Morphological diagnosis of hematologic malignancy using feature fusion-based deep convolutional neural network.

Leukemia is a cancer of white blood cells characterized by immature lymphocytes. Due to blood cancer, many people die every year. Hence, the early detection of these blast cells is necessary for avoiding blood cancer. A novel deep convolutional neural network (CNN) 3SNet that has depth-wise convolution blocks to reduce the computation costs has been developed to aid the diagnosis of leukemia cells. The proposed method includes three inputs to the deep CNN model. These inputs are grayscale and their corresponding histogram of gradient (HOG) and local binary pattern (LBP) images. The HOG image finds the local shape, and the LBP image describes the leukaemia cell's texture pattern. The suggested model was trained and tested with images from the AML-Cytomorphology_LMU dataset. The mean average precision (MAP) for the cell with less than 100 images in the dataset was 84%, whereas for cells with more than 100 images in the dataset was 93.83%. In addition, the ROC curve area for these cells is more than 98%. This confirmed proposed model could be an adjunct tool to provide a second opinion to a doctor.

Efficient diagnosis of hematologic malignancies using bone marrow microscopic images: A method based on MultiPathGAN and MobileViTv2.

BACKGROUND AND OBJECTIVES: Hematologic malignancies, including the associated multiple subtypes, are critically threatening to human health. The timely detection of malignancies is crucial for their effective treatment. In this regard, the examination of bone marrow smears constitutes a crucial step. Nonetheless, the conventional approach to cell identification and enumeration is laborious and time-intensive. Therefore, the present study aimed to develop a method for the efficient diagnosis of these malignancies directly from bone marrow microscopic images. METHODS: A deep learning-based framework was developed to facilitate the diagnosis of common hematologic malignancies. First, a total of 2033 microscopic images of bone marrow analysis, including the images for 6 disease types and 1 healthy control, were collected from two Chinese medical websites. Next, the collected images were classified into the training, validation, and test datasets in the ratio of 7:1:2. Subsequently, a method of stain normalization to multi-domains (stain domain augmentation) based on the MultiPathGAN model was developed to equalize the stain styles and expand the image datasets. Afterward, a lightweight hybrid model named MobileViTv2, which integrates the strengths of both CNNs and ViTs, was developed for disease classification. The resulting model was trained and utilized to diagnose patients based on multiple microscopic images of their bone marrow smears, obtained from a cohort of 61 individuals. RESULTS: MobileViTv2 exhibited an average accuracy of 94.28% when applied to the test set, with multiple myeloma, acute lymphocytic leukemia, and lymphoma revealed as the three diseases diagnosed with the highest accuracy values of 98%, 96%, and 96%, respectively. Regarding patient-level prediction, the average accuracy of MobileViTv2 was 96.72%. This model outperformed both CNN and ViT models in terms of accuracy, despite utilizing only 9.8 million parameters. When applied to two public datasets, MobileViTv2 exhibited accuracy values of 99.75% and 99.72%, respectively, and outperformed previous methods. CONCLUSIONS: The proposed framework could be applied directly to bone marrow microscopic images with different stain styles to efficiently establish the diagnosis of common hematologic malignancies.

European Association of Nuclear Medicine (EANM) Focus 4 consensus recommendations: molecular imaging and therapy in haematological tumours.

Given the paucity of high-certainty evidence, and differences in opinion on the use of nuclear medicine for hematological malignancies, we embarked on a consensus process involving key experts in this area. We aimed to assess consensus within a panel of experts on issues related to patient eligibility, imaging techniques, staging and response assessment, follow-up, and treatment decision-making, and to provide interim guidance by our expert consensus. We used a three-stage consensus process. First, we systematically reviewed and appraised the quality of existing evidence. Second, we generated a list of 153 statements based on the literature review to be agreed or disagreed with, with an additional statement added after the first round. Third, the 154 statements were scored by a panel of 26 experts purposively sampled from authors of published research on haematological tumours on a 1 (strongly disagree) to 9 (strongly agree) Likert scale in a two-round electronic Delphi review. The RAND and University of California Los Angeles appropriateness method was used for analysis. Between one and 14 systematic reviews were identified on each topic. All were rated as low to moderate quality. After two rounds of voting, there was consensus on 139 (90%) of 154 of the statements. There was consensus on most statements concerning the use of PET in non-Hodgkin and Hodgkin lymphoma. In multiple myeloma, more studies are required to define the optimal sequence for treatment assessment. Furthermore, nuclear medicine physicians and haematologists are awaiting consistent literature to introduce volumetric parameters, artificial intelligence, machine learning, and radiomics into routine practice.

Differential diagnosis of infectious diseases, drug-induced lung injury, and pulmonary infiltration due to underlying malignancy in patients with hematological malignancy using HRCT.

PURPOSE: To differentiate among infectious diseases, drug-induced lung injury (DILI) and pulmonary infiltration due to underlying malignancy (PIUM) based on high-resolution computed tomographic (HRCT) findings from patients with hematological malignancies who underwent chemotherapy or hematopoietic stem cell transplantation. MATERIALS AND METHODS: A total of 221 immunocompromised patients with hematological malignancies who had proven chest complications (141 patients with infectious diseases, 24 with DILI and 56 with PIUM) were included. Two chest radiologists evaluated the HRCT findings, including ground-glass opacity, consolidation, nodules, and thickening of bronchovascular bundles (BVBs) and interlobular septa (ILS). After comparing these CT findings among the three groups using the χ2test, multiple logistic regression analyses (infectious vs noninfectious diseases, DILI vs non-DILI, and PIUM vs non-PIUM) were performed to detect useful indicators for differentiation. RESULTS: Significant differences were detected in many HRCT findings by the χ2 test. The results from the multiple logistic regression analyses identified several indicators: nodules without a perilymphatic distribution [p = 0.012, odds ratio (95% confidence interval): 4.464 (1.355-11.904)], nodules with a tree-in-bud pattern [p = 0.011, 8.364 (1.637-42.741)], and the absence of ILS thickening[p = 0.003, 3.621 (1.565-8.381)] for infectious diseases, the presence of ILS thickening [p = 0.001, 7.166 (2.343-21.915)] for DILI, and nodules with a perilymphatic distribution [p = 0.011, 4.256 (1.397-12.961)] and lymph node enlargement (p = 0.008, 3.420 (1.385-8.441)] for PIUM. CONCLUSION: ILS thickening, nodules with a perilymphatic distribution, tree-in-bud pattern, and lymph node enlargement could be useful indicators for differentiating among infectious diseases, DILI, and PIUM in patients with hematological malignancies.

Sonographic features of secondary involvement of skin and subcutaneous tissues by hematologic malignancies.

OBJECTIVES: To evaluate the sonographic features of secondary involvement of skin and subcutaneous tissues by hematologic malignancies. METHODS: A review of the ultrasound and pathology databases yielded 10 cases with 13 skin and subcutaneous tissue lesions secondary to hematologic neoplasms, which were confirmed by pathology. We used ultrasound to assess the number, location, size, depth of involvement, echogenicity, and vascularity of the lesions. RESULTS: The study involved five male and five female patients, including four leukemia, two multiple myeloma, and four lymphoma patients. The average age was 45 years (17-66 years). Three patients presented with one lesion, four with two lesions, and three with more than two lesions. All the lesions were located in the trunk and extremities. The lesions ranged from 1.2 to 8.3 cm in size. A total of 10 lesions involved subcutaneous fat tissue. A total of 10 lesions displayed hypoechoic foci within a hyperechoic background, and three appeared hypoechoic, and most of them exhibited abundant vascularity (12 of 13 lesions). CONCLUSIONS: Secondary involvement of skin and subcutaneous tissues by hematologic malignancies often present with multiple palpable masses showing the following ultrasound features: (1) subcutaneous fat infiltration, (2) hypoechoic foci with a hyperechoic background, and (3) abundant vascularity.

High-resolution CT findings of pulmonary infections in patients with hematologic malignancy: comparison between patients with or without hematopoietic stem cell transplantation.

PURPOSE: To evaluate the high-resolution CT (HRCT) findings of pulmonary infections in patients with hematologic malignancy and compare them between patients with or without hematopoietic stem cell transplantation (HSCT). MATERIALS AND METHODS: A total of 128 patients with hematologic malignancy and pulmonary infection were included in this study. The diagnoses of the patients consisted of bacterial pneumonia (37 non-HSCT cases and 14 HSCT cases), pneumocystis pneumonia (PCP) (29 non-HSCT cases and 11 HSCT cases), and fungal infection other than PCP (20 non-HSCT cases and 17 HSCT cases). Two chest radiologists retrospectively evaluated the HRCT criteria and compared them using chi-squared tests and a multiple logistic regression analysis. RESULTS: According to the multiple logistic regression analysis, nodules were an indicator in HSCT patients with PCP (p = 0.025; odds ratio, 5.8; 95% confidence interval, 1.2-26.6). The centrilobular distribution of nodules was the most frequent (n = 4, 36%) in HSCT patients with PCP. A mosaic pattern was an indicator of PCP in both HSCT and non-HSCT patients. There were no significant differences in other infections. CONCLUSION: The mosaic pattern could be an indicator of PCP in both HSCT and non-HSCT patients. Nodules with centrilobular distribution might be relatively frequent HRCT findings of PCP in HSCT patients.

Imaging of C-X-C Motif Chemokine Receptor 4 Expression in 690 Patients with Solid or Hematologic Neoplasms Using 68Ga-Pentixafor PET.

In recent years, molecular imaging addressing the C-X-C motif chemokine receptor 4 (CXCR4) has increasingly been used in various clinical settings. Here, we aimed to assess radiopharmaceutical uptake and image contrast to determine the most relevant clinical applications for CXCR4-directed imaging. We also investigated the impact of specific activity on scan contrast. Methods: Patients (n = 690) with a variety of neoplasms underwent a total of 777 PET/CT scans with 68Ga-Pentixafor, serving as the CXCR4-specific radioligand. A semiquantitative target lesion analysis was conducted (providing SUVmax and target-to-blood pool ratio [TBR], defined as SUVmax [from target lesion] divided by SUVmean [from blood pool]). The applied specific activity (in MBq/µg) was compared with semiquantitative assessments. Results: Of the 777 scans, 242 did not show discernible uptake in disease sites, leaving 535 PET scans (68.9%) for further analysis. Very high tracer uptake (SUVmax > 12) was found in multiple myeloma (n = 113), followed by adrenocortical carcinoma (n = 30), mantle cell lymphoma (n = 20), adrenocortical adenoma (n = 6), and small cell lung cancer (n = 12). Providing information on image contrast, comparable results for TBR were recorded, with TBR (>8) in multiple myeloma, mantle cell lymphoma, and acute lymphoblastoid leukemia (n = 6). When comparing specific activity with semiquantitative parameters, no significant correlation was found for SUVmax or TBR (P ≥ 0.612). Conclusion: In this large cohort, 68Ga-Pentixafor demonstrated high image contrast in a variety of neoplasms, particularly for hematologic malignancies, small cell lung cancer, and adrenocortical neoplasms. The present analysis may provide a roadmap for detecting patients who may benefit from CXCR4-targeted therapies.

Central nervous system manifestations of systemic haematological malignancies and key differentials.

Central nervous system (CNS) involvement by haematological malignancies is uncommon, and generally associated with a poor prognosis. Neuroimaging plays a key role in the accurate diagnosis, including in the critical differentiation from other processes such as infection and treatment-related toxicity. This review illustrates a variety of manifestations of CNS involvement by haematological malignancies and relevant differential diagnoses. CNS involvement can be seen in lymphoma (both primary and secondary), Waldenström macroglobulinaemia, multiple myeloma, leukaemia, and the malignant histiocytoses. The typical patterns vary between the different disorders, for example, in the most common sites of involvement and the relative frequency of parenchymal and meningeal involvement. Adjacent structures may also be involved. Nevertheless, there is some overlap in the imaging appearances, with common features including pre-contrast hyperdensity on computed tomography (CT), diffusion restriction, and avid post-contrast enhancement. In the post-treatment context, it is also important to distinguish between disease relapse and post-treatment effects. This includes opportunistic infections and the effects of chemotherapy and/or radiotherapy, including toxic effects and radiotherapy-induced neoplasms.

Mesial Temporal Sclerosis as Late Consequence of Posterior Reversible Encephalopathy Syndrome in Pediatric Hemato-oncologic Patients.

OBJECTIVES: Drug resistant epilepsy has rarely been reported following posterior reversible encephalopathy syndrome (PRES), with few cases of mesial temporal sclerosis (MTS). The aim of this study was to report clinical and neuroimaging features of MTS subsequent to PRES in hemato-oncologic/stem cell transplanted children. MATERIALS AND METHODS: Among 70 children treated in 2 pediatric hemato-oncologic Italian centers between 1994 and 2018 and presenting an episode of PRES, we retrospectively identified and analyzed a subgroup of patients who developed epilepsy and MTS. RESULTS: Nine of 70 patients (12.8%) developed post-PRES persistent seizures with magnetic resonance imaging evidence of MTS. One patient died few months after MTS diagnosis, because of hematologic complications; the remaining 8 patients showed unprovoked seizures over time leading to the diagnosis of epilepsy, focal in all and drug resistant in 4. At PRES diagnosis, all patients with further evidence of epilepsy and MTS suffered of convulsive seizures, evolving into status epilepticus in 3. In 3 patients a borderline cognitive level or intellectual disability were diagnosed after the onset of epilepsy, and 2 had behavioral problems impacting their quality of life. CONCLUSIONS: MTS and long-term focal epilepsy, along with potential cognitive and behavioral disorders, are not uncommon in older pediatric patients following PRES.

ImmunoPET imaging of hematological malignancies: From preclinical promise to clinical reality.

Immuno-positron emission tomography (immunoPET) imaging is a paradigm-shifting imaging technique for whole-body and all-lesion tumor detection, based on the combined specificity of tumor-targeting vectors [e.g., monoclonal antibodies (mAbs), nanobodies, and bispecific antibodies] and the sensitivity of PET imaging. By noninvasively, comprehensively, and serially revealing heterogeneous tumor antigen expression, immunoPET imaging is gradually improving the theranostic prospects for hematological malignancies. In this review, we summarize the available literature regarding immunoPET in imaging hematological malignancies. We also highlight the pros and cons of current conjugation strategies, and modular chemistry that can be leveraged to develop novel immunoPET probes for hematological malignancies. Lastly, we discuss the use of immunoPET imaging in guiding antibody drug development.

PET Imaging for Hematologic Malignancies.

Hematologic malignancies are a broad category of cancers arising from the lymphoid and myeloid cell lines. The 2016 World Health Organization classification system incorporated molecular markers as part of the diagnostic criteria and includes more than 100 subtypes. This article focuses on the subtypes for which imaging with positron emission tomography/computed tomography (PET/CT) has become an integral component of the patient's evaluation, that is, lymphoma and multiple myeloma. Leukemia and histiocytic and dendritic cell neoplasms are also discussed as these indications for PET/CT are less common, but increasingly seen in clinic.

FDG-PET/CT Variants and Pitfalls in Haematological Malignancies.

Hematologic malignancies represent a vast group of hematopoietic and lymphoid cancers that typically involve the blood, the bone marrow, and the lymphatic organs. Due to extensive research and well defined and standardized response criteria, the role of [18F]FDG-PET/CT is well defined in these malignancies. Never the less, the reliability of visual and quantitative interpretation of PET/CT may be impaired by several factors including inconsistent scanning protocols and image reconstruction methods. Furthermore, the uptake of [18F]FDG not only reflects tissue glucose consumption by malignant lesions, but also in other situations such as in inflammatory lesions, local and systemic infections, benign tumors, reactive thymic hyperplasia, histiocytic infiltration, among others; or following granulocyte colony stimulating factors therapy, radiation therapy, chemotherapy or surgical interventions, all of which are a potential source of false-positive or negative interpretations. Therefore it is of paramount importance for the Nuclear Medicine Physician to be familiar with, not only the normal distribution of [18F]FDG in the body, but also with the most frequent findings that may hamper a correct interpretation of the scan, which could ultimately alter the patients management. In this review, we describe these myriad of situations so the interpreting physician can be familiar with them, providing tools for their correct identification and interpretation when possible.

Assessment and Management of Cardiotoxicity in Hematologic Malignancies.

With the increasing overall survival of cancer patients due to recent discoveries in oncology, the incidence of side effects is also rising, and along with secondary malignancies, cardiotoxicity is one of the most concerning side effects, affecting the quality of life of cancer survivors. There are two types of cardiotoxicity associated with chemotherapy; the first one is acute, life-threatening but, fortunately, in most of the cases, reversible; and the second one is with late onset and mostly irreversible. The most studied drugs associated with cardiotoxicity are anthracyclines, but many new agents have demonstrated unexpected cardiotoxic effect, including those currently used in multiple myeloma treatment (proteasome inhibitors and immunomodulatory agents), tyrosine kinase inhibitors used in the treatment of chronic myeloid leukemia and some forms of acute leukemia, and immune checkpoint inhibitors recently introduced in treatment of refractory lymphoma patients. To prevent irreversible myocardial damage, early recognition of cardiac toxicity is mandatory. Traditional methods like echocardiography and magnetic resonance imaging are capable of detecting structural and functional changings, but unable to detect early myocardial damage; therefore, more sensible biomarkers like troponins and natriuretic peptides have to be introduced into the current practice. Baseline assessment of patients allows the identification of those with high risk for cardiotoxicity, while monitoring during and after treatment is important for early detection of cardiotoxicity and prompt intervention.

Association of Thoracic Computed Tomographic Measurements and Outcomes in Patients with Hematologic Malignancies Requiring Mechanical Ventilation.

Rationale: Patients with hematologic malignancies requiring mechanical ventilation have historically experienced poor outcomes.Objectives: We aimed to determine whether body composition characteristics derived from thoracic computed tomographic (CT) imaging were associated with time to liberation from mechanical ventilation.Methods: We evaluated mechanically ventilated patients with hematological malignancies admitted between 2014 and 2018. We included patients with thoracic CT imaging completed between 1 month before and 48 hours after intensive care unit (ICU) admission. We assessed the association of carinal skeletal muscle cross-sectional area (CSA), subcutaneous fat CSA, and fat index (fat/skeletal muscle ratio) with time to liberation from mechanical ventilation within 28 days. We accounted for the competing event of death within 28 days of mechanical ventilation.Results: One hundred fifty-six patients were included; the mean age was 57 years (standard deviation 14) and 39% were female. Thirty-seven percent had received a hematopoietic stem cell transplant, and the median ratio of arterial oxygen tension/pressure to fraction of inspired oxygen was 134 mm Hg (interquartile range [IQR], 92-205). Median skeletal muscle CSA was 68 cm2 (IQR, 54-88) and subcutaneous fat CSA was 38 cm2 (IQR, 27-52). Forty-two percent of patients were liberated from mechanical ventilation within 28 days and 56% died in the ICU. Subcutaneous fat CSA (subdistribution hazard ratio [sHR], 0.81; 95% confidence interval [95% CI], -0.68 to 0.97) and fat index (sHR, 0.81; 95% CI, -0.68 to 0.97) were significantly associated with longer time to liberation from mechanical ventilation. Skeletal muscle CSA was not associated with time to liberation from ventilation (sHR, 1.08; 95% CI, -0.94 to 1.23).Conclusions: Body composition measurements based on thoracic CT scans were associated with time to liberation from ventilation. These could represent novel surrogate markers of physical frailty in patients with hematologic malignancies receiving mechanical ventilation.

Application of Dynamic Enhanced Magnetic Resonance Imaging in the Diagnosis of Hematological Malignancies.

The use of dynamic enhanced magnetic resonance imaging technology can effectively explore the diagnosis and clinical application of hematological malignancies. This paper selected 60 patients with hematological malignancies from 2015 to 2019; all of whom were diagnosed with hematological malignancies, including 40 men and 20 women, aged between 40 and 77 years. The main clinical manifestations of the patient are hematological malignancies, fever, and other symptoms. We used Siemens 3.0T to perform MRI and dynamic enhanced MRI examinations on 30 patients with hematological tumors. The PACS system was used to collect and organize clinical data. All patients were pathologically confirmed and clinically diagnosed with hematological malignancies. Based on the clinical data of the patients, retrospective analysis and summary were conducted and the clinical manifestations of hematological malignancies were discussed. The results showed that the diagnostic accuracy of 30 cases of dynamic enhanced MRI was 100%, while the diagnostic accuracy of ordinary MRI was lower than that of dynamic enhanced MRI, P < 0.05, and the difference was statistically significant. In addition, compared with dynamic enhanced MRI and MRI, P > 0.05, the difference was not statistically significant. Therefore, the application of dynamic enhanced MRI in the diagnosis of hematological malignancies is valuable.

Cerebrospinal fluid findings in patients with hematologic neoplasms and meningeal infiltration.

Neoplastic cell infiltration into the central nervous system (CNS) is a serious complication of hematological neoplasms. Cytomorphology (CM) and flow cytometry (FC) have been used to detect meningeal infiltration. The association between CSF findings with the results of CM and FC is still poorly understood. We retrospectively evaluated CSF findings in 72 patients with hematological neoplasm and meningeal infiltration detected either by CM or FC. We compared CSF cell count, total protein concentration, and lactate concentration according to the type of hematological neoplasm. We also compared these CSF findings according to the FC and CM results (FC + CM + , FC + CM-, and FC-CM +). The proportion of patients with positive FC was higher than with CM (FC - 91.7%; CM - 63.9%). Thirty-five (48.6%) patients with meningeal infiltration had normal CSF cell count, normal total protein concentration, and normal lactate concentration. The proportion of cases in which these CSF parameters were normal did not differ according to the type of hematological neoplasm. The positivity of CM was significantly higher in patients with > 3 cell/mm3 (P = 0.015) but the positivity of FC was not significantly different between patients with > 3 cell/mm3 or ≤ 3 cells/mm3. Patients with positive CM had more CSF cells (P = 0.0005) and higher lactate concentration (P = 0.0165) than patients with negative CM. The absence of CSF changes in cell count and total protein and lactate concentrations does not exclude the presence of meningeal infiltration. Although CM is considered the gold standard, the probability of positive CM is low in patients without CSF abnormalities in these parameters. Patients with hematological neoplasm with suspected meningeal infiltration should be investigated with both methods.