Open-source, fully-automated hybrid cardiac substructure segmentation: development and optimisation.

Radiotherapy for thoracic and breast tumours is associated with a range of cardiotoxicities. Emerging evidence suggests cardiac substructure doses may be more predictive of specific outcomes, however, quantitative data necessary to develop clinical planning constraints is lacking. Retrospective analysis of patient data is required, which relies on accurate segmentation of cardiac substructures. In this study, a novel model was designed to deliver reliable, accurate, and anatomically consistent segmentation of 18 cardiac substructures on computed tomography (CT) scans. Thirty manually contoured CT scans were included. The proposed multi-stage method leverages deep learning (DL), multi-atlas mapping, and geometric modelling to automatically segment the whole heart, cardiac chambers, great vessels, heart valves, coronary arteries, and conduction nodes. Segmentation performance was evaluated using the Dice similarity coefficient (DSC), mean distance to agreement (MDA), Hausdorff distance (HD), and volume ratio. Performance was reliable, with no errors observed and acceptable variation in accuracy between cases, including in challenging cases with imaging artefacts and atypical patient anatomy. The median DSC range was 0.81-0.93 for whole heart and cardiac chambers, 0.43-0.76 for great vessels and conduction nodes, and 0.22-0.53 for heart valves. For all structures the median MDA was below 6 mm, median HD ranged 7.7-19.7 mm, and median volume ratio was close to one (0.95-1.49) for all structures except the left main coronary artery (2.07). The fully automatic algorithm takes between 9 and 23 min per case. The proposed fully-automatic method accurately delineates cardiac substructures on radiotherapy planning CT scans. Robust and anatomically consistent segmentations, particularly for smaller structures, represents a major advantage of the proposed segmentation approach. The open-source software will facilitate more precise evaluation of cardiac doses and risks from available clinical datasets.

Standardising Breast Radiotherapy Structure Naming Conventions: A Machine Learning Approach.

In progressing the use of big data in health systems, standardised nomenclature is required to enable data pooling and analyses. In many radiotherapy planning systems and their data archives, target volumes (TV) and organ-at-risk (OAR) structure nomenclature has not been standardised. Machine learning (ML) has been utilised to standardise volumes nomenclature in retrospective datasets. However, only subsets of the structures have been targeted. Within this paper, we proposed a new approach for standardising all the structures nomenclature by using multi-modal artificial neural networks. A cohort consisting of 1613 breast cancer patients treated with radiotherapy was identified from Liverpool & Macarthur Cancer Therapy Centres, NSW, Australia. Four types of volume characteristics were generated to represent each target and OAR volume: textual features, geometric features, dosimetry features, and imaging data. Five datasets were created from the original cohort, the first four represented different subsets of volumes and the last one represented the whole list of volumes. For each dataset, 15 sets of combinations of features were generated to investigate the effect of using different characteristics on the standardisation performance. The best model reported 99.416% classification accuracy over the hold-out sample when used to standardise all the nomenclatures in a breast cancer radiotherapy plan into 21 classes. Our results showed that ML based automation methods can be used for standardising naming conventions in a radiotherapy plan taking into consideration the inclusion of multiple modalities to better represent each volume.

Impact of salvage treatment modalities in patients with positive FDG-PET/CT after R-CHOP chemotherapy for aggressive B-cell non-Hodgkin lymphoma.

INTRODUCTION: To compare outcomes of different salvage treatment modalities in patients with aggressive B-cell non-Hodgkin lymphoma (NHL) who remain FDG-PET positive after R-CHOP chemotherapy. Existing data on these patients with FDG-PET primary refractory disease are limited. METHODS: Patients with diffuse large B-cell lymphoma or grade 3 follicular lymphoma were retrospectively reviewed from the Prince of Wales Hospital databases. Eligibility criteria were: age≥18 years, treated with R-CHOP, with positive post-chemotherapy FDG-PET. Salvage treatment modalities were: radical radiotherapy (RT, dose≥30 Gy), high dose chemotherapy and autologous stem cell transplant (ASCT), or non-radical management. Survival was calculated from date of post-chemotherapy FDG-PET to last follow-up. RESULTS: Twenty-six patients from 2003-2015 met the inclusion criteria. Median age was 60 (range 19-84). Most had adverse baseline features: 21 (81%) stage III-IV, 24 (92%) bulky disease and nine (35%) skeletal involvement. Characteristics of PET-positivity post-chemotherapy were single site in 16 (62%), sites of prior bulk in 24 of 24, skeletal sites in five of nine, and able to be encompassed by RT in 21 (81%). Salvage treatment was: radical RT in 17 (65%), ASCT in four (15%) and non-radical in five (20%). Median follow-up of surviving patients was 31 months. Kaplan-Meier estimates of 3-year PFS and OS were 41% and 52%, respectively. By salvage modality, 3-year PFS was 51% for RT, 25% for ASCT and 20% for non-radical treatment, (P = 0.453); 3-year OS was respectively 65%, 25% and 40% (P = 0.173). CONCLUSION: Patients with FDG-PET positive disease after R-CHOP for aggressive B-cell NHL are salvageable with radiotherapy.

Role of bacterial and genetic factors in gastric cancer in Costa Rica.

AIM: To evaluate several risk factors for gastric cancer (GC) in Costa Rican regions with contrasting GC incidence rate (GCIR). METHODS: According to GCIR, 191 Helicobacter pylori (H pylori)-positive patients were classified into groups A (high GCIR, n = 101) and B (low GCIR, n = 90). Human DNA obtained from biopsy specimens was used in the determination of polymorphisms of the genes coding for interleukin (IL)-1beta and IL-10 by PCR-RFLP, and IL-1RN by PCR. H pylori DNA extractions obtained from clinical isolates of 83 patients were used for PCR-based genotyping of H pylori cagA, vacA and babA2. Human DNA from gastric biopsies of 52 GC patients was utilized for comparative purposes. RESULTS: Cytokine polymorphisms showed no association with GCIR variability. However, gastric atrophy, intestinal metaplasia and strains with different vacA genotypes in the same stomach (mixed strain infection) were more frequently found in group A than in group B, and cagA and vacA s1b were significantly associated with high GCIR (P = 0.026 and 0.041, respectively). IL-1beta+3954_T/C (OR 2.1, 1.0-4.3), IL-1RN*2/L (OR 3.5, 1.7-7.3) and IL-10-592_C/A (OR 3.2, 1.5-6.8) were individually associated with GC, and a combination of these cytokine polymorphisms with H pylori vacA s1b and m1 further increased the risk (OR 7.2, 1.4-36.4). CONCLUSION: Although a proinflammatory cytokine genetic profile showed an increased risk for developing GC, the characteristics of H pylori infection, in particular the status of cagA and vacA genotype distribution seemed to play a major role in GCIR variability in Costa Rica.

Serum pepsinogen levels, Helicobacter pylori CagA Status, and cytokine gene polymorphisms associated with gastric premalignant lesions in Costa Rica.

The detection of gastric premalignant lesions, atrophic gastritis, corpus atrophic gastritis, and intestinal metaplasia, using several potential markers was examined in Costa Rica. Depending on the lesion investigated, from a total of 223 dyspeptic patients, 58 (26.0%), 31 (13.9%), or 23 (10.3%) were histologically diagnosed with atrophic gastritis, corpus atrophic gastritis, or intestinal metaplasia, respectively. Sera were used for the measurement of pepsinogen (PG) and Helicobacter pylori CagA antibody (CagA-ab) levels by ELISA, and human genomic DNAs were used for the genotyping of interleukin (IL)-1beta (-511 and +3954), IL-10 (-1082 and -592), and IL-1RN intron 2 by PCR and RFLP. Multivariate analysis was done adjusting for sex, age, and H. pylori seropositivity. Low PG levels (L-PG; PG I < or = 70 microg/L + PG I/II < or = 3), very low PG levels (VL-PG; PG I < or = 30 microg/L + PG I/II < or = 2), and CagA-ab were individually associated with all premalignant lesions whereas IL-1beta +3954T-carrier and IL-1RN homozygous 2 allele were associated with intestinal metaplasia. VL-PG, for corpus atrophic gastritis detection, was the single marker with the highest combination of test characteristics, sensitivity (77.4%), specificity (80.7%), positive predictive value (39.3%), negative predictive value (95.7%), and seropositivity rate (27.4%), expected to improve after periodic measurements. Combined examinations of VL-PG and CagA-ab improved the specificity (92.7%) and positive predictive value (62.2%), with similar sensitivity (74.2%) and negative predictive value (95.7%). In conclusion, corpus atrophic gastritis detection with periodic measurements of serum PG, alone or in combination with CagA-ab status, to identify high gastric cancer risk, seems to be the method best suited for mass screening in Costa Rica.

Diversity of Helicobacter pylori cagA and vacA genes in Costa Rica: its relationship with atrophic gastritis and gastric cancer.

BACKGROUND: Associations between Helicobacter pylori gene diversity and gastric cancer have not been reported on in Costa Rica, despite its being one of the countries with the highest gastric cancer incidence and mortality rates in the world. The aim of this study was to determine the prevalence of H. pylori cagA and vacA genes and investigate whether it could be correlated with atrophic gastritis (AG) and gastric cancer (GC) in Costa Rica. MATERIALS AND METHODS: Genomic DNAs from isolates of 104 patients classified into two groups: non-atrophic gastritis group (n = 68) and atrophic gastritis group (n = 36), were subjected to PCR-based genotyping of cagA and vacA genes and their correlation with clinical outcome was investigated. Total DNA extractions from gastric tissues of 25 H. pylori-infected gastric cancer patients were utilized for comparative purposes. RESULTS: The presence of cagA (75.3%), vacA s1b (75.3%), and vacA m1 (74.2%) was detected, and colonization by strains with different vacA genotypes in the same stomach was found in 9.7% of the patients. Age- and sex-adjusted vacA s1b and vacA m1 were associated with GC while only vacA m1 was significantly associated with AG. A tendency for association between cagA and vacA s1b, and AG was reported. CONCLUSIONS: The prevalence status of the cagA and vacA (s1/m1) genes in Costa Rica seems to fall between that found in European/North American and East Asian countries, and both cagA and vacA seem to have clinical relevance in this country.

Helicobacter pylori CagA status associated with gastric cancer incidence rate variability in Costa Rican regions.

BACKGROUND: We evaluated several risk factors for gastric cancer in Costa Rican regions having contrasting gastric cancer incidence rates, despite the small dimensions of the country. METHODS: A total of 180 dyspeptic patients were classified into two groups according to the gastric cancer incidence (GCI) rate in their Costa Rican region: group A, with a high GCI rate (n = 91) and group B, with a low GCI rate (n = 89). Helicobacter pylori infection was detected by rapid urease test, Gram staining, and histological observation. Antral and corpus specimens were obtained to assess the grade of inflammation, topography of gastritis, gastric atrophy, and intestinal metaplasia by histological examination. Serum CagA antibody was measured by an antigen-specific enzyme-linked immunosorbent assay. RESULTS: There was no significant difference in H. pylori prevalence between groups A (73%) and B (63%); however, serum CagA antibody was more frequently detected in group A (79%) than in group B (54%) [P = 0.02; odds ratio (OR), 2.68]. Among patients under 60 years of age, serum CagA antibody was even more frequently detected in group A (81%) than in group B (49%) (P < 0.01; OR, 4.50). The prevalence of corpus-predominant gastritis, atrophic gastritis, and moderate/severe grades of neutrophilic infiltration was higher in serum CagA antibody-positive patients than in CagA antibody-negative patients (P = 0.003, 0.04, and 0.002, respectively). CONCLUSIONS: Infection with H. pylori possessing the cagA gene is associated with the development of severe gastric damage such as gastric atrophy, leading to gastric cancer, and probably influences the differences in GCI between Costa Rican regions.

Triple sinergia: un método eficaz sedoanalgesia para colonoscopías / Tripple synergy: an effective method sedoanalgesia for colonoscopy

Con el desarrollo de las nuevas tecnologías en el campo de diagnóstico, tanto radiológico como endoscópico, apareció un nuevo tipo de paciente, el cual se volvió un reto para el anestesiólogo, ya que por ser sometido a procedimientos invasivos y poco dolorosos, permitía el avance en las técnicas de sedación. Objetivos: A) Comprobar la eficacia del método de sedación con triple sinergia. B) Determinar el número de bolos de propofol necesarios para realizar el procedimiento. C) Evaluar las complicaciones asociadas a la técnica de sedación hipnótica. Materiales y método: Se estudió en forma prospectiva una muestra consecutiva de 1500 pacientes entre los años 1998 y 2000, de edades mayores de los 18 años, que fueron sometidos a colonoscopías, en el Centro Endoscópico Internacional, San José, Costa Rica. Se aplicó la sedación hipnótica con la técnica de la triple en sinergia (midazolam/fentanil/propofol). Resultados: De los 1500 pacientes, un 57,8 por ciento fueron hombres y un 42,4 por ciento , mujeres. La mayoría fueron de la categoría ASAII (n=725). Un total de 208 pacientes (13,8 por ciento) presentaron desaturación. La técnica de sedación fue placentera en el 94 por ciento de los pacientes. Un 1 por ciento presentó náusea y vómito. El 98 por ciento de los pacientes tuvieron amnesia del procedimiento. Solo el 2 por ciento de los casos sintió dolor. Conclusiones: La sedacción hipnótica con la técnica de la triple sinergia es un método efectivo y seguro, tanto para el endoscopista como para el paciente. El costo es mínimo y la incidencia del paciente. Dado a que la desaturación es un efecto relativamente frecuente, recomendamos colocar a todos los pacientes una cánula nasal con oxígeno profiláctico y monitorizarlos con pulsoximetría y electrocardiográficamente, en especial los casos de la categoría ASA III. Descriptores: Sedacción hipnótica, triple sinergia, colonoscopía.