Glioblastomas and brain metastases differentiation following an MRI texture analysis-based radiomics approach

PurposeTo evaluate the potential of 2D texture features extracted from magnetic resonance (MR) images for differentiating brain metastasis (BM) and glioblastomas (GBM) following a radiomics approach.MethodsThis retrospective study included 50 patients with BM and 50 with GBM who underwent T1-weighted MRI between December 2010 and January 2017. Eighty-eight rotation-invariant texture features were computed for each segmented lesion using six texture analysis methods. These features were also extracted from the four images obtained after applying the discrete wavelet transform (88 features × 4 images). Three feature selection methods and five predictive models were evaluated. A 5-fold cross-validation scheme was used to randomly split the study group into training (80 patients) and testing (20 patients), repeating the process ten times. Classification was evaluated computing the average area under the receiver operating characteristic curve. Sensibility, specificity and accuracy were also computed. The whole process was tested quantizing the images with different gray-level values to evaluate their influence in the final results.ResultsHighest classification accuracy was obtained using the original images quantized with 128 gray-levels and a feature selection method based on the p-value. The best overall performance was achieved using a support vector machine model with a subset of 32 features (AUC = 0.896 ± 0.067, sensitivity of 82% and specificity of 80%). Naïve Bayes and k-nearest neighbors models showed also valuable results (AUC ≈ 0.8) with a lower number of features (<13), thus suggesting that these models may be more generalizable when using external validations.ConclusionThe proposed radiomics MRI approach is able to discriminate between GBM and BM with high accuracy employing a set of 2D texture features, thus helping in the diagnosis of brain lesions in a fast and non-invasive way.     

Breast lesion characterization using Quantitative Ultrasound (QUS) and derivative texture methods

PurposeAccurate and timely diagnosis of breast cancer is extremely important because of its high incidence and high morbidity. Early diagnosis of breast cancer through screening can improve overall prognosis. Currently, biopsy remains as the gold standard for tumor pathological confirmation. Development of diagnostic imaging techniques for rapid and accurate characterization of breast lesions is required. We aim to evaluate the usefulness of texture-derivate features of QUS spectral parametric images for non-invasive characterization of breast lesions.MethodsQUS Spectroscopy was used to determine parametric images of mid-band fit (MBF), spectral slope (SS), spectral intercept (SI), average scatterer diameter (ASD), and average acoustic concentration (AAC) in 204 patients with suspicious breast lesions. Subsequently, texture analysis techniques were used to generate texture maps from parametric images to quantify heterogeneities of QUS parametric images. Further, a second-pass texture analysis was applied to obtain texture-derivate features. QUS parameters, texture-parameters and texture-derivate parameters were determined from both tumor core and a 5-mm tumor margin and were used in comparison to histopathological analysis in order to develop a diagnostic model for classifying breast lesions as either benign or malignant. Both leave-one-out and hold-out cross-validations were used to evaluate the performance of the diagnostic model. Three standard classification algorithms including a linear discriminant analysis (LDA), k-nearest neighbors (KNN), and support vector machines-radial basis function (SVM-RBF) were evaluated.ResultsCore and margin information using the SVM-RBF attained the best classification performance of 90% sensitivity, 92% specificity, 91% accuracy, and 0.93 AUC utilizing QUS parameters and their texture derivatives, evaluated using leave-one-out cross-validation. Implementation of hold-out cross-validation using combination of both core and margin information and SVM-RBF achieved average accuracy and AUC of 88% and 0.92, respectively.ConclusionsQUS-based framework and derivative texture methods enable accurate classification of breast lesions. Evaluation of the proposed technique on a large cohort using hold-out cross-validation demonstrates its robustness and its generalization.     

Follow-up assessment of patients with Pure Neural Leprosy in a reference center in Rio de Janeiro—Brazil

IntroductionPure Neural Leprosy (PNL) is a rare clinical form of leprosy in which patients do not present with the classical skin lesions but have a high burden of the disability associated with the disease. Clinical characteristics and follow up of patients in PNL are still poorly described in the literature.ObjectiveThis paper aims to describe the clinical, electrophysiological and histopathological characteristics of PNL patients, as well as their evolution after multidrug therapy (MDT).MethodsFifty-two PNL patients were selected. Clinical, nerve conduction studies (NCS), histopathological and anti-PGL-1serology were evaluated. Patients were also assessed monthly during the MDT. At the end of the MDT, all of the patients had a new neurological examination and 44 were submitted to another NCS.ResultsParesthesia was the complaint most frequently reported by patients, and in the neurological examination the most common pattern observed was impairment in sensory and motor examination and a mononeuropathy multiplex. Painful nerve enlargement, a classical symptom of leprosy neuropathy, was observed in a minority of patients and in the motor NCS axonal injuries, alone or in combination with demyelinating features, were the most commonly observed. 88% of the patients did not present any leprosy reaction during MDT. There was no statistically significant difference between the neurological examinations, nor the NCS pattern, performed before and after the MDT.DiscussionThe classical hallmarks of leprosy neuropathy are not always present in PNL making the diagnosis even more challenging. Nerve biopsy is an important tool for PNL diagnosis as it may guide therapeutic decisions. This paper highlights unique characteristics of PNL in the spectrum of leprosy in an attempt to facilitate the diagnosis and management of these patients.     

Rest-activity rhythm in breast cancer survivors: an update based on non-parametric indices

ABSTRACT

Recently we evaluated by actigraphy the rest-activity circadian rhythm (RAR) in breast cancer (BC) survivors at 5 years from primary diagnosis, as well as in a control group with similar age and body mass index (BMI). RAR, analyzed by Cosinor method, resulted significantly different in BC survivors compared to healthy subjects: BC survivors showed lower values of MESOR and Amplitude (A), while acrophase (φ) was similar in the two groups.

Now, using non-parametric methods we have detected Interdaily Stability (IS), Intradaily Variability (IV), nocturnal activity (L5), and daily activity (M10) on the same sample of previous study: 15 BC survivors at 5 years from the primary diagnosis (mean age = 56.7 ± 6.6 yrs; mean BMI = 24.5 ± 3.8 Kg/m²) and 13 healthy controls (mean age = 54.4 ± 7.2 yrs; mean BMI = 25.2 ± 2.8 Kg/m²).

The non-parametric indices showed that in BC-group IV was significantly higher than in Ctrl-group (0.86 vs. 0.65 a.u. in BC and Ctrl, respectively; p <.01), while L5 (11.27 vs. 34.41 a.c. in BC and Ctrl, respectively; p <.0001) and M10 (326.82 vs. 428.07 a.c. in BC and Ctrl, respectively; p <.01) were significantly lower compared to Ctrl-group.

The data suggest that BC patients need constant clinical assessment of RAR characteristics along the years following the primary diagnosis. The analysis of RAR in all its components, parametric and non-parametric, is important to detect alterations in the sleep-wake cycle and can be useful for developing new strategies for health protection, such as structured and tailored physical activity programs, to improve circadian activity level in order to raise the quality of life in BC survivors.     

缓解性血清阴性对称性滑膜炎伴凹陷性水肿22例临床回顾

目的:总结缓解性血清阴性对称性滑膜炎伴凹陷性水肿(RS3PE)的临床特点及转归,为该病的诊治提供更多的参考依据。方法:回顾性分析我院2004年1月至2016年2月的22例我院收治的RS3PE住院病例的临床数据和随访资料,总结该病的特点。结果:本研究纳入男性13例(59.1%),女性9例(40.9%),平均发病年龄(65.6±10.0)岁(40.0～79.0岁),发病至确诊时间12天至2年,均为急性起病,并伴有不同程度的对称性多关节炎及肢端水肿。发热6例(27.3%)。主要实验室指标:血红蛋白(109.1±16.0)g·L~(-1),血小板计数(277.0±91.8)×109·L~(-1),血清白蛋白(36.1±5.0)g·L~(-1),C反应蛋白(62.5±67.4)mg·L~(-1),血沉(58.1±33.9)mm/h。测血清铁蛋白的6例中有5例(83.3%)升高,测肿瘤标记的10例均正常。18例(81.8%)行关节影像学检查,均未见骨破坏。5例行唇腺粘膜活检、2例行骨髓穿刺活检均无异常。住院期间治疗后症状均明显好转,4例(18.2%)单用糖皮质激素,18例(81.8%)在激素基础上联用慢作用药物。电话随访4.0月～8.0年(中位随访时间5.0年),有5例(22.7%)关节症状和肢端水肿持续缓解,4例(18.2%)症状减轻,1例(4.5%)反复,3例(13.6%)发展为类风湿关节炎,9例(40.9%)失访,确诊淋巴瘤1例(4.5%)。老年组RS3PE患者红细胞计数、血红蛋白、血清总蛋白及白蛋白低于非老年组。结论:RS3PE多为老年患者,急性起病,为对称性多关节炎伴肢端水肿,常伴贫血及低蛋白血症,对糖皮质激素敏感,有时可伴肿瘤,关于RS3PE的随访观察病例数迄今仍不足。     

Insecticidal efficacy of two isolates of Beauveria bassiana (Bals.) (Vuill.), on the second larval stage of Leptinotarsa decemlineata (Say) (Col.: Chrysomelidae)

Leptinotarsa decemlineata is a serious potato pest throughout the world that has developed resistance to many insecticides. This study investigated the virulence of two indigenous isolates of Beauveria bassiana, consisting of DEBI007 and IR1217C. Five different concentrations of each isolate, 1 × 10⁷, 3.1 × 10⁷, 1 × 10⁸, 3.1 × 10⁸ and 1 × 10⁹ conidia ml^?1, were applied in bioassay using dipping method on second instar larvae. Control insects were treated with distilled water containing 0.01% Tween-80 (Sigma, Germany) as surfactant. The results showed that mortality percentages due to DEBI007 and IR1217C isolates in the highest concentration were 60% and 57.77%, respectively. There was a significant difference between efficacies of different conidial concentrations; therefore, with rise in conidial concentration, the mortality percentage was increased. The DEBI007 isolate that was previously isolated from the soil was more effective than IR1217C isolate. Calculated LC₅₀ for this effective isolate was 1.4 × 10⁷ conidia ml^?1, which, respectively, is the highest concentration.     

In vitro susceptibility of human Blastocystis subtypes to simeprevir

Introduction and aimBlastocystis is a common enteric parasite, having a worldwide distribution. Many antimicrobial agents are effective against it, yet side effects and drug resistance have been reported. Thus, ongoing trials are being conducted for exploring anti-Blastocystis alternatives. Proteases are attractive anti-protozoal drug targets, having documented roles in Blastocystis. Serine proteases are present in both hepatitis C virus and Blastocystis. Since drug repositioning is quite trendy, the in vitro efficacy of simeprevir (SMV), an anti-hepatitis serine protease inhibitor, against Blastocystis was investigated in the current study.MethodsStool samples were collected from patients, Alexandria, Egypt. Concentrated stools were screened using direct smears, trichrome, and modified Ziehl-Neelsen stains to exclude parasitic co-infections. Positive stool isolates were cultivated, molecularly subtyped for assessing the efficacy of three SMV doses (100,150, and 200 μg/ml) along 72 hours (h), on the most common subtype, through monitoring parasite growth, viability, re-culture, and also via ultrastructure verification. The most efficient dose and duration were later tested on other subtypes.ResultsResults revealed that Blastocystis was detected in 54.17% of examined samples. Molecularly, ST3 predominated (62%), followed by ST1 (8.6%) and ST2 (3.4%). Ascending concentrations of SMV progressively inhibited growth, viability, and re-culture of treated Blastocystis, with a non-statistically significant difference when compared to the therapeutic control metronidazole (MTZ). The most efficient dose and duration against ST3 was 150 µg/ml for 72 h. This dose inhibited the growth of ST3, ST1, and ST2 with percentages of 95.19%, 94.83%, and 94.74%, successively and viability with percentages of 98.30%, 98.09%, and 97.96%, successively. This dose abolished Blastocystis upon re-culturing. Ultra-structurally, SMV induced rupture of Blastocystis cell membrane leading to necrotic death, versus the reported apoptotic death caused by MTZ. In conclusion, 150 µg/ml SMV for 72 h proved its efficacy against ST1, ST2, and ST3 Blastocystis, thus sparing the need for pre-treatment molecular subtyping in developing countries.     

Modulation of Mutational Landscape in HER2-Positive Breast Cancer after Neoadjuvant Chemotherapy

IntroductionIn early-stage HER2 positive breast cancer (BC) patients, tumor response to neoadjuvant chemotherapy (NACT) predict survival outcomes. Patients achieving less than pathological complete response (pCR) have a worse prognosis, however, this group is heterogeneous. Nowadays limited data on predictive/prognostic biomarkers in patients with residual cancer disease are available.MethodsUsing next-generation sequencing technology, we evaluated a panel of 21 cancer genes in a group of HER2 positive BC patients with residual disease after NACT. A control group of patients who achieved the pCR was selected too. The BC mutational profile was analyzed on both the tumor diagnostic biopsy and matched residual disease.ResultsOverall, the detection rate of mutations was 79% in the No-pCR group versus 90% in the pCR cohort and 98% in the residual BC. The most mutated genes were TP53 and PIK3CA. No correlations between single gene mutations and survival outcomes were found. In no-pCR cohort, 52% of patients had different mutational profile after NACT, 69% of them had an increased in the number of mutated genes. Mutational profile changes from diagnostic biopsy to residual BC were a negative prognostic factor in term of relapse free survival: recurrence probability in different gene profile sub-group was 42% vs 0% in the same profile one (P = .019).ConclusionsTreatment selective pressure on tumor cells due to NACT changed the gene mutational profile in more than half of BC patient with residual tumor disease. Treatment-induced gene mutations significantly increase the risk of relapse. Profiling primary and residual BC is a major step in order to further personalized adjuvant treatment strategy.     

两种免疫磁珠分离法分选小鼠骨髓粒-单核祖细胞的效率比较

摘要 目的：提取小鼠骨髓细胞（bone marrow cell, BMC）,用两种不同的免疫磁珠分离（magnetic activated cell sorting, MACS）试剂盒从小鼠BMC中分选提纯粒-单核祖细胞（granulocyte-monocyte progenitor, GMP）,比较这两种免疫磁珠的分选效率。方法：从小鼠股骨和胫骨中提取BMC,通过两种不同的MACS试剂盒,即Lineage阳性细胞清除试剂盒和CD117阳性细胞分选试剂盒,分别得到Lineage-细胞群和CD117⁺细胞群,用代表GMP细胞表面标志物的荧光抗体标记,孵育后通过流式细胞荧光分选技术得到GMP细胞,并且对比得到GMP细胞的效率。结果：每2只野生型C57BL/6J小鼠可共收集骨髓细胞（7.02±1.24）×10⁷个,细胞活力为（91.86±5.24）%。经过Lineage阳性细胞清除试剂盒得到的细胞数量为（5.71±2.86）×10⁶个;经过CD117阳性细胞分选试剂盒得到的细胞数量为（2.70±0.56）×10⁶个。Lineage磁珠分选纯化得到的GMP细胞数占总细胞数的比例为（10.90±1.37）%,CD117磁珠分选纯化得到的GMP细胞数占总细胞数的比例为（4.83±2.08）%。结论：Lineage阳性细胞清除试剂盒能更有效分选小鼠骨髓细胞中的粒-单核祖细胞。     

A French single-center experience on allogeneic stem cell transplant cryopreservation during severe acute respiratory syndrome coronavirus 2 pandemic

Background aimsAllogeneic hematopoietic stem cell transplantation (allo-SCT) is a curative treatment for chemo-resistant hematological malignancies. Because of transport restriction imposed by the coronavirus disease 2019 pandemic, regulatory bodies and societies recommended graft cryopreservation before recipient conditioning. However, the freezing and thawing processes, including washing steps, might impair CD34+ cell recovery and viability, thereby impacting the recipient engraftment. Over 1 year (between March 2020 and May 2021), we aimed to analyze the results of frozen/thawed peripheral blood stem cell allografts in terms of stem cell quality and clinical outcomes.MethodsTransplant quality was evaluated by comparing total nucleated cells (TNCs), CD34+ cells and colony-forming unit–granulocyte/macrophage (CFU-GM)/kg numbers as well as TNC and CD34+ cell viabilities before and after thawing. Intrinsic biological parameters such as granulocyte, platelet and CD34+ cell concentrations were analyzed, as they might be responsible for a quality loss. The impact of the CD34+ cell richness of the graft on TNC and CD34 yields was evaluated by designing three groups of transplants based on their CD34 /kg value at collection: >8 × 10 ⁶/kg, between 6 and 8 × 10⁶/kg and <6 × 10⁶/kg. The consequences of cryopreservation were compared in the fresh and thawed group by evaluating the main transplant outcomes.ResultsOver 1 year, 76 recipients were included in the study; 57 patients received a thawed and 19 patients a fresh allo-SCT. None received allo-SCT from a severe acute respiratory syndrome coronavirus 2–positive donor. The freezing of 57 transplants led to the storage of 309 bags, for a mean storage time (between freezing and thawing) of 14 days. For the fresh transplant group, only 41 bags were stored for potential future donor lymphocyte infusions. Regarding the graft characteristics at collection, median number of cryopreserved TNC and CD34+ cells/kg were greater than those for fresh infusions. After thawing, median yields were 74.0%, 69.0% and 48.0% for TNC, CD34+ cells and CFU-GM, respectively. The median TNC dose/kg obtained after thawing was 5.8 × 10⁸, with a median viability of 76%. The median CD34+ cells/kg was 5 × 10⁶, with a median viability of 87%. In the fresh transplant group, the median TNC/kg was 5.9 × 10⁸/kg, and the median CD34+ cells/kg and CFU-GM/kg were 6 × 10⁶/kg and 276.5 × 10⁴/kg, respectively. Sixty-one percent of the thawed transplants were out of specifications regarding the CD34+ cells/ kg requested cell dose (6 × 10⁶/kg) and 85% of them would have had this dose if their hematopoietic stem cell transplant had been infused fresh. Regarding fresh grafts, 15.8% contained less than 6 × 10⁶ CD34+ cells /kg and came from peripheral blood stem cells that did not reach 6 × 10⁶ CD34+ cells /kg at collection. Regarding the factor that impaired CD34 and TNC yield after thawing, no significant impact of the granulocyte count, the platelet count or the CD34+ cells concentration/µL was observed. However, grafts containing more than 8 × 10 ⁶/kg at collection showed a significantly lower TNC and CD34 yield.ConclusionsTransplant outcomes (engraftment, graft-versus-host disease, infections, relapse or death) were not significantly different between the two groups.     

Non-computer-assisted liquid-based cytology for diagnosis of oral squamous cell carcinoma

Abstract

The development of oral squamous cell carcinoma (OSCC) occasionally follows the neoplastic progression of other premalignant lesions. Although biopsy is the definitive diagnostic method, liquid-based cytology is an adequate method for screening suspicious lesions. We compared liquid-based cytology to histology for diagnosis of OSCC in patients with oral lesions that raised clinical suspicion of malignancy. Our sample consisted of 48 patients. Cytological samples were obtained by scraping the lesion superficially using Cytobrush^®. We conducted cytological and histopathological evaluation of all preparations. We estimated sensitivity and specificity levels as well as positive and negative predictive values. The degree of inter-observer agreement for both methods was assessed using the kappa index. Twenty-eight (58.3%) of the cases finally were diagnosed with OSCC and 20 (41.7%) were determined to be premalignant lesions. We observed eight false negatives and no false positives; OSCC prevalence was 56.5%. The values for diagnostic indices were: sensitivity, 69% (CI 95%, prevalence 51.87); specificity, 100%; positive predictive value, 100%; negative predictive value, 71% (CI 95% 54.82). A kappa index of 0.622 (CI 95% 0.93, 0.39) was observed.     

Nonclassic Presentation of Pheochromocytoma: Difficulties in Diagnosis and Management of the Normotensive Patient

ObjectiveTo report an unusual presentation of pheochromocytoma along with challenges in diagnosis and management.MethodsWe report a clinical case history and describe diagnostic methods and pitfalls. The preoperative medical preparation in a normotensive patient is described.ResultsA 33-year-old man fell off a ladder, resulting in C6 paraplegia. After C7-T1 laminectomy, he was transferred to a rehabilitation center where he reported lightheadedness, shortness of breath, and chest pain during therapy sessions. A left adrenal mass was incidentally discovered during the workup to rule out a pulmonary embolism. He reported no history of hypertension, and no elevated blood pressure readings had been documented. Magnetic resonance imaging of the adrenal glands showed a well-defined left adrenal mass measuring 3.9 × 3.2 × 3.3 cm, which was hyperintense on T2-weighted images. Twenty-four hour urinary catecholamine concentrations were unremarkable; urinary metanephrines were markedly elevated. During hospital admission, blood pressure was in the low to normotensive range, requiring a cautious approach to α-adrenergic blockade and surgical preparation. He underwent uneventful laparoscopic left adrenalectomy; surgical pathology was consistent with pheochromocytoma.ConclusionsThis case illustrates a nonclassic presentation of pheochromocytoma and demonstrates that urinary catecholamines alone are not sufficient for a biochemical diagnosis of large pheochromocytomas. Preoperative preparation in normotensive patients can be achieved with α-adrenergic blockade, hydration, and liberal salt intake.     

Acridinium ester conjugated to lectin as chemiluminescent histochemistry marker

Abstract

Cell differentiation/dedifferentiation includes changes in oligosaccharide composition and distribution in the cell surface glycoconjugates. Lectins have been used as auxiliary tools in histopathological diagnosis of mammary, uterus and brain pathologies. Acridinium ester (AE) conjugated to biomolecules has been employed in chemiluminescent analytical applications. This work aimed to use a lectin, concanavalin A (Con A), conjugated to AE as a chemiluminescent histochemistry tool. Biopsies of normal and infiltrating duct carcinoma (IDC) of mammary tissues were treated by a Con A–AE derivative. Photon emission, observed during the breakage of the chemical bound between Con A and AE, was quantified, expressed in relative light units (RLU) and correlated to the labelling of the normal and transformed tissues. The results demonstrated that RLU presented a linear relationship with the labelled tissue area in the range 0.125–1.0?cm² (r=0.98). Furthermore, RLU was much higher for the IDC (1283.920×10³±220.621×10³) than the normal tissue (2.565×10³±0.247×10³), namely, about 500 times higher. The Con A–AE conjugation efficiency, differential staining of normal and IDC tissues, and quantification of results contribute to a decrease in the subjectivity in routine histopathological diagnoses and indicate that acrydinum ester can join other lectin marker to be used in histochemistry.     

基于染色体片段导入系发掘抗玉米丝黑穗病主效QTL

选用抗玉米丝黑穗病自交系Mo17和SH15为供体,与受体感病自交系黄早四和昌7-2构建回交群体(BC3F1\BC4F2),通过田间人工接种玉米丝黑穗病原菌鉴定抗病性表现,评价群体抗病性。研究结果显示黄早四×(黄早四×Mo17)BC4F2群体发病率明显高于BC3F1群体;两个BC4F2黄早四×(黄早四×Mo17)和昌7-2×(昌7-2×SH15)群体的发病率差异较大。采用SSR标记分析抗病株的供体染色体导入片段,发现随着回交次数的增多,导入片段数量减少,但不同回交群体中供体导入片段数目明显不同。通过连锁不平衡分析,在染色体2.09和3.04区段发掘和验证2个抗玉米丝黑穗病主效QTL,连锁标记分别为umc2077和phio53或bnlg1965。本文研究结果为抗丝黑穗病基因精细定位和分子聚合育种提供了信息和材料。     

The delay of breast cancer diagnosis and management during the Syrian war

BackgroundEarly detection of breast cancer (BC) is crucial for better prognosis especially in low-income countries, where advanced cancer stages are common. The Syrian war severely affected the healthcare system restraining the proper timely management of BC cases. We aimed to investigate the prevalence of patient- and system-related delays in BC diagnosis and management in Syria in addition to their predisposing characteristics and impact on the staging.MethodsThis is a cross-sectional retrospective cohort study on patients followed by the BC unit at Al-Bairouni main cancer center in Syria. The data were collected through personal interviews and retrospective revision of patients’ records.ResultsA total number of 519 patients were recruited; A quarter of them (n = 126) reported more than three months intervals between symptoms recognition and presentation to a physician. Additionally, 72 (13.9 %) patients received a confirmed diagnosis more than three months after presentation, and 12 (2.3 %) started treatment at least three months after the diagnosis. Patients who suffered from war-related inaccessibility to healthcare were 2.55 [1.58–4.11] times more likely to report significant delays. Additionally, the most common self-reported reasons for patient delay were the lack of awareness, which was more evident for less common symptoms like the change in breasts size, and shyness. Patients who reported significant delays were more likely to receive an advanced-stage diagnosis.ConclusionWar-related inaccessibility to healthcare rendered a significant group of BC patient susceptible to evident delay. This combined with significant system delays because of the overwhelmed hospitals, high costs, and insufficient personnel, equipment, medications, and training. However, personal factors, which might not be directly related to the war, like the inadequate awareness of rare symptoms and shyness still necessitate urgent interventions on the public knowledge and performed screening practices. These delays associated with receiving advanced-stage diagnoses and minimizing them can return better prognoses.     

Automating chest radiograph imaging quality control

PurposeTo automate diagnostic chest radiograph imaging quality control (lung inclusion at all four edges, patient rotation, and correct inspiration) using convolutional neural network models.MethodsThe data comprised of 2589 postero-anterior chest radiographs imaged in a standing position, which were divided into train, validation, and test sets. We increased the number of images for the inclusion by cropping appropriate images, and for the inclusion and the rotation by flipping the images horizontally. The image histograms were equalized, and the images were resized to a 512 × 512 resolution. We trained six convolutional neural networks models to detect the image quality features using manual image annotations as training targets. Additionally, we studied the inter-observer variability of the image annotation.ResultsThe convolutional neural networks’ areas under the receiver operating characteristic curve were >0.88 for the inclusions, and >0.70 and >0.79 for the rotation and the inspiration, respectively. The inter-observer agreement between two human annotators for the assessed image-quality features were: 92%, 90%, 82%, and 88% for the inclusion at patient’s left, patient’s right, cranial, and caudal edges, and 78% and 89% for the rotation and inspiration, respectively. Higher inter-observer agreement was related to a smaller variance in the network confidence.ConclusionsThe developed models provide automated tools for the quality control in a radiological department. Additionally, the convolutional neural networks could be used to obtain immediate feedback of the chest radiograph image quality, which could serve as an educational instrument.