techniques and up-to-date findings to their repertoire and make information from previous cases available to aid decision making. The modular-based format of the tool allows integration of analyses that are not readily available clinically and streamlines the future developments.

Evaluation of the added value of 1H-magnetic resonance spectroscopy for the diagnosis of pediatric brain lesions in clinical practice

Background

Magnetic resonance spectroscopy (MRS) aids noninvasive diagnosis of pediatric brain tumors, but use in clinical practice is not well documented. We aimed to review clinical use of MRS, establish added value in noninvasive diagnosis, and investigate potential impact on patient care.

Methods

Sixty-nine children with lesions imaged using MRS and reviewed by the tumor board from 2014 to 2016 met inclusion criteria. Contemporaneous MRI diagnosis, spectroscopy analysis, histopathology, and clinical information were reviewed. Final diagnosis was agreed on by the tumor board at study end.

Results

Five cases were excluded for lack of documented MRI diagnosis. The principal MRI diagnosis by pediatric radiologists was correct in 59%, increasing to 73% with addition of MRS. Of the 73%, 19.1% (95% CI, 9.1%-33.3%) were incorrectly diagnosed with MRI alone. MRS led to a significant improvement in correct diagnosis over all tumor types (P = .012). Of diagnoses correctly made with MRI, confidence increased by 37% when adding MRS, with no patients incorrectly re-diagnosed. Indolent lesions were diagnosed noninvasively in 85% of cases, with MRS a major contributor to 91% of these diagnoses. Of all patients, 39% were managed without histopathological diagnosis. MRS contributed to diagnosis in 68% of this group, modifying it in 12%. MRS influenced management in 33% of cases, mainly through avoiding and guiding biopsy and aiding tumor characterization.

Conclusion

MRS can improve accuracy and confidence in noninvasive diagnosis of pediatric brain lesions in clinical practice. There is potential to improve outcomes through avoiding biopsy of indolent lesions, aiding tumor characterization, and facilitating earlier family discussions and treatment planning.

Prediction of severe bacterial infection in children with an emergency department diagnosis of infection

OBJECTIVE

To determine the relationship between near-patient-test (NPT) lactate, white blood cell count (WBC) and C-reactive protein (CRP) and severe bacterial infection (SBI) in children presenting to the emergency department (ED) with infection.

METHODS

An observational cohort study was undertaken in a paediatric emergency department of a large urban teaching hospital. Data were collected from January 2007 until December 2007. Inclusion criteria were age <16 years, blood test including NPT lactate obtained in the ED and infection-related ED diagnosis. Patients were pre-assigned to risk groups according to their NPT lactate, WBC and CRP.

RESULTS

506 children were included in the study, of which 42 (8.3%) had SBI. NPT lactate, WBC and CRP were significantly higher in the SBI cohort. High-risk NPT lactate (≥4 mmol/l) had a sensitivity of 38.1% (95% CI 23.6% to 54.4%) and a specificity of 89.7% (95% CI 86.5% to 92.3%); high-risk WBC (<5 or ≥15×10(9)/l) had a sensitivity of 51.2% (95% CI 35.1% to 67.1%) and a specificity of 73.8% (95% CI 69.4% to 77.8%); and high-risk CRP (≥50 mg/l) had a sensitivity of 36.8% (95% CI 21.8% to 54.1%) and a specificity of 83.6% (95% CI 79.4% to 87.2%) for SBI. All three high-risk markers combined yielded a sensitivity of 5.3% (95% CI 1.5% to 17.3%) and a specificity of 99.2% (95% CI 97.6% to 99.7%) for SBI.

CONCLUSION

The data from our study suggest that NPT lactate provides early diagnostic information about the risk of SBI in children presenting to the ED with a suspected infection. Combining NPT lactate with WBC and CRP resulted in a promising rule-in-tool for SBI in children in the ED which, with prospective validation, has the potential to aid early identification of SBI in children.

Fractures and recurrent fractures in children; varying effects of environmental factors as well as bone size and mass

BACKGROUND

Fractures are frequent in childhood and cause considerable morbidity. Previous reports have indicated a variety of potential contributors to fracture risk including low bone mineral content and density, milk avoidance, lack of exercise, asthma, obesity, and a high consumption of carbonated beverages.

AIMS

We wished to test the hypothesis that children who sustain recurrent fractures have a lower bone mass and a higher prevalence of underlying risk factors for fracture than those who fracture once or not at all.

METHODS

We studied 150 children aged 4-16 years: 50 who had suffered recurrent fractures, 50 who had fractured for the first time, and 50 fracture-free controls. Subjects underwent assessment of bone size and mass by total body (TB) and lumbar spine (L2-4) dual energy X-ray absorptiometry (DXA). Values were adjusted for body size, based on the control group measurements as unadjusted DXA values are substantially influenced by size in children. Anthropometry and grip dynamometry were carried out, and information about factors implicated in fracture aetiology such as milk intake, physical activity levels, asthma prevalence and carbonated beverage consumption were recorded using questionnaires.

RESULTS

Children who had sustained one or more fractures had a significantly lower BMC and aBMD at all sites than controls after conversion to size adjusted z scores (L2-4 BMC P = 0.0002; L2-4 aBMD P < 0.0001; TB BMC P < 0.0001; TB aBMD P < 0.0001); estimates for TB excluded fracture sites. There was, however, no difference in adjusted bone mass between children with one and those with recurrent fractures. Children with recurrent fractures had a significantly lower milk intake, lower levels of physical activity, a higher BMI, and a higher consumption of carbonated beverages than controls. The prevalence of risk factors was not, however, significantly higher than controls in children with a single fracture.

CONCLUSIONS

Children with fractures have a lower bone mass for body size than children without fractures. Modifiable risk factors such as diet and exercise increase the risk of recurrent fractures.

The serpin-enzyme complex receptor recognizes soluble, nontoxic amyloid-beta peptide but not aggregated, cytotoxic amyloid-beta peptide

There is now extensive evidence that amyloid-beta peptide is toxic to neurons and that its cytotoxic effects can be attributed to a domain corresponding to amyloid-beta 25-35, GSNKGAIIGLM. We have shown recently that the serine proteinase inhibitor (serpin)-enzyme complex receptor (SEC-R), a receptor initially identified for binding of alpha1-antitrypsin (alpha1-AT) and other serine protease inhibitors, also recognizes the amyloid-beta 25-35 domain. In fact, by recognizing the amyloid-beta 25-35 domain, SEC-R mediates cell surface binding, internalization, and degradation of soluble amyloid-beta peptide. In this study, we examined the possibility that SEC-R mediates the neurotoxic effect of amyloid-beta peptide. A series of peptides based on the sequences of amyloid-beta peptide and alpha1-AT was prepared soluble in dimethyl sulfoxide or insoluble in water and examined in assays for SEC-R binding, for cytotoxicity in neuronal PC12 cells and murine cortical neurons in primary culture, and for aggregation in sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) analysis. The results show that amyloid-beta peptide 25-35 and amyloid-beta peptide 1-40 prepared soluble in dimethyl sulfoxide compete for binding to SEC-R, are nontoxic, and migrate as monomers in SDS-PAGE analysis. In contrast, the same peptides aged in water did not compete for binding to SEC-R but were toxic and migrated as aggregates in SDS-PAGE. An all-D-amyloid-beta 25-35 peptide was not recognized at all by SEC-R but retained full toxic/aggregating properties. Using a series of deleted, substituted, and chimeric ambeta/alpha1-AT peptides, toxicity correlated well with aggregation but poorly with SEC-R recognition. In a subclone of PC12 cells which developed resistance to the toxic effect of aggregated amyloid-beta 25-35 there was a 2.5-3-fold increase in the number of SEC-R molecules/cell compared with the parent PC12 cell line. These data show that SEC-R does not mediate the cytotoxic effect of aggregated amyloid-beta peptide. Rather, SEC-R could play a protective role by mediating clearance and catabolism of soluble, monomeric amyloid-beta peptide, if soluble amyloid-beta peptide proves to be an in vivo precursor of the insoluble, toxic peptide.

Specificity in recognition of amyloid-beta peptide by the serpin-enzyme complex receptor in hepatoma cells and neuronal cells

The serpin-enzyme complex (SEC) receptor was originally identified using a synthetic peptide (peptide 105Y) based on the sequence of a candidate receptor-binding domain of alpha 1-antitrypsin (1-AT) and was subsequently shown to be a receptor on the surface of hepatocytes, monocytes, and neutrophils for recognition of alpha 1-AT-elastase and several other serpin-enzyme complexes (Perlmutter, D. H., Glover, G. I., Rivetna, M., Schasteen, C. S., and Fallon, R.J. (1990) Proc. Natl. Acad. Sci. U.S.A. 87, 3753-3757). Studies of the minimal requirements for binding to SEC receptor (SEC-R) showed that a pentapeptide FVFLM within the carboxyl-terminal tail of alpha 1-AT was sufficient for binding to SEC-R and interacted with SEC-R in a sequence-specific manner (Joslin, G., Krause, J. E., Hershey, A. D., Adams, S. P., Fallon, R. J., and Perlmutter, D. H. (1991) J. Biol. Chem. 266, 21897-21902). Sequence motifs bearing homology with this pentapeptide domain were found in the amyloid-beta peptide, and amyloid-beta peptide 1-42 was shown to compete for binding to SEC-R on hepatoma cells (Joslin, G., Fallon, R. J., Bullock, J., Adams, S. P., and Perlmutter, D. H. (1991) J. Biol. Chem, 266, 11281-11288). In this study we examined the sequence specificity by which amyloid-beta peptide competes for binding to SEC-R and examined the possibility that SEC-R is expressed in cells of neuronal origin. The results show that amyloid-beta-(25-35) and amyloid-beta-(31-35) compete for binding to SEC-R as effectively as amyloid-beta-(1-39), amyloid-beta-(1-40), and amyloid-beta-(1-42). Amyloid-beta-(1-16) does not compete for binding to SEC-R. There is cross-competition for binding to the same site by 125I-peptide 105Y and amyloid-beta-(25-35) as well as by 125I-Y amyloid-beta-(25-35) and peptide 105Y. By deletions and substitutions within amyloid-beta-(25-35) and generation of chimeric amyloid-beta-alpha 1-AT peptides, amyloid-beta-(31-35) is shown to be critical for binding to the SEC receptor. However, the upstream region, amyloid-beta-(25-30), also contributes to recognition by SEC-R. The SEC-R is present on the surface of a neuronal cell line PC12 as well as that of murine cortical neurons in primary culture, and the specificity of neuronal SEC-R for amyloid-beta peptide is identical to that on hepatoma cells. Finally SEC-R mediates internalization and degradation of amyloid beta-peptide in PC12 cells.(ABSTRACT TRUNCATED AT 400 WORDS)